SlideShare a Scribd company logo

Lda2vec text by the bay 2016 with notes

ïżœ
👋 Christopher Moody

The document discusses lda2vec, a model combining word2vec and LDA, focusing on how word vectors can capture semantic relationships and topic distributions in text data. It explains the mechanisms of word2vec for word prediction based on context and contrasts it with the global document structure of LDA. The presentation also emphasizes practical applications and computational efficiency, highlighting how these models can effectively represent language and meaning in machine learning tasks.

Technology
1 of 146
Downloaded 74 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
lda2vec (word2vec, and lda) Christopher Moody @ Stitch Fix Welcome, thanks for coming, having me, organizer NLP can be a messy aïŹ€air because you have to teach a computer about the irregularities and ambiguities of the English language and have to teach it this sort of hierarchical & sparse nature of english grammar & vocab 3rd trimester, pregnant “wears scrubs” — medicine taking a trip — a ïŹx for vacation clothing power and promise of word vectors is to sweep away a lot of issues
About @chrisemoody Caltech Physics PhD. in astrostats supercomputing sklearn t-SNE contributor Data Labs at Stitch Fix github.com/cemoody Gaussian Processes t-SNE chainer deep learning Tensor Decomposition
word2vec lda 1 2 3lda2vec
1. king - man + woman = queen 2. Huge splash in NLP world 3. Learns from raw text 4. Pretty simple algorithm 5. Comes pretrained word2vec 1. Learns what words mean — can solve analogies cleanly. 1. Not treating words as blocks, but instead modeling relationships 2. Distributed representations form the basis of more complicated deep learning systems 3.
1. Set up an objective function 2. Randomly initialize vectors 3. Do gradient descent word2vec
w ord2vec word2vec: learn word vector w from it’s surrounding context w 1. not mention neural networks 2. Let’s talk about training ïŹrst 3. n-grams transition probability vs tf-idf / LSI co-occurence matrices 4. Here we will learn the embedded representation directly, with no intermediates, update it w/ every example
w ord2vec “The fox jumped over the lazy dog” Maximize the likelihood of seeing the words given the word over. P(the|over) P(fox|over) P(jumped|over) P(the|over) P(lazy|over) P(dog|over) 
instead of maximizing the likelihood of co-occurrence counts. 1. Context — the words surrounding the training word 2. Naively assume, BoW, not recurrent, no state 3. Still a pretty simple assumption! Conditioning on just *over* no other secret parameters or anything
w ord2vec P(fox|over) What should this be?
w ord2vec P(vfox|vover) Should depend on the word vectors. P(fox|over) Trying to learn the word vectors, so let’s start with those (we’ll randomly initialize them to begin with)
w ord2vec “The fox jumped over the lazy dog” P(w|c) Extract pairs from context window around every input word.
w ord2vec “The fox jumped over the lazy dog” c P(w|c) Extract pairs from context window around every input word. IN = training word = context
w ord2vec “The fox jumped over the lazy dog” w P(w|c) c Extract pairs from context window around every input word.
w ord2vec P(w|c) w c “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
w ord2vec “The fox jumped over the lazy dog” P(w|c) w c Extract pairs from context window around every input word.
w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. inner most for-loop v_in was ïŹx over the for loop increment v_in to point to the next word
w ord2vec P(w|c) w c “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. called skip grams 
So that at a high level is what we want word2vec to do. two for loops
objective Measure loss between w and c? How should we deïŹne P(w|c)? Now we’ve deïŹned the high-level update path for the algorithm. Need to deïŹne this prob exactly in order to deïŹne our updates. Boils down to diïŹ€ between word & context — want to make as similar as possible, and then the probability will go up.
objective w . c How should we deïŹne P(w|c)? Measure loss between w and c? Use cosine sim. could imagine euclidean dist, mahalonobis dist
w ord2vec w . c ~ 1 objective w c vcanada . vsnow ~ 1 Dot product has these properties: Similar vectors have similarly near 1 (if normed)
w ord2vec w . c ~ 0 objective w c vcanada . vdesert ~0 Orthogonal vectors have similarity near 0
w ord2vec w . c ~ -1 objective w c Orthogonal vectors have similarity near -1
w ord2vec w . c ∈ [-1,1] objective But the inner product ranges from -1 to 1 (when normalized)
w ord2vec But we’d like to measure a probability. w . c ∈ [-1,1] objective 
and we’d like a probability
w ord2vec But we’d like to measure a probability. objective ∈ [0,1]σ(c·w) Transform again using sigmoid
w ord2vec But we’d like to measure a probability. objective ∈ [0,1]σ(c·w) w c w c SimilarDissimilar Transform again using sigmoid
w ord2vec Loss function: objective L=σ(c·w) Logistic (binary) choice. Is the (context, word) combination from our dataset? Are these 2 similar? This is optimized with identical vectors
 everything looks exactly the same
w ord2vec The skip-gram negative-sampling model objective Trivial solution is that context = word for all vectors L=σ(c·w) w c no contrast! add negative samples
w ord2vec The skip-gram negative-sampling model L = σ(c·w) + σ(-c·wneg) objective Draw random words in vocabulary. no contrast! add negative samples discrimanate that this (w, c) is from observed vocabulary, this is a randomly drawn word example: (fox, jumped) but (-fox, career) no popularity guessing as in softmax
w ord2vec The skip-gram negative-sampling model objective Discriminate positive from negative samples Multiple Negative L = σ(c·w) + σ(-c·wneg) +
+ σ(-c·wneg) mult samples (fox, jumped) not: (fox, career) (fox, android) (fox, sunlight) remind: this L is being computed, and we’ll nudge all of the values of c & w via GD to optimize L so that’s SGNS / w2v That’s it! A bit disservice to make this a giant network
w ord2vec The SGNS Model PM I ci·wj = PMI(Mij) - log k 
is extremely similar to matrix factorization! Levy & Goldberg 2014 L = σ(c·w) + σ(-c·wneg) explain what matrix is, row, col, entries e.g. can solve word2vec via SVD if we want deterministically One of the most cited NLP papers of 2014
w ord2vec The SGNS Model PM I Levy & Goldberg 2014 ‘traditional’ NLP L = σ(c·w) + σ(-c·wneg) ci·wj = PMI(Mij) - log k 
is extremely similar to matrix factorization! Dropped indices for readability most cited because connection to PMI Info-theoretic measure association of (w and c)
w ord2vec The SGNS Model L = σ(c·w) + ÎŁÏƒ(-c·w) PM I ci·wj = log Levy & Goldberg 2014 #(ci,wj)/n k #(wj)/n #(ci)/n ‘traditional’ NLP instead of looping over all words, you can count and roll them up the way props are just counts divided by number of obs word2vec adds this bias term cool that all of for looping did before can be reduced to this form
w ord2vec The SGNS Model L = σ(c·w) + ÎŁÏƒ(-c·w) PM I ci·wj = log Levy & Goldberg 2014 popularity of c,w k (popularity of c) (popularity of w) ‘traditional’ NLP props are just counts divided by number of obs word2vec adds this bias term down weights rare terms More frequent words are weighted more so than infrequent words. Rare words are downweighted. Make sense given that words distribution follows Zipf law. This weighting is what makes SGNS so powerful. theoretical is nice, but practical is even better
w ord2vec PM I 99% of word2vec is counting. And you can count words in SQL
w ord2vec PM I Count how many times you saw c·w Count how many times you saw c Count how many times you saw w
w ord2vec PM I 
and this takes ~5 minutes to compute on a single core. Computing SVD is a completely standard math library.
word2vec explain table
intuition about word vectors
Showing just 2 of the ~500 dimensions. EïŹ€ectively we’ve PCA’d it only 4 of 100k words
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
If we only had locality and not regularity, this wouldn’t necessarily be true
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
So we live in a vector space where operations like addition and subtraction are semantically meaningful. So here’s a few examples of this working. Really get the idea of these vectors as being ‘mixes’ of other ideas & vectors
ITEM_3469 + ‘Pregnant’ SF is a person service Box
+ ‘Pregnant’ I love the stripes and the cut around my neckline was amazing someone else might write ‘grey and black’ subtlety and nuance in that language For some items, we have several times the collected works of shakespeare
= ITEM_701333 = ITEM_901004 = ITEM_800456
Stripes and are safe for maternity And also similar tones and ïŹ‚owy — still great for expecting mothers
what about?LDA?
LDA on Client Item Descriptions This shows the incredible amount of structure
LDA on Item Descriptions (with Jay) clunky jewelry dangling delicate jewelry elsewhere
LDA on Item Descriptions (with Jay) topics on patterns, styles — this cluster is similarly described as high contrast tops with popping colors
LDA on Item Descriptions (with Jay) bright dresses for a warm summer LDA helps us model topics over documents in an interpretable way
lda vs word2vec
Bayesian Graphical ModelML Neural Model
word2vec is local: one word predicts a nearby word “I love ïŹnding new designer brands for jeans” as if the world where one very long text string. no end of documents, no end of sentence, etc. and a window across words
“I love ïŹnding new designer brands for jeans” But text is usually organized. as if the world where one very long text string. no end of documents, no end of sentence, etc.
“I love ïŹnding new designer brands for jeans” But text is usually organized. as if the world where one very long text string. no end of documents, no end of sentence, etc.
“I love ïŹnding new designer brands for jeans” In LDA, documents globally predict words. doc 7681 these are client comment which are short, only predict dozens of words but could be legal documents, or medical documents, 10k words
