130 likes | 256 Views
Unsupervised Discovery of Fine-Grained Topic Clusters in Twitter Posts Vita G. Markman. 1. Introduction. Social media == new source of information and the ground for social interaction Twitter: Noisy and content-sparse data Question: Can we carve out fine grained topics within the
E N D
Unsupervised Discovery of Fine-Grained Topic Clusters in Twitter Posts Vita G. Markman
1. Introduction Social media == new source of information and the ground for social interaction Twitter: Noisy and content-sparse data Question: Can we carve out fine grained topics within the micro-documents, e.g. topics such as “food”, “movies”, “music”, etc ?
2. Related Work Ritter et al. (2010) uncover dialogue acts such as “Status”, “Question to Followers”, “Comment”, “Reaction” within Twitter post via unsupervised modeling. Ramageet al. (2010) cluster a set of Twitter conversations using supervised LDA. Main finding: Twitter is 11% substance, 5% status, 16% style, 10% social, and 56% other.
3. Data 1119 posts were selected from Twitter corpus, with the following topics: Food 276 Movies 141 Sickness 39 Sports 34 Music 153 Relationships 66 Computers 57 Travel 32 Other 321 The topic labels were not included in the data.
4. Algorithm Latent Dirichlet Allocation (Blei et al, 2003) was used to model topics Each document == a mixture of topics Topic == a probability distribution over words Intuition behind LDA == words that co-occur across documents have similar meaning and indicate the topic of the document. Example: if “café” and “food” co-occur in multiple documents, they are related to the same topic.
5. Experiment 1 Individual posts broken into words and a random topic between 0 and 8 was assigned to each word. Only nouns and verbs were left in the data. A representation of a tweet such as (1) is in (2) (1) “hey, whats going on lets get some food in this café” (2) [(‘food’, t1), (‘cafe’, t2)]
6. Unpadded Data – Results t0music 67%; movies 33% t1 other 50%; travel 12.5%; sports 12.5%; food 25% t2music 50%; relationships 25%; sickness 12.5%; computers 12.5% t3food 80%; movies 20% t4food 74%; music 11%; sickness 10%; other 5% t5music 67%; other 14%; movies 9%; food 5%; sports 5% t6food 50%; other 14%; sickness 14%; computers 14%; movies 8% t7food 70%; music 20%; relationships 10% t8food 45%; travel 28%; relationships 9%; other 9%; computers 9%
7. Problems & Solutions Results: Tweets on the same topic scattered across clusters Clusters contained various unrelated tweets. Problem: tweets on the same topic had no words in common Solution: ‘Thesaurus’ of words related to the topics in posts was created Most generic words from the thesaurus were selected as “pads” Sample thesaurus entry: (3) food_topic [“pizza”, “café” , “food”, eat..] food topic padding set: [“food”, “café”] Posts shorter than 3 words were padded with 2-3 words from the relevant “padding” set Padding ensured that some shared words appear in posts Unpadded posts: (4) “café eat” (5) “food” Padded posts: (4’) “café food eat food” (5’) “café food food”.
8. Padded Data – Results t0 music 77%; food 9%; computers 4.54%; sports 4.54%; travel 4.54% t1movies 80%; relationships 3%; sickness 10%; music 5%; food 2% t2food 69%; movies 16%; music 5%; sickness 5%; sports 5% t3 food 93%; sickness 4%; other 3%; movies 1% t4 other 65%; travel 29%; movies 6% t5 sickness 100% t6music 46%; other 24%; sports 20%; computers 10% t7food 49%; computers 22%; other 20%; music 8%; movies 1% t8 relationships 55%; food 23%; music 10%; sickness 6%; movies 6%
9. Experiment 2 - Discussion 5 out of 9 clusters contained at least one strongly dominant semantic topic (over 60%) 6 topics emerge from the padded data clustering: food, movies, music, other, sickness, and relationships. Food is distributed across 3 clusters and is the dominant topic in all of them The most cohesive clusters are t5, t3, t2, t1, and t0; they have a dominant semantic topic – one that occurs in over 60% of the posts within the cluster
10. General Discussion Differences across the two data sets: • Non-padded data has no pure clusters and no clusters where one topic occurs < 80% of the time • 3 clusters found in non-padded data: music, • food, and other. • 6 distinct clusters found inthe padded data: music, movies, food, relationships, other, and sickness Similarities in the padded and the non-padded data sets: • movie posts were given the same topic ID as food posts • Food, music, and movies were distributed across several topic assignments
11. Conclusion Information sparsity has a strong negative effect on topic clustering. Topic clustering is improved, when: a) only nouns and verbs are used in posts b) posts are padded with similar words Future investigation: Can padding help discover topics in large sets of tweets? Thank you!
12. References • Barzilay, R. and L. Lee. 2004. Catching the Drift: Probabilistic Content Models, with Applications to Generation and Summarization. In Proceedings of HLT-NAACL, 113-120. Stroudsburg, PA. • Blei, D., Ng, A. and Jordan, M. 2003. Latent Dirichlet Allocation. Journal of Machine Learning: 3, 993-1022. • Griffiths, T. and Steyvers, M. 2003. Prediction and Semantic Association. In Neural Information Processing Systems 15: 11-18. • Griffiths, T. and Steyvers, M. 2004. Finding Scientific Topics. Proceedings of Natural Academic of Sciences, 101 Supplement 1:5228-5235. • Griffiths, T. and Steyvers, M. 2007. Probabilistic Topic Models. In Landauer, T., McNamara, D., D. Dennis, S. and Kintsch, W. (eds.) Latent Semantic Analysis: A Road to Meaning. Laurence Erlbaum. • Hoffman, T. 1999. Probabilistic Latent Semantic Analysis. In Proceedings of the Fifteenth Conference on Uncertainty in Artificial Intelligence, 289-296. Stockholm, Sweden: Morgan Kaufman. • Hoffman, T. 2001. Unsupervised Learning by Probabilistic Latent Semantic Analysis. Machine Learning Journal, 42(1), 177-196. • Ramage, D., Dumais, S., and Liebling, D. 2010. Characterizing Microblogs with Topic Models. Association for the Advancement of Artificial Intelligence. • Ritter, A., Cherry, C., and Dolan, B. 2010. Unsupervised Modeling of Twitter Conversations. In Proceedings of HLT-NAACL 2010. 113-120. Stroudsburg, PA. • Ritter, A. et. al. 2011. Status Messages: A Unique Textual Source of Realtime and Social Information. Presented at Natural Language Seminar, Information Sciences Institute, USC. • Yao, L., Mimno,D., and McCallum, A. 2009. Efficient Methods for Topic Model Inference on Streaming Document Collections. In Proceedings of the 15th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 937-946.