BDACA1516s2 - Lecture4
1 like368 views
The document discusses various analytical approaches to sentiment analysis and the removal of stopwords in the context of big data and automated content analysis, focusing on applications like querying APIs and handling data formats like CSV and JSON. It highlights sentiment analysis as a method for extracting subjective information from texts, explaining advanced techniques and their applications across different fields such as social science and web analytics. Additionally, the document explores both bag-of-words and more advanced methods for analyzing sentiment, addressing their pros and cons.
![Different types of analysis Sentiment analysis Stopword removal Next meetings
Bag-of-words approaches
How to do this
(given a string tekst that you want to analyze and two lists of strings with negative
and positive words, lijstpos=["great","fantastic",...,"perfect"] and
lijstneg)
1 sentiment=0
2 for woord in tekst.split():
3 if woord in lijstpos:
4 sentiment=sentiment+1 #same as sentiment+=1
5 elif woord in lijstneg:
6 sentiment=sentiment-1 #same as sentiment-=1
7 print (sentiment)
Big Data and Automated Content Analysis Damian Trilling](https://image.slidesharecdn.com/lecture4-160413151351/85/BDACA1516s2-Lecture4-30-320.jpg)
![Different types of analysis Sentiment analysis Stopword removal Next meetings
A simple algorithm
Stopword removal: How
1 testo=’He gives her a beer and a cigarette.’
2 testonuovo=""
3 stopwords=[’and’,’the’,’a’,’or’,’he’,’she’,’him’,’her’]
4 for verbo in testo.split():
5 if verbo not in stopwords:
6 testonuovo=testonuovo+verbo+" "
What do we get if we do:
1 print (testonuovo)
Can you explain the algorithm?
Big Data and Automated Content Analysis Damian Trilling](https://image.slidesharecdn.com/lecture4-160413151351/85/BDACA1516s2-Lecture4-57-320.jpg)
![Different types of analysis Sentiment analysis Stopword removal Next meetings
A simple algorithm
Stopword removal
1 testo=’He gives her a beer and a cigarette.’
2 testonuovo=""
3 stopwords=[’and’,’the’,’a’,’or’,’he’,’she’,’him’,’her’]
4 for verbo in testo.split():
5 if verbo.lower() not in stopwords:
6 testonuovo=testonuovo+verbo+" "
Big Data and Automated Content Analysis Damian Trilling](https://image.slidesharecdn.com/lecture4-160413151351/85/BDACA1516s2-Lecture4-59-320.jpg)
BDACA1516s2 - Lecture4
- 1. Different types of analysis Sentiment analysis Stopword removal Next meetings Big Data and Automated Content Analysis Week 4 – Monday »Sentiment Analysis« Damian Trilling d.c.trilling@uva.nl @damian0604 www.damiantrilling.net Afdeling Communicatiewetenschap Universiteit van Amsterdam 18 April 2016 Big Data and Automated Content Analysis Damian Trilling
- 2. Different types of analysis Sentiment analysis Stopword removal Next meetings Today 1 Different types of analysis What can we do? Systematizing analytical approaches 2 Data analysis 1: Sentiment analysis What is it? Bag-of-words approaches Advanced approaches A sentiment analysis tailored to your needs! 3 NLP-preview: Stopword removal Natural language processing A simple algorithm 4 Take-home message, next meetings, & exam Big Data and Automated Content Analysis Damian Trilling
- 3. Different types of analysis Sentiment analysis Stopword removal Next meetings What we already can do with regard to data collection: • query a (JSON-based) API (GoogleBooks, Twitter) • handle CSV files • handle JSON files Big Data and Automated Content Analysis Damian Trilling
- 4. Different types of analysis Sentiment analysis Stopword removal Next meetings What we already can do with regard to data collection: • query a (JSON-based) API (GoogleBooks, Twitter) • handle CSV files • handle JSON files with regard to analysis: Not much. We counted some frequencies and calculated some averages. Big Data and Automated Content Analysis Damian Trilling
- 5. Different types of analysis Sentiment analysis Stopword removal Next meetings What can we do? Data analysis 0: Overview What can we do? Big Data and Automated Content Analysis Damian Trilling
- 6. Different types of analysis Sentiment analysis Stopword removal Next meetings What can we do? Data analysis 0: Overview What can we do? What do you think? What are interesting methods to analyze large data sets (like, e.g., social media data? What questions can they answer?) Big Data and Automated Content Analysis Damian Trilling
- 7. Different types of analysis Sentiment analysis Stopword removal Next meetings What can we do? What else can we do? For example • sentiment analysis • automated coding with regular expressions • natural language processing • supervised and unsupervised machine learning • network analysis Big Data and Automated Content Analysis Damian Trilling
- 8. Different types of analysis Sentiment analysis Stopword removal Next meetings What can we do? What else can we do? Or ideally. . . . . . a combination of these techniques. Big Data and Automated Content Analysis Damian Trilling