typical word2vec vector [ 0%, 9%, 78%, 11%] typical LDA document vector [ -0.75, -1.25, -0.55, -0.12, +2.2] All sum to 100%All real values
5D word2vec vector [ 0%, 9%, 78%, 11%] 5D LDA document vector [ -0.75, -1.25, -0.55, -0.12, +2.2] Sparse All sum to 100% Dimensions are absolute Dense All real values Dimensions relative much easier to say to another human 78% than it is +2.2 of something and -1.25 of something else w2v an address — 200 main st. — ïŹgure out from neighbors LDA is a *mixture* model 78% of some ingredient but w2v isn’t -1.25 of some ingredient ingredient = topics
100D word2vec vector [ 0%0%0%0%0% 
 0%, 9%, 78%, 11%] 100D LDA document vector [ -0.75, -1.25, -0.55, -0.27, -0.94, 0.44, 0.05, 0.31 
 -0.12, +2.2] Sparse All sum to 100% Dimensions are absolute Dense All real values Dimensions relative dense sparse lda is sparse, has 95 dims close to zero
100D word2vec vector [ 0%0%0%0%0% 
 0%, 9%, 78%, 11%] 100D LDA document vector [ -0.75, -1.25, -0.55, -0.27, -0.94, 0.44, 0.05, 0.31 
 -0.12, +2.2] Similar in fewer ways (more interpretable) Similar in 100D ways (very ïŹ‚exible) +mixture +sparse
can we do both? lda2vec series of exp grain of salt
-1.9 0.85 -0.6 -0.3 -0.5 Lufthansa is a German airline and when fox #hidden units Skip grams from sentences Word vector Negative sampling loss Lufthansa is a German airline and when German word2vec predicts locally: one word predicts a nearby word read example We extract pairs of pivot and target words that occur in a moving window that scans across the corpus. For every pair, the pivot word is used to predict the nearby target word.
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German Document vector predicts a word from a global context 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German: French or Spanish +airlines a document vector similar to the word vector for airline. we know we can add vectors German + airline: like Lufthansa, Condor Flugdienst, and Aero Lloyd. A latent vector is randomly initialized for every document in the corpus. Very similar to doc2vec and paragraph vectors.
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when We’re missing mixtures & sparsity! German Good for training sentiment models, great scores. interpretability
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when We’re missing mixtures & sparsity! German Too many documents. about as interpretable a hash
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Now it’s a mixture. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German document X is +0.34 in topic 0, -0.1 in topic 2 and 0.17 in topic 3 model with 3 topics. before had 500 DoF; now has just a few. better choose really good topics, because I only have a few available to summarize the entire document
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Trinitarian baptismal Pentecostals Bede schismatics excommunication 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight Each topic has a distributed representation that lives in the same space as the word vectors. While each topic is not literally a token present in the corpus, it is similar to other tokens.
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when topic 1 = “religion” Trinitarian baptismal Pentecostals Bede schismatics excommunication 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Milosevic absentee Indonesia Lebanese Isrealis Karadzic 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight notice one column over in the topic matrix
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when topic 2 = “politics” Milosevic absentee Indonesia Lebanese Isrealis Karadzic 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight topic vectors, document vectors, and word vectors all live in the same space
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when The document weights are softmax transformed weights to yield the document proportions. similar to logistic, but instead of 0-1 now a vectors of % 100% and indicates the topic proportions of a single document one document might be 41% in topic 0, 26% in topic 1, and 34% in topic 3 very close to LDA-like representations
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 1st time i did this, still very dense in percentages. so if 100 topics, had 1% each might be, but still dense! a zillion categories, 1% of this, 1% of this, 1% of this
 mathematically works, addition of lots of little bits (distributed)
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Sparsity! 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 34% 32% 34% t=0 41% 26% 34% t=10 99% 1% 0% t=∞ tim e init balanced, but dense Dirichlet likelihood loss encourages proportions vectors to become sparser over time. relatively simple start pink, go to white and sparse
0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German end up with something that’s quite a bit more complicated but it achieves our goals: mixes word vectors w/ sparse interpretable document representations
@chrisemoody Example Hacker News comments Word vectors: https://github.com/cemoody/ lda2vec/blob/master/examples/ hacker_news/lda2vec/ word_vectors.ipynb read examples play around at home — on
@chrisemoody Example Hacker News comments Topics: http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/ hacker_news/lda2vec/lda2vec.ipynb http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/hacker_news/lda2vec/lda2vec.ipynb topic 16 — sci, phys topic 1 — housing topic 8 — ïŹnance, bitcoin topic 23 — programming languages topic 6 — transportation topic 3 — education
@chrisemoody lda2vec.com http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
+ API docs + Examples + GPU + Tests @chrisemoody lda2vec.com http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
@chrisemoody lda2vec.com human-interpretable doc topics, use LDA. machine-useable word-level features, use word2vec. if you like to experiment a lot, and have topics over user / doc / region / etc. features, use lda2vec. (and you have a GPU) If you want
 http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
?@chrisemoody Multithreaded Stitch Fix
@chrisemoody lda2vec.com http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
Credit Large swathes of this talk are from previous presentations by: ‱ Tomas Mikolov ‱ David Blei ‱ Christopher Olah ‱ Radim Rehurek ‱ Omer Levy & Yoav Goldberg ‱ Richard Socher ‱ Xin Rong ‱ Tim Hopper Richar & Xin Rong both ludic explanations of the word2vec gradient
“PS! Thank you for such an awesome idea” @chrisemoody doc_id=1846 Can we model topics to sentences? lda2lstm Data Labs @ SF is all about mixing cutting edge algorithms but we absolutely need interpretability. initial vector is a dirichlet mixture — moves us from bag of words to sentence-level LDA give a sent that’s 80% religion, 10% politics word2vec on word level, LSTM on the sentence level, LDA on document level Dirichlet-squeeze internal states and manipulations, that’ll help us understand the science of LSTM dynamics
Can we model topics to sentences? lda2lstm “PS! Thank you for such an awesome idea”doc_id=1846 @chrisemoody Can we model topics to images? lda2ae TJ Torres
and now for something completely crazy 4 Fun StuïŹ€
translation (using just a rotation matrix) M ikolov 2013 English Spanish Matrix Rotation Blow mind Explain plot Not a complicated NN here Still have to learn the rotation matrix — but it generalizes very nicely. Have analogies for every linalg op as a linguistic operator Robust framework and tools to do science on words
deepwalk Perozzi etal2014 learn word vectors from sentences “The fox jumped over the lazy dog” vOUT vOUT vOUT vOUT vOUTvOUT ‘words’ are graph vertices ‘sentences’ are random walks on the graph word2vec
Playlists at Spotify context sequence learning ‘words’ are song indices ‘sentences’ are playlists
Playlists at Spotify context Erik Bernhardsson Great performance on ‘related artists’
Fixes at Stitch Fix sequence learning Let’s try: ‘words’ are items ‘sentences’ are ïŹxes
Fixes at Stitch Fix context Learn similarity between styles because they co-occur Learn ‘coherent’ styles sequence learning
Fixes at Stitch Fix? context sequence learning Got lots of structure!
Fixes at Stitch Fix? context sequence learning
Fixes at Stitch Fix? context sequence learning Nearby regions are consistent ‘closets’ What sorts of sequences do you have at Quora? What kinds of things can you learn from context?
?@chrisemoody Multithreaded Stitch Fix
context dependent Levy & G oldberg 2014 Australian scientist discovers star with telescope context +/- 2 words
context dependent context Australian scientist discovers star with telescope Levy & G oldberg 2014 What if we
context dependent context Australian scientist discovers star with telescope context Levy & G oldberg 2014
context dependent context BoW DEPS topically-similar vs ‘functionally’ similar Levy & G oldberg 2014
?@chrisemoody Multithreaded Stitch Fix
Lda2vec text by the bay 2016 with notes
Crazy Approaches Paragraph Vectors (Just extend the context window) Content dependency (Change the window grammatically) Social word2vec (deepwalk) (Sentence is a walk on the graph) Spotify (Sentence is a playlist of song_ids) Stitch Fix (Sentence is a shipment of ïŹve items)
See previous
CBOW “The fox jumped over the lazy dog” Guess the word given the context ~20x faster. (this is the alternative.) vOUT vIN vINvIN vIN vIN vIN SkipGram “The fox jumped over the lazy dog” vOUT vOUT vIN vOUT vOUT vOUTvOUT Guess the context given the word Better at syntax. (this is the one we went over) CBOW sums words vectors, loses the order in the sentence Both are good at semantic relationships Child and kid are nearby Or gender in man, woman If you blur words over the scale of context — 5ish words, you lose a lot grammatical nuance But skipgram preserves order Preserves the relationship in pluralizing, for example
lda2vec vDOC = a vtopic1 + b vtopic2 +
 Let’s make vDOC sparse Too many documents. I really like that document X is 70% in topic 0, 30% in topic1, 

