Skip to main content
SpringerLink
Log in

Efficient temporal mining of micro-blog texts and its application to event discovery

  • Published:
Data Mining and Knowledge Discovery Aims and scope Submit manuscript

Abstract

In this paper we present a novel method for clustering words in micro-blogs, based on the similarity of the related temporal series. Our technique, named SAX*, uses the Symbolic Aggregate ApproXimation algorithm to discretize the temporal series of terms into a small set of levels, leading to a string for each. We then define a subset of “interesting” strings, i.e. those representing patterns of collective attention. Sliding temporal windows are used to detect co-occurring clusters of tokens with the same or similar string. To assess the performance of the method we first tune the model parameters on a 2-month 1 % Twitter stream, during which a number of world-wide events of differing type and duration (sports, politics, disasters, health, and celebrities) occurred. Then, we evaluate the quality of all discovered events in a 1-year stream, “googling” with the most frequent cluster n-grams and manually assessing how many clusters correspond to published news in the same temporal slot. Finally, we perform a complexity evaluation and we compare SAX* with three alternative methods for event discovery. Our evaluation shows that SAX* is at least one order of magnitude less complex than other temporal and non-temporal approaches to micro-blog clustering.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8

Similar content being viewed by others

Notes

  1. http://trec.nist.gov/data/tweets/ sampled from Jan 23rd to Feb 8th, 2011.

  2. http://wordnet.princeton.edu.

  3. In any case a limit is fixed a priori for the number of topics.

  4. https://dev.twitter.com/docs/streaming-apis.

  5. Words are stemmed to reduce sparseness, even though, as discussed in the paper, this might not be strictly necessary with more dense Twitter streams. In what follows we will refer to clustered items interchangeably as words, stems, or tokens.

  6. http://code.google.com/p/jmotif/wiki/ZNormalization.

  7. http://en.wikipedia.org/wiki/2012.

  8. https://www.google.it/trends/.

  9. http://libalf.informatik.rwth-aachen.de/index.php?page=home.

  10. See Fig. 8 of the mentioned paper, in which 6 shapes of attention of Twitter hashtags are shown.

  11. For example, in many algorithms the number of cluster K is a parameter.

  12. We use the euclidean distance, but other measures, e.g. the edit distance, produce very similar results.

  13. For example, there are many available implementations of LDA.

  14. Some of the events shown in the related papers are world-wide, but several are local events, e.g. “Super Junior’s Yesung (@shfly3424) created his Twitter account”.

  15. https://github.com/lintool/twitter-tools/wiki/TREC-2013-Track-Guidelines.

  16. In what follows we omit the “big-o” notation for simplicity: complexity formulas are all to be interpreted as “order of”.

  17. [\(1{\ldots }B\)] in the original paper (Xie et al. 2013).

  18. Table I of (Xie et al. 2013).

  19. http://en.wikipedia.org/wiki/Computational_complexity_of_mathematical_operations.

  20. In agreement with http://firstmonday.org/ojs/index.php/fm/article/view/4366/3654.

  21. This is also confirmed by the fact that we