- 9. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Overview Systematizing analytical approaches Big Data and Automated Content Analysis Damian Trilling
- 10. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Systematizing analytical approaches Taking the example of Twitter: Analyzing the structure • Number of Tweets over time • singleton/retweet ratio • Distribution of number of Tweets per user • Interaction networks Bruns, A., & Stieglitz, S. (2013). Toward more systematic Twitter analysis: metrics for tweeting activities. International Journal of Social Research Methodology. doi:10.1080/13645579.2012.756095 Big Data and Automated Content Analysis Damian Trilling
- 11. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Systematizing analytical approaches Taking the example of Twitter: Analyzing the structure • Number of Tweets over time • singleton/retweet ratio • Distribution of number of Tweets per user • Interaction networks ⇒ Focus on the amount of content and on the question who interacts with whom, not on what is said Bruns, A., & Stieglitz, S. (2013). Toward more systematic Twitter analysis: metrics for tweeting activities. International Journal of Social Research Methodology. doi:10.1080/13645579.2012.756095 Big Data and Automated Content Analysis Damian Trilling
- 12. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Systematizing analytical approaches Taking the example of Twitter: Analyzing the content • Sentiment analysis • Word frequencies, searchstrings • Co-word analysis (⇒frames) Big Data and Automated Content Analysis Damian Trilling
- 13. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Systematizing analytical approaches Taking the example of Twitter: Analyzing the content • Sentiment analysis • Word frequencies, searchstrings • Co-word analysis (⇒frames) ⇒ Focus on what is said Big Data and Automated Content Analysis Damian Trilling
- 14. Different types of analysis Sentiment analysis Stopword removal Next meetings Systematizing analytical approaches Systematizing analytical approaches ⇒ It depends on your reserach question which approach is more interesting! Big Data and Automated Content Analysis Damian Trilling
- 15. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? Data analysis 1: Sentiment analysis Big Data and Automated Content Analysis Damian Trilling
- 16. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts Big Data and Automated Content Analysis Damian Trilling
- 17. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts • the author’s attitude towards the topic of the text Big Data and Automated Content Analysis Damian Trilling
- 18. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts • the author’s attitude towards the topic of the text • polarity: negative—positive Big Data and Automated Content Analysis Damian Trilling
- 19. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts • the author’s attitude towards the topic of the text • polarity: negative—positive • subjectivity: neutral—subjective * Big Data and Automated Content Analysis Damian Trilling
- 20. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts • the author’s attitude towards the topic of the text • polarity: negative—positive • subjectivity: neutral—subjective * • advanced approaches: different emotions Big Data and Automated Content Analysis Damian Trilling
- 21. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? What is sentiment analysis? Extracting subjective information from texts • the author’s attitude towards the topic of the text • polarity: negative—positive • subjectivity: neutral—subjective * • advanced approaches: different emotions * Less sophisticated approaches do not see this as a sperate dimension but simply calculate objectivity = 1 − (negativity + positivity) Big Data and Automated Content Analysis Damian Trilling
- 22. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? Applications Who uses it? • Companies • especially for Web Analytics • Social Scientists • applications in data journalism, politics, . . . Many references to examples in Mostafa (2013). ⇒ Cases in which you have a huge amount of data or real-time data and you want to get an idea of the tone. Mostafa, M. M. (2013). More than words: Social networks’ text mining for consumer brand sentiments. Expert Systems with Applications, 40(10), 4241– 4251. doi:10.1016/j.eswa.2013.01.019 Big Data and Automated Content Analysis Damian Trilling
- 23. Different types of analysis Sentiment analysis Stopword removal Next meetings What is it? Example 1 >>> sentiment("Great service by @NSHighspeed") 2 (0.8, 0.75) 3 >>> sentiment("Bad service by @NSHighspeed") 4 (-0.6166666666666667, 0.6666666666666666) (polarity, subjectivity) with −1 ≤ polarity ≤ +1 0 ≤ subjectivity ≤ +1 ) This is the module pattern.nl, available for Python 2 only. De Smedt, T., & Daelemans W. (2012). Pattern for Python. Journal of Machine Learning Research, 13, 2063-2067. Big Data and Automated Content Analysis Damian Trilling