lda2vec This works! 😀 But vDOC isn’t as interpretable as the topic vectors. 😔 vDOC = topic0 + topic1 Let’s say that vDOC ads Too many documents. I really like that document X is 70% in topic 0, 30% in topic1, 

lda2vec softmax(vOUT * (vIN+ vDOC)) we want k *sparse* topics
Shows that are many words similar to vacation actually come in lots of ïŹ‚avors — wedding words (bachelorette, rehearsals) — holiday/event words (birthdays, brunch, christmas, thanksgiving) — seasonal words (spring, summer,) — trip words (getaway) — destinations
theory of lda2vec lda2vec
pyLDAvis of lda2vec lda2vec
LDA Results context H istory I loved every choice in this ïŹx!! Great job! Great Stylist Perfect There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
LDA Results context H istory Body Fit My measurements are 36-28-32. If that helps. I like wearing some clothing that is ïŹtted. Very hard for me to ïŹnd pants that ïŹt right. There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
LDA Results context H istory Sizing Really enjoyed the experience and the pieces, sizing for tops was too big. Looking forward to my next box! Excited for next There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
LDA Results context H istory Almost Bought It was a great ïŹx. Loved the two items I kept and the three I sent back were close! Perfect There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
All of the following ideas will change what ‘words’ and ‘context’ represent. But we’ll still use the same w2v algo
paragraph vector What about summarizing documents? On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that
On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that The framework nuclear agreement he reached with Iran on Thursday did not provide the deïŹnitive answer to whether Mr. Obama’s audacious gamble will pay off. The ïŹst Iran has shaken at the so-called Great Satan since 1979 has not completely relaxed. paragraph vector Normal skipgram extends C words before, and C words after. IN OUT OUT Except we stay inside a sentence
On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that The framework nuclear agreement he reached with Iran on Thursday did not provide the deïŹnitive answer to whether Mr. Obama’s audacious gamble will pay off. The ïŹst Iran has shaken at the so-called Great Satan since 1979 has not completely relaxed. paragraph vector A document vector simply extends the context to the whole document. IN OUT OUT OUT OUTdoc_1347
from gensim.models import Doc2Vec fn = “item_document_vectors” model = Doc2Vec.load(fn) model.most_similar('pregnant') matches = list(filter(lambda x: 'SENT_' in x[0], matches)) # ['...I am currently 23 weeks pregnant...', # '...I'm now 10 weeks pregnant...', # '...not showing too much yet...', # '...15 weeks now. Baby bump...', # '...6 weeks post partum!...', # '...12 weeks postpartum and am nursing...', # '...I have my baby shower that...', # '...am still breastfeeding...', # '...I would love an outfit for a baby shower...'] sentence search
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes
Lda2vec text by the bay 2016 with notes

More Related Content

PDF
lda2vec Text by the Bay 2016
👋 Christopher Moody
 
PDF
word2vec, LDA, and introducing a new hybrid algorithm: lda2vec
👋 Christopher Moody
 
PDF
word2vec, LDA, and introducing a new hybrid algorithm: lda2vec
👋 Christopher Moody
 
PDF
AI at Stitch Fix 2017
👋 Christopher Moody
 
PDF
Introduction to word embeddings with Python
Pavel Kalaidin
 
PDF
Word2vec ultimate beginner
Sungmin Yang
 
PDF
Word2vec and Friends
Bruno Gonçalves
 
PPT
13 Strings and text processing
maznabili
 
lda2vec Text by the Bay 2016
👋 Christopher Moody
 
word2vec, LDA, and introducing a new hybrid algorithm: lda2vec
👋 Christopher Moody
 
word2vec, LDA, and introducing a new hybrid algorithm: lda2vec
👋 Christopher Moody
 
AI at Stitch Fix 2017
👋 Christopher Moody
 
Introduction to word embeddings with Python
Pavel Kalaidin
 
Word2vec ultimate beginner
Sungmin Yang
 
Word2vec and Friends
Bruno Gonçalves
 
13 Strings and text processing
maznabili
 

Viewers also liked (10)

PDF
Discussion on the Distributed Search Engine
Yusuke Fujisaka
 
PDF
P2p search engine
Yusuke Fujisaka
 
PDF
Fabrikatyr lda topic modelling practical application
Tim Carnus
 
PPTX
Topic Modelling to identify behavioral trends in online communities
Conor Duke
 
PDF
Drawing word2vec
Kai Sasaki
 
PDF
EMNLP2014èȘ­ăżäŒš "Efficient Non-parametric Estimation of Multiple Embeddings per ...
Yuya Unno
 
PPTX
EmnlpèȘ­ăżäŒšèł‡æ–™
Jiro Nishitoba
 
PDF
[Yang, Downey and Boyd-Graber 2015] Efficient Methods for Incorporating Knowl...
Shuyo Nakatani
 
PDF
Running Word2Vec with Chinese Wikipedia dump
Billy Yang
 
POTX
LDA Beginner's Tutorial
Wayne Lee
 
Discussion on the Distributed Search Engine
Yusuke Fujisaka
 
P2p search engine
Yusuke Fujisaka
 
Fabrikatyr lda topic modelling practical application
Tim Carnus
 
Topic Modelling to identify behavioral trends in online communities
Conor Duke
 
Drawing word2vec
Kai Sasaki
 
EMNLP2014èȘ­ăżäŒš "Efficient Non-parametric Estimation of Multiple Embeddings per ...
Yuya Unno
 
EmnlpèȘ­ăżäŒšèł‡æ–™
Jiro Nishitoba
 
[Yang, Downey and Boyd-Graber 2015] Efficient Methods for Incorporating Knowl...
Shuyo Nakatani
 
Running Word2Vec with Chinese Wikipedia dump
Billy Yang
 
LDA Beginner's Tutorial
Wayne Lee
 
Ad

Similar to Lda2vec text by the bay 2016 with notes (20)

