How NLP is revolutionizing marketing and communications

by#SK#Reddy
Chief#Product#Officer#AI#&#ML
linkedin.com/in/sk@reddy/
Digitalist#(www.digitalist.global)
How
NLP#
is#revolutionizing#
Marketing#and#Communications

McKinsey)Global)Institute:)AI)the)next)Digital)frontier)2017;)https://www.mckinsey.com/businessEfunctions/mckinseyEanalytics/ourE
insights/howEartificialEintelligenceEcanEdeliverErealEvalueEtoEcompanies

McKinsey)Global)Institute:)AI)the)next)Digital)frontier)2017;)https://www.mckinsey.com/businessEfunctions/mckinseyEanalytics/ourEinsights/howEartificialEintelligenceE
canEdeliverErealEvalueEtoEcompanies

https://venturebeat.com/2015/06/01/the7guy7who7made7this7insane720007company7marketing7landscape7chart7is7sorry/

https://www.accenture.com/t20160928T230416__w__/us;en/_acnmedia/PDF;32/AI_in_Management_Report.pdf
NLP$vs$Human$processing$

https://www.techemergence.com/machine2learning2marketing/
Where%do%I%focus?

https://www.techemergence.com/artificial4intelligence4in4marketing4and4advertising454examples4of4real4traction/
• Search
• Keyword
• Recommendation2Engines
• Collaborative2filtering,2Content<based2filtering
• Advertising
• News2paper,2TV,2Online
• Marketing2Forecasting
• Research
• Speech2/2Text2Recognition
• ??2
AI#in#Marketing
• Search
• Auto2spell2correction,2non<key2words,2image2
search,2audio2search
• Recommendation2Engines
• Deep2Learning2based2on2multiple<inputs
• Programmatic2Advertising
• automated2buying2and2selling2
• Marketing2Forecasting
• ML,2Streaming2data,2Satellite2Images,2tweets,2
emails,2Social2media,2IoT,2
• Speech2/2Text2Recognition2
• ChatBots,2AudioBots,2NLP,2Sentiment2
analyses,2
Pre.AI AI#enabled

• Email'management
• Email'Sorting
• Insight'Surfacing
• Contact'Data'Extraction
• Intent'Determination
• Sentiment'Analysis
Communications+&+Conversational+Marketing
https://www.prime?journal.com/lets?talk?how?conversational?marketing?can?make?you?a?better?practitioner/

“The%biggest%difference%between%
companies%who%are%poised%to%succeed%in%this%new%era%of%business%and%
those%that%aren’t%is%
the$ability$to$use$software$and$other$
technologies$to$create$new$market$
opportunities”%
– MIT%Tech%Review
https://www.technologyreview.com/s/608095/it9pays9to9be9smart/

https://www.techemergence.com/artificial4intelligence4in4marketing4and4advertising454examples4of4real4traction/
https://www.techemergence.com/machine4learning4marketing/
Challenges

The$unreasonable$effectiveness$of$RNNs
http://karpathy.github.io/2015/05/21/rnn5effectiveness/
RNN:=3=Layer=RNN=with=512=nodes=per=layer;=4.4=MB=input=file
PANDARUS:
Alas,=I=think=he=shall=be=come=approached=and=the=day=When=little=srain=would=be=attain'd=into=being=never=fed,=And=
who=is=but=a=chain=and=subjects=of=his=death,=I=should=not=sleep.
Second=Senator:
They=are=away=this=miseries,=produced=upon=my=soul,=Breaking=and=strongly=should=be=buried,=when=I=perish=The=
earth=and=thoughts=of=many=states.
DUKE=VINCENTIO:
Well,=your=wit=is=in=the=care=of=side=and=that.
Second=Lord:
They=would=be=ruled=after=this=chamber,=and=my=fair=nues=begun=out=of=the=fact,=to=be=conveyed,=Whose=noble=souls=
I'll=have=the=heart=of=the=wars.
Clown:
Come,=sir,=I=will=make=did=behold=your=worship.
VIOLA:
I'll=drink=it.

Ref:%https://stanfordnlp.github.io/CoreNLP/
Stanford)CoreNLP

Ref:%Google’s%Neural%Machine%Translation%System:%Bridging%the%Gap%between%Human%and%Machine%Translation@%https://arxiv.org/pdf/1609.08144.pdf
Google&Neural&Machine&Translation
• 8"layer"encoder"and"8"layer"decoder
• Bi.Directional"encoder"in"layer1"
• Residual"connections
• Attention"(from"bottom"to"top)
• Low"precision"arithmetic"during"inference
• Beam"search"(Length"normalization"&"coverage"
penalty)
• Rare"words

Ref:%A%Deep%Reinforced%Model%for%Abstractive%Summarization;%R%Paulus,%C%Xiong,%R%Socher;%https://arxiv.org/pdf/1705.04304.pdf
Neural'Model'for'summarization
To%avoid%repetition,%set%
for%a%triKgram

https://www.kaggle.com
You3can3solve3these3too…
1. Predict*which*shoppers*will*become*repeat*buyers
2. Predict*which*ads*contain*illicit*items*or*content
3. Predict*which*people*are*influential*in*a*social*network
4. Predict*if*a*customer*is*satisfied*or*dissatisfied
5. Can*you*predict*which*recommended*content*each*
user*will*click?
6. Predict*if*context*ads*will*earn*a*user's*click
7. Identify*individual*users*across*their*digital*devices
8. Which*customers*will*respond*to*a*direct*mail*offer
9. Predict*where*a*new*user*will*book*their*first*travel*
experience
10. Automatically*spot*duplicate*ads
11. Predicting*users’*demographic*characteristics*based*
on*their*app*usage,*geolocation,*and*mobile*device*
properties
12. Identify*which*customers*have*the*most*potential*
business*value*
13. How*much*interest*will*a*new*rental*listing*receive?
14. Customer*churn*prediction
15. Predict*if*a*driver*will*file*an*insurance*claim*next*year
16. Forecast*future*traffic*to*our*website
17. Can*you*assign*accurate*description*labels*to*images*
of*apparel*products?
18. Maximize*sales*and*minimize*returns*of*goods

Thank&you
SK#Reddy
www.linkedin.com/in/sk3reddy/

How NLP is revolutionizing marketing and communications

More Related Content

Similar to How NLP is revolutionizing marketing and communications (20)

More from SK Reddy (19)

Recently uploaded (20)

How NLP is revolutionizing marketing and communications