- 24. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Data analysis 1: Sentiment analysis Bag-of-words approaches Big Data and Automated Content Analysis Damian Trilling
- 25. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches How does it work? • We take each word of a text and look if it’s positive or negative. Big Data and Automated Content Analysis Damian Trilling
- 26. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches How does it work? • We take each word of a text and look if it’s positive or negative. • Most simple way: compare it with a list of negative words and with a list of positive words (That’s what Mostafa (2013) did) Big Data and Automated Content Analysis Damian Trilling
- 27. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches How does it work? • We take each word of a text and look if it’s positive or negative. • Most simple way: compare it with a list of negative words and with a list of positive words (That’s what Mostafa (2013) did) • More advanced: look up a subjectivity score from a table Big Data and Automated Content Analysis Damian Trilling
- 28. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches How does it work? • We take each word of a text and look if it’s positive or negative. • Most simple way: compare it with a list of negative words and with a list of positive words (That’s what Mostafa (2013) did) • More advanced: look up a subjectivity score from a table • e.g., add up the scores and average them. Big Data and Automated Content Analysis Damian Trilling
- 29. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches How to do this If you were to run an analyis like the one by Mostafa (2013), how could you do this? Big Data and Automated Content Analysis Damian Trilling
- 30. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches How to do this (given a string tekst that you want to analyze and two lists of strings with negative and positive words, lijstpos=["great","fantastic",...,"perfect"] and lijstneg) 1 sentiment=0 2 for woord in tekst.split(): 3 if woord in lijstpos: 4 sentiment=sentiment+1 #same as sentiment+=1 5 elif woord in lijstneg: 6 sentiment=sentiment-1 #same as sentiment-=1 7 print (sentiment) Big Data and Automated Content Analysis Damian Trilling
- 31. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Do we need to have the lists in our program itself? No. You could have them in a separate text file, one per row, and then read that file directly to a list. 1 poslijst=open("filewithonepositivewordperline.txt").read().splitlines() 2 neglijst=open("filewithonenegativewordperline.txt").read().splitlines() Big Data and Automated Content Analysis Damian Trilling
- 33. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches More advanced versions • CSV files or similar tables with weights • Or some kind of dict? Big Data and Automated Content Analysis Damian Trilling
- 36. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Mustafa 2013: Interpreting the output Big Data and Automated Content Analysis Damian Trilling
- 37. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Mustafa 2013: Interpreting the output Your thoughts? Big Data and Automated Content Analysis Damian Trilling
- 38. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Mustafa 2013: Interpreting the output Your thoughts? • each word counts equally (1) • many tweets contain no words from the list. What does this mean? • Ways to improve BOW approaches? Big Data and Automated Content Analysis Damian Trilling
- 39. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches e.g., Schut, L. (2013). Verenigde Staten vs. Verenigd Koningrijk: Een automatische inhoudsanalyse naar verklarende factoren voor het gebruik van positive campaigning en negative campaigning door vooraanstaande politici en politieke partijen op Twitter. Bachelor Thesis, Universiteit van Amsterdam. Big Data and Automated Content Analysis Damian Trilling
- 40. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches pro • easy to implement • easy to modify: • add or remove words • make new lists for other languages, other categories (than positive/negative), . . . • easy to understand (transparency, reproducability) e.g., Schut, L. (2013). Verenigde Staten vs. Verenigd Koningrijk: Een automatische inhoudsanalyse naar verklarende factoren voor het gebruik van positive campaigning en negative campaigning door vooraanstaande politici en politieke partijen op Twitter. Bachelor Thesis, Universiteit van Amsterdam. Big Data and Automated Content Analysis Damian Trilling