PDF
Yoav Goldberg: Word Embeddings What, How and Whither
MLReview
 
PPTX
Lecture1.pptx
jonathanG19
 
PPTX
A Panorama of Natural Language Processing
Ted Xiao
 
PPTX
word vector embeddings in natural languag processing
ReetShinde
 
PPTX
L6.pptxsdv dfbdfjftj hgjythgfvfhjyggunghb fghtffn
RwanEnan
 
PPTX
CVDL Unit-5.pptx this is relates to computer vision
srinathbadugu7070
 
PPTX
Word2 vec
ankit_ppt
 
PPTX
Text Mining for Lexicography
Leiden University
 
PPTX
Text features
Shruti kar
 
PPTX
What is word2vec?
Traian Rebedea
 
PPTX
DL-CO2 -Session 3 Learning Vectorial Representations of Words.pptx
Kv Sagar
 
PDF
Word2vec and Friends
Bruno Gonçalves
 
PDF
MACHINE-DRIVEN TEXT ANALYSIS
Massimo Schenone
 
PPTX
Word2vec slide(lab seminar)
Jinpyo Lee
 
PPTX
Deep Learning Bangalore meet up
Satyam Saxena
 
PPTX
DLBLR talk
Anuj Gupta
 
PDF
Word2Vec
hyunyoung Lee
 
PPTX
Word embedding
ShivaniChoudhary74
 
PDF
(Deep) Neural Networks朹 NLP 撌 Text Mining 总结
搛 ć»–
 
PPTX
Tutorial on word2vec
Leiden University
 
Yoav Goldberg: Word Embeddings What, How and Whither
MLReview
 
Lecture1.pptx
jonathanG19
 
A Panorama of Natural Language Processing
Ted Xiao
 
word vector embeddings in natural languag processing
ReetShinde
 
L6.pptxsdv dfbdfjftj hgjythgfvfhjyggunghb fghtffn
RwanEnan
 
CVDL Unit-5.pptx this is relates to computer vision
srinathbadugu7070
 
Word2 vec
ankit_ppt
 
Text Mining for Lexicography
Leiden University
 
Text features
Shruti kar
 
What is word2vec?
Traian Rebedea
 
DL-CO2 -Session 3 Learning Vectorial Representations of Words.pptx
Kv Sagar
 
Word2vec and Friends
Bruno Gonçalves
 
MACHINE-DRIVEN TEXT ANALYSIS
Massimo Schenone
 
Word2vec slide(lab seminar)
Jinpyo Lee
 
Deep Learning Bangalore meet up
Satyam Saxena
 
DLBLR talk
Anuj Gupta
 
Word2Vec
hyunyoung Lee
 
Word embedding
ShivaniChoudhary74
 
(Deep) Neural Networks朹 NLP 撌 Text Mining 总结
搛 ć»–
 
Tutorial on word2vec
Leiden University
 
Ad

Recently uploaded (20)

PDF
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
PDF
Per capita expenditure prediction using model stacking based on satellite ima...
IAESIJAI
 
PPTX
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
PDF
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
Libreria ERP
 
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
PDF
Encapsulation_ Review paper, used for researhc scholars
gurumoop
 
PPTX
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
PDF
How UI/UX Design Impacts User Retention in Mobile Apps.pdf
SynapseIndia
 
PDF
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
PPTX
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
PDF
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 
PPT
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
PPTX
Cloud computing and distributed systems.
kkkkkhan
 
PPTX
sap open course for s4hana steps from ECC to s4
sreeni2106invites
 
PDF
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
Julien SIMON
 
PPTX
Spectroscopy.pptx food analysis technology
nawahassan45
 
PDF
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
PDF
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
PDF
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
Per capita expenditure prediction using model stacking based on satellite ima...
IAESIJAI
 
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
Libreria ERP
 
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
Encapsulation_ Review paper, used for researhc scholars
gurumoop
 
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
How UI/UX Design Impacts User Retention in Mobile Apps.pdf
SynapseIndia
 
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
Cloud computing and distributed systems.
kkkkkhan
 
sap open course for s4hana steps from ECC to s4
sreeni2106invites
 
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
Julien SIMON
 
Spectroscopy.pptx food analysis technology
nawahassan45
 
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 

Lda2vec text by the bay 2016 with notes

  • 1. lda2vec (word2vec, and lda) Christopher Moody @ Stitch Fix Welcome, thanks for coming, having me, organizer NLP can be a messy aïŹ€air because you have to teach a computer about the irregularities and ambiguities of the English language and have to teach it this sort of hierarchical & sparse nature of english grammar & vocab 3rd trimester, pregnant “wears scrubs” — medicine taking a trip — a ïŹx for vacation clothing power and promise of word vectors is to sweep away a lot of issues
  • 2. About @chrisemoody Caltech Physics PhD. in astrostats supercomputing sklearn t-SNE contributor Data Labs at Stitch Fix github.com/cemoody Gaussian Processes t-SNE chainer deep learning Tensor Decomposition
  • 4. 1. king - man + woman = queen 2. Huge splash in NLP world 3. Learns from raw text 4. Pretty simple algorithm 5. Comes pretrained word2vec 1. Learns what words mean — can solve analogies cleanly. 1. Not treating words as blocks, but instead modeling relationships 2. Distributed representations form the basis of more complicated deep learning systems 3.
  • 5. 1. Set up an objective function 2. Randomly initialize vectors 3. Do gradient descent word2vec
  • 6. w ord2vec word2vec: learn word vector w from it’s surrounding context w 1. not mention neural networks 2. Let’s talk about training ïŹrst 3. n-grams transition probability vs tf-idf / LSI co-occurence matrices 4. Here we will learn the embedded representation directly, with no intermediates, update it w/ every example
  • 7. w ord2vec “The fox jumped over the lazy dog” Maximize the likelihood of seeing the words given the word over. P(the|over) P(fox|over) P(jumped|over) P(the|over) P(lazy|over) P(dog|over) 
instead of maximizing the likelihood of co-occurrence counts. 1. Context — the words surrounding the training word 2. Naively assume, BoW, not recurrent, no state 3. Still a pretty simple assumption! Conditioning on just *over* no other secret parameters or anything
  • 9. w ord2vec P(vfox|vover) Should depend on the word vectors. P(fox|over) Trying to learn the word vectors, so let’s start with those (we’ll randomly initialize them to begin with)
  • 10. w ord2vec “The fox jumped over the lazy dog” P(w|c) Extract pairs from context window around every input word.
  • 11. w ord2vec “The fox jumped over the lazy dog” c P(w|c) Extract pairs from context window around every input word. IN = training word = context
  • 12. w ord2vec “The fox jumped over the lazy dog” w P(w|c) c Extract pairs from context window around every input word.
  • 13. w ord2vec P(w|c) w c “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
  • 14. w ord2vec “The fox jumped over the lazy dog” P(w|c) w c Extract pairs from context window around every input word.
  • 15. w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
  • 16. w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
  • 17. w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. inner most for-loop v_in was ïŹx over the for loop increment v_in to point to the next word
  • 18. w ord2vec P(w|c) w c “The fox jumped over the lazy dog” Extract pairs from context window around every input word.
  • 19. w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
  • 20. w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
  • 21. w ord2vec P(w|c) cw “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
  • 22. w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. 
So that at a high level is what we want word2vec to do.
  • 23. w ord2vec P(w|c) c w “The fox jumped over the lazy dog” Extract pairs from context window around every input word. called skip grams 
So that at a high level is what we want word2vec to do. two for loops
  • 24. objective Measure loss between w and c? How should we deïŹne P(w|c)? Now we’ve deïŹned the high-level update path for the algorithm. Need to deïŹne this prob exactly in order to deïŹne our updates. Boils down to diïŹ€ between word & context — want to make as similar as possible, and then the probability will go up.
  • 25. objective w . c How should we deïŹne P(w|c)? Measure loss between w and c? Use cosine sim. could imagine euclidean dist, mahalonobis dist
  • 26. w ord2vec w . c ~ 1 objective w c vcanada . vsnow ~ 1 Dot product has these properties: Similar vectors have similarly near 1 (if normed)
  • 27. w ord2vec w . c ~ 0 objective w c vcanada . vdesert ~0 Orthogonal vectors have similarity near 0
  • 28. w ord2vec w . c ~ -1 objective w c Orthogonal vectors have similarity near -1
  • 29. w ord2vec w . c ∈ [-1,1] objective But the inner product ranges from -1 to 1 (when normalized)
  • 30. w ord2vec But we’d like to measure a probability. w . c ∈ [-1,1] objective 
and we’d like a probability
  • 31. w ord2vec But we’d like to measure a probability. objective ∈ [0,1]σ(c·w) Transform again using sigmoid
  • 32. w ord2vec But we’d like to measure a probability. objective ∈ [0,1]σ(c·w) w c w c SimilarDissimilar Transform again using sigmoid
  • 33. w ord2vec Loss function: objective L=σ(c·w) Logistic (binary) choice. Is the (context, word) combination from our dataset? Are these 2 similar? This is optimized with identical vectors
 everything looks exactly the same
  • 34. w ord2vec The skip-gram negative-sampling model objective Trivial solution is that context = word for all vectors L=σ(c·w) w c no contrast! add negative samples
  • 35. w ord2vec The skip-gram negative-sampling model L = σ(c·w) + σ(-c·wneg) objective Draw random words in vocabulary. no contrast! add negative samples discrimanate that this (w, c) is from observed vocabulary, this is a randomly drawn word example: (fox, jumped) but (-fox, career) no popularity guessing as in softmax
  • 36. w ord2vec The skip-gram negative-sampling model objective Discriminate positive from negative samples Multiple Negative L = σ(c·w) + σ(-c·wneg) +
+ σ(-c·wneg) mult samples (fox, jumped) not: (fox, career) (fox, android) (fox, sunlight) remind: this L is being computed, and we’ll nudge all of the values of c & w via GD to optimize L so that’s SGNS / w2v That’s it! A bit disservice to make this a giant network
  • 37. w ord2vec The SGNS Model PM I ci·wj = PMI(Mij) - log k 
is extremely similar to matrix factorization! Levy & Goldberg 2014 L = σ(c·w) + σ(-c·wneg) explain what matrix is, row, col, entries e.g. can solve word2vec via SVD if we want deterministically One of the most cited NLP papers of 2014
  • 38. w ord2vec The SGNS Model PM I Levy & Goldberg 2014 ‘traditional’ NLP L = σ(c·w) + σ(-c·wneg) ci·wj = PMI(Mij) - log k 
is extremely similar to matrix factorization! Dropped indices for readability most cited because connection to PMI Info-theoretic measure association of (w and c)
  • 39. w ord2vec The SGNS Model L = σ(c·w) + ÎŁÏƒ(-c·w) PM I ci·wj = log Levy & Goldberg 2014 #(ci,wj)/n k #(wj)/n #(ci)/n ‘traditional’ NLP instead of looping over all words, you can count and roll them up the way props are just counts divided by number of obs word2vec adds this bias term cool that all of for looping did before can be reduced to this form
  • 40. w ord2vec The SGNS Model L = σ(c·w) + ÎŁÏƒ(-c·w) PM I ci·wj = log Levy & Goldberg 2014 popularity of c,w k (popularity of c) (popularity of w) ‘traditional’ NLP props are just counts divided by number of obs word2vec adds this bias term down weights rare terms More frequent words are weighted more so than infrequent words. Rare words are downweighted. Make sense given that words distribution follows Zipf law. This weighting is what makes SGNS so powerful. theoretical is nice, but practical is even better
  • 41. w ord2vec PM I 99% of word2vec is counting. And you can count words in SQL
  • 42. w ord2vec PM I Count how many times you saw c·w Count how many times you saw c Count how many times you saw w
  • 43. w ord2vec PM I 
and this takes ~5 minutes to compute on a single core. Computing SVD is a completely standard math library.
  • 46. Showing just 2 of the ~500 dimensions. EïŹ€ectively we’ve PCA’d it only 4 of 100k words
  • 57. If we only had locality and not regularity, this wouldn’t necessarily be true
  • 60. So we live in a vector space where operations like addition and subtraction are semantically meaningful. So here’s a few examples of this working. Really get the idea of these vectors as being ‘mixes’ of other ideas & vectors
  • 61. ITEM_3469 + ‘Pregnant’ SF is a person service Box
  • 62. + ‘Pregnant’ I love the stripes and the cut around my neckline was amazing someone else might write ‘grey and black’ subtlety and nuance in that language For some items, we have several times the collected works of shakespeare
  • 64. Stripes and are safe for maternity And also similar tones and ïŹ‚owy — still great for expecting mothers
  • 66. LDA on Client Item Descriptions This shows the incredible amount of structure
  • 67. LDA on Item Descriptions (with Jay) clunky jewelry dangling delicate jewelry elsewhere
  • 68. LDA on Item Descriptions (with Jay) topics on patterns, styles — this cluster is similarly described as high contrast tops with popping colors
  • 69. LDA on Item Descriptions (with Jay) bright dresses for a warm summer LDA helps us model topics over documents in an interpretable way
  • 72. word2vec is local: one word predicts a nearby word “I love ïŹnding new designer brands for jeans” as if the world where one very long text string. no end of documents, no end of sentence, etc. and a window across words
  • 73. “I love ïŹnding new designer brands for jeans” But text is usually organized. as if the world where one very long text string. no end of documents, no end of sentence, etc.
  • 74. “I love ïŹnding new designer brands for jeans” But text is usually organized. as if the world where one very long text string. no end of documents, no end of sentence, etc.
  • 75. “I love ïŹnding new designer brands for jeans” In LDA, documents globally predict words. doc 7681 these are client comment which are short, only predict dozens of words but could be legal documents, or medical documents, 10k words
  • 76. typical word2vec vector [ 0%, 9%, 78%, 11%] typical LDA document vector [ -0.75, -1.25, -0.55, -0.12, +2.2] All sum to 100%All real values
  • 77. 5D word2vec vector [ 0%, 9%, 78%, 11%] 5D LDA document vector [ -0.75, -1.25, -0.55, -0.12, +2.2] Sparse All sum to 100% Dimensions are absolute Dense All real values Dimensions relative much easier to say to another human 78% than it is +2.2 of something and -1.25 of something else w2v an address — 200 main st. — ïŹgure out from neighbors LDA is a *mixture* model 78% of some ingredient but w2v isn’t -1.25 of some ingredient ingredient = topics
  • 78. 100D word2vec vector [ 0%0%0%0%0% 
 0%, 9%, 78%, 11%] 100D LDA document vector [ -0.75, -1.25, -0.55, -0.27, -0.94, 0.44, 0.05, 0.31 
 -0.12, +2.2] Sparse All sum to 100% Dimensions are absolute Dense All real values Dimensions relative dense sparse lda is sparse, has 95 dims close to zero
  • 79. 100D word2vec vector [ 0%0%0%0%0% 
 0%, 9%, 78%, 11%] 100D LDA document vector [ -0.75, -1.25, -0.55, -0.27, -0.94, 0.44, 0.05, 0.31 
 -0.12, +2.2] Similar in fewer ways (more interpretable) Similar in 100D ways (very ïŹ‚exible) +mixture +sparse
  • 80. can we do both? lda2vec series of exp grain of salt
  • 81. -1.9 0.85 -0.6 -0.3 -0.5 Lufthansa is a German airline and when fox #hidden units Skip grams from sentences Word vector Negative sampling loss Lufthansa is a German airline and when German word2vec predicts locally: one word predicts a nearby word read example We extract pairs of pivot and target words that occur in a moving window that scans across the corpus. For every pair, the pivot word is used to predict the nearby target word.
  • 82. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German Document vector predicts a word from a global context 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German: French or Spanish +airlines a document vector similar to the word vector for airline. we know we can add vectors German + airline: like Lufthansa, Condor Flugdienst, and Aero Lloyd. A latent vector is randomly initialized for every document in the corpus. Very similar to doc2vec and paragraph vectors.
  • 83. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when We’re missing mixtures & sparsity! German Good for training sentiment models, great scores. interpretability
  • 84. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when We’re missing mixtures & sparsity! German Too many documents. about as interpretable a hash
  • 85. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Now it’s a mixture. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German document X is +0.34 in topic 0, -0.1 in topic 2 and 0.17 in topic 3 model with 3 topics. before had 500 DoF; now has just a few. better choose really good topics, because I only have a few available to summarize the entire document
  • 86. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Trinitarian baptismal Pentecostals Bede schismatics excommunication 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight Each topic has a distributed representation that lives in the same space as the word vectors. While each topic is not literally a token present in the corpus, it is similar to other tokens.
  • 87. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when topic 1 = “religion” Trinitarian baptismal Pentecostals Bede schismatics excommunication 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight
  • 88. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Milosevic absentee Indonesia Lebanese Isrealis Karadzic 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight notice one column over in the topic matrix
  • 89. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when topic 2 = “politics” Milosevic absentee Indonesia Lebanese Isrealis Karadzic 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 #topics Document weight topic vectors, document vectors, and word vectors all live in the same space
  • 90. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when
  • 91. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when The document weights are softmax transformed weights to yield the document proportions. similar to logistic, but instead of 0-1 now a vectors of % 100% and indicates the topic proportions of a single document one document might be 41% in topic 0, 26% in topic 1, and 34% in topic 3 very close to LDA-like representations
  • 92. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 1st time i did this, still very dense in percentages. so if 100 topics, had 1% each might be, but still dense! a zillion categories, 1% of this, 1% of this, 1% of this
 mathematically works, addition of lots of little bits (distributed)
  • 93. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when Sparsity! 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when 34% 32% 34% t=0 41% 26% 34% t=10 99% 1% 0% t=∞ tim e init balanced, but dense Dirichlet likelihood loss encourages proportions vectors to become sparser over time. relatively simple start pink, go to white and sparse
  • 94. 0.34 -0.1 0.17 41% 26% 34% -1.4 -0.5 -1.4 -1.9-1.7 0.75 0.96-0.7 -1.9 -0.2-1.1 0.6 -0.7 -0.4 -0.7 -0.3 -0.3-1.9 0.85 -0.6 -0.3 -0.5 -2.6 0.45 -1.3 -0.6 -0.8 Lufthansa is a German airline and when #topics #topics fox #hiddenunits #topics #hidden units#hidden units #hidden units Skip grams from sentences Word vector Negative sampling loss Topic matrix Document proportion Document weight Document vector Context vector x + Lufthansa is a German airline and when German end up with something that’s quite a bit more complicated but it achieves our goals: mixes word vectors w/ sparse interpretable document representations
  • 95. @chrisemoody Example Hacker News comments Word vectors: https://github.com/cemoody/ lda2vec/blob/master/examples/ hacker_news/lda2vec/ word_vectors.ipynb read examples play around at home — on
  • 96. @chrisemoody Example Hacker News comments Topics: http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/ hacker_news/lda2vec/lda2vec.ipynb http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/hacker_news/lda2vec/lda2vec.ipynb topic 16 — sci, phys topic 1 — housing topic 8 — ïŹnance, bitcoin topic 23 — programming languages topic 6 — transportation topic 3 — education
  • 98. + API docs + Examples + GPU + Tests @chrisemoody lda2vec.com http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
  • 99. @chrisemoody lda2vec.com human-interpretable doc topics, use LDA. machine-useable word-level features, use word2vec. if you like to experiment a lot, and have topics over user / doc / region / etc. features, use lda2vec. (and you have a GPU) If you want
 http://nbviewer.jupyter.org/github/cemoody/lda2vec/blob/master/examples/twenty_newsgroups/lda.ipynb
  • 102. Credit Large swathes of this talk are from previous presentations by: ‱ Tomas Mikolov ‱ David Blei ‱ Christopher Olah ‱ Radim Rehurek ‱ Omer Levy & Yoav Goldberg ‱ Richard Socher ‱ Xin Rong ‱ Tim Hopper Richar & Xin Rong both ludic explanations of the word2vec gradient
  • 103. “PS! Thank you for such an awesome idea” @chrisemoody doc_id=1846 Can we model topics to sentences? lda2lstm Data Labs @ SF is all about mixing cutting edge algorithms but we absolutely need interpretability. initial vector is a dirichlet mixture — moves us from bag of words to sentence-level LDA give a sent that’s 80% religion, 10% politics word2vec on word level, LSTM on the sentence level, LDA on document level Dirichlet-squeeze internal states and manipulations, that’ll help us understand the science of LSTM dynamics
  • 104. Can we model topics to sentences? lda2lstm “PS! Thank you for such an awesome idea”doc_id=1846 @chrisemoody Can we model topics to images? lda2ae TJ Torres
  • 105. and now for something completely crazy 4 Fun StuïŹ€
  • 106. translation (using just a rotation matrix) M ikolov 2013 English Spanish Matrix Rotation Blow mind Explain plot Not a complicated NN here Still have to learn the rotation matrix — but it generalizes very nicely. Have analogies for every linalg op as a linguistic operator Robust framework and tools to do science on words
  • 107. deepwalk Perozzi etal2014 learn word vectors from sentences “The fox jumped over the lazy dog” vOUT vOUT vOUT vOUT vOUTvOUT ‘words’ are graph vertices ‘sentences’ are random walks on the graph word2vec
  • 108. Playlists at Spotify context sequence learning ‘words’ are song indices ‘sentences’ are playlists
  • 110. Fixes at Stitch Fix sequence learning Let’s try: ‘words’ are items ‘sentences’ are ïŹxes
  • 111. Fixes at Stitch Fix context Learn similarity between styles because they co-occur Learn ‘coherent’ styles sequence learning
  • 114. Fixes at Stitch Fix? context sequence learning Nearby regions are consistent ‘closets’ What sorts of sequences do you have at Quora? What kinds of things can you learn from context?
  • 117. context dependent context Australian scientist discovers star with telescope Levy & G oldberg 2014 What if we
  • 118. context dependent context Australian scientist discovers star with telescope context Levy & G oldberg 2014
  • 119. context dependent context BoW DEPS topically-similar vs ‘functionally’ similar Levy & G oldberg 2014
  • 122. Crazy Approaches Paragraph Vectors (Just extend the context window) Content dependency (Change the window grammatically) Social word2vec (deepwalk) (Sentence is a walk on the graph) Spotify (Sentence is a playlist of song_ids) Stitch Fix (Sentence is a shipment of ïŹve items)
  • 124. CBOW “The fox jumped over the lazy dog” Guess the word given the context ~20x faster. (this is the alternative.) vOUT vIN vINvIN vIN vIN vIN SkipGram “The fox jumped over the lazy dog” vOUT vOUT vIN vOUT vOUT vOUTvOUT Guess the context given the word Better at syntax. (this is the one we went over) CBOW sums words vectors, loses the order in the sentence Both are good at semantic relationships Child and kid are nearby Or gender in man, woman If you blur words over the scale of context — 5ish words, you lose a lot grammatical nuance But skipgram preserves order Preserves the relationship in pluralizing, for example
  • 125. lda2vec vDOC = a vtopic1 + b vtopic2 +
 Let’s make vDOC sparse Too many documents. I really like that document X is 70% in topic 0, 30% in topic1, 