- 41. Different types of analysis Sentiment analysis Stopword removal Next meetings Bag-of-words approaches Bag-of-words approaches con • simplistic assumptions • e.g., intensifiers cannot be interpreted ("really" in "really good" or "really bad") • or, even more important, negations. Big Data and Automated Content Analysis Damian Trilling
- 42. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Data analysis 1: Sentiment analysis Advanced approaches Big Data and Automated Content Analysis Damian Trilling
- 43. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Improving the BOW approach Example: The Sentistrenght algorithm • −5 . . . − 1 and +1 . . . + 5 • spelling correction • "booster word list" for strengthening/weakening the effect of the following word • interpreting repeated letters ("baaaaaad"), CAPITALS and !!! • idioms • negation • ldots Thelwall, M., Buckley, K., & Paltoglou, G. (2012). Sentiment strength detection for the social Web. Journal of the American Society for Information Science and Technology, 63(1), 163-173. Big Data and Automated Content Analysis Damian Trilling
- 44. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Advanced approaches Take the structure of a text into account • Try to apply linguistics concepts to identify sentence structure • can identify negations • can interpret intensifiers Big Data and Automated Content Analysis Damian Trilling
- 45. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Example 1 from pattern.nl import sentiment 2 >>> sentiment("Great service by @NSHighspeed") 3 (0.8, 0.75) 4 >>> sentiment("Really") 5 (0.0, 1.0) 6 >>> sentiment("Really Great service by @NSHighspeed") 7 (1.0, 1.0) (polarity, subjectivity) with −1 ≤ polarity ≤ +1 0 ≤ subjectivity ≤ +1 ) Unlike in pure bag-of-words approaches, here, the overall sentiment is not just the sum or the average of its parts! De Smedt, T., & Daelemans W. (2012). Pattern for Python. Journal of Machine Learning Research, 13, 2063-2067. Big Data and Automated Content Analysis Damian Trilling
- 46. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Advanced approaches Big Data and Automated Content Analysis Damian Trilling
- 47. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Advanced approaches pro • understand intensifiers or negation • thus: higher accuracy Big Data and Automated Content Analysis Damian Trilling
- 48. Different types of analysis Sentiment analysis Stopword removal Next meetings Advanced approaches Advanced approaches pro • understand intensifiers or negation • thus: higher accuracy con • Black box? Or do we understand the algorithm? • Difficult to adapt to own needs • really much better results? Big Data and Automated Content Analysis Damian Trilling
- 49. Different types of analysis Sentiment analysis Stopword removal Next meetings A sentiment analysis tailored to your needs! Data analysis 1: Sentiment analysis A sentiment analysis tailored to your needs! Big Data and Automated Content Analysis Damian Trilling
- 50. Different types of analysis Sentiment analysis Stopword removal Next meetings A sentiment analysis tailored to your needs! A sentiment analysis tailored to your needs! Identifying suicidal texts • Bag-of-words-approach with very specific dictionary • added negation • added regular expression search for key phrases • Very specific design requirements: False positives are OK, false negatives not! Huang, Y.-P., Goh, T., & Liew, C.L. (2007). Hunting suicide notes in web 2.0 – preliminary findings. Ninth IEEE International Symposium on Multimedia. Retrieved from http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4476021 Big Data and Automated Content Analysis Damian Trilling
- 51. Different types of analysis Sentiment analysis Stopword removal Next meetings A sentiment analysis tailored to your needs! Already this still relatively simple approach seems to work satisfactory, but if 106 scientists from 24 competing teams (!) work on it, they can Pestian, J.P.; Matykiewicz, P., Linn-Gust, M., South, B., Uzuner, O., Wiebe, J., Cohen, K.B., Hurdle, J., & Brew, C. (2012). Sentiment analysis of suicide notes: A shared task. Biomedical Informatics Insights, 5(1), p. 3-16. Retrieved from http://europepmc.org/articles/PMC3299408?pdf=render Big Data and Automated Content Analysis Damian Trilling
- 52. Different types of analysis Sentiment analysis Stopword removal Next meetings A sentiment analysis tailored to your needs! Already this still relatively simple approach seems to work satisfactory, but if 106 scientists from 24 competing teams (!) work on it, they can group suidide notes by these characteristics: • swear • family • friend • positive emotion • negative emotion • anxiety • anger • sad • cognitive process • biology • sexual • ingestion • religion • death Pestian, J.P.; Matykiewicz, P., Linn-Gust, M., South, B., Uzuner, O., Wiebe, J., Cohen, K.B., Hurdle, J., & Brew, C. (2012). Sentiment analysis of suicide notes: A shared task. Biomedical Informatics Insights, 5(1), p. 3-16. Retrieved from http://europepmc.org/articles/PMC3299408?pdf=render Big Data and Automated Content Analysis Damian Trilling
- 53. Different types of analysis Sentiment analysis Stopword removal Next meetings A sentiment analysis tailored to your needs! Sidenote An alternative state-of-the-art approach: Use supervised machine learning • Instead of defining rules, hand-code (“annotate”) the sentiment of some tweets manually and let the computer find out which words or characters (“features”) predict sentiment • Then use this model to predict sentiment for other tweets • Essentially the same like what you know since the second year of your Bachelor: regression analysis (but now with DV sentiment and IV’s word occurrences) • ⇒ week 8 Gonzalez-Bailon, S., & Paltoglou, G. (2015). Signals of public opinion in online communication: A comparison of methods and data sources. The ANNALS of the American Academy of Political and Social Science, 659(1), 95-–107. Big Data and Automated Content Analysis Damian Trilling
- 54. Different types of analysis Sentiment analysis Stopword removal Next meetings Natural Language Processing preview: Stopword removal Big Data and Automated Content Analysis Damian Trilling
- 55. Different types of analysis Sentiment analysis Stopword removal Next meetings Natural Language Processing preview: Stopword removal Why now? — Because the logic of the algorithm is very much related to the one of our first simple sentiment analysis Big Data and Automated Content Analysis Damian Trilling
- 56. Different types of analysis Sentiment analysis Stopword removal Next meetings Natural language processing Stopword removal: What and why? Why remove stopwords? • If we want to identify key terms (e.g., by means of a word count), we are not interested in them • If we want to calculate document similarity, it might be inflated • If we want to make a word co-occurance graph, irrelevant information will dominate the picture Big Data and Automated Content Analysis Damian Trilling
- 57. Different types of analysis Sentiment analysis Stopword removal Next meetings A simple algorithm Stopword removal: How 1 testo=’He gives her a beer and a cigarette.’ 2 testonuovo="" 3 stopwords=[’and’,’the’,’a’,’or’,’he’,’she’,’him’,’her’] 4 for verbo in testo.split(): 5 if verbo not in stopwords: 6 testonuovo=testonuovo+verbo+" " What do we get if we do: 1 print (testonuovo) Can you explain the algorithm? Big Data and Automated Content Analysis Damian Trilling
- 58. Different types of analysis Sentiment analysis Stopword removal Next meetings A simple algorithm We get: 1 >>> print (testonuovo) 2 ’He gives beer cigarette. ’ Why is "He" still in there? How can we fix this? Big Data and Automated Content Analysis Damian Trilling
- 59. Different types of analysis Sentiment analysis Stopword removal Next meetings A simple algorithm Stopword removal 1 testo=’He gives her a beer and a cigarette.’ 2 testonuovo="" 3 stopwords=[’and’,’the’,’a’,’or’,’he’,’she’,’him’,’her’] 4 for verbo in testo.split(): 5 if verbo.lower() not in stopwords: 6 testonuovo=testonuovo+verbo+" " Big Data and Automated Content Analysis Damian Trilling
- 60. Different types of analysis Sentiment analysis Stopword removal Next meetings Take-home message Mid-term take-home exam Next meetings Big Data and Automated Content Analysis Damian Trilling
- 61. Different types of analysis Sentiment analysis Stopword removal Next meetings Take-home messages What you should be familiar with: • You should have completely understood last week’s exercise. Re-read it if neccessary. • Approaches to the analysis (e.g., structure vs. content) • Types of sentiment analysis, application areas, pros and cons Big Data and Automated Content Analysis Damian Trilling
- 62. Different types of analysis Sentiment analysis Stopword removal Next meetings Mid-term take home exam Week 5: Monday, 25 April, to Sunday, 1 May • You get the exam on Monday at the end of the meeting • Answers have to be handed in no later than Sunday, 23.59 • 30% of final grade • 3 questions: 1 Literature question: Different methods (“Explain how. . . is done”) and/or epistemological or theoretical implications (“What does this mean for social-scientific research?”) 2 Empirical question (conceptual) 3 Empirical question (actual programming task) If you fully understood all exercises until now, it shouldn’t be difficult and won’t take too long. But give yourself a lot of buffer time!!! Big Data and Automated Content Analysis Damian Trilling
- 63. Different types of analysis Sentiment analysis Stopword removal Next meetings Next meetings Wednesday, 20–4 We work together on the examples in Chapter 5. You can bring your own data (you will probably learn more!), but then already think about how to write some script to read the data (as we did last week and as described in Chapter 4). Big Data and Automated Content Analysis Damian Trilling