  • 126. lda2vec This works! 😀 But vDOC isn’t as interpretable as the topic vectors. 😔 vDOC = topic0 + topic1 Let’s say that vDOC ads Too many documents. I really like that document X is 70% in topic 0, 30% in topic1, 

  • 127. lda2vec softmax(vOUT * (vIN+ vDOC)) we want k *sparse* topics
  • 128. Shows that are many words similar to vacation actually come in lots of ïŹ‚avors — wedding words (bachelorette, rehearsals) — holiday/event words (birthdays, brunch, christmas, thanksgiving) — seasonal words (spring, summer,) — trip words (getaway) — destinations
  • 131. LDA Results context H istory I loved every choice in this ïŹx!! Great job! Great Stylist Perfect There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
  • 132. LDA Results context H istory Body Fit My measurements are 36-28-32. If that helps. I like wearing some clothing that is ïŹtted. Very hard for me to ïŹnd pants that ïŹt right. There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
  • 133. LDA Results context H istory Sizing Really enjoyed the experience and the pieces, sizing for tops was too big. Looking forward to my next box! Excited for next There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
  • 134. LDA Results context H istory Almost Bought It was a great ïŹx. Loved the two items I kept and the three I sent back were close! Perfect There are k tags Issues "Cancel Disappointed" Delivery "ProïŹle, Pinterest" "Weather Vacation" “Corrections for Next" "Wardrobe Mix" "Requesting SpeciïŹc" "Requesting Department" "Requesting Style" "Style, Positive" "Style, Neutral"
  • 135. All of the following ideas will change what ‘words’ and ‘context’ represent. But we’ll still use the same w2v algo
  • 136. paragraph vector What about summarizing documents? On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that
  • 137. On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that The framework nuclear agreement he reached with Iran on Thursday did not provide the deïŹnitive answer to whether Mr. Obama’s audacious gamble will pay off. The ïŹst Iran has shaken at the so-called Great Satan since 1979 has not completely relaxed. paragraph vector Normal skipgram extends C words before, and C words after. IN OUT OUT Except we stay inside a sentence
  • 138. On the day he took ofïŹce, President Obama reached out to America’s enemies, offering in his ïŹrst inaugural address to extend a hand if you are willing to unclench your ïŹst. More than six years later, he has arrived at a moment of truth in testing that The framework nuclear agreement he reached with Iran on Thursday did not provide the deïŹnitive answer to whether Mr. Obama’s audacious gamble will pay off. The ïŹst Iran has shaken at the so-called Great Satan since 1979 has not completely relaxed. paragraph vector A document vector simply extends the context to the whole document. IN OUT OUT OUT OUTdoc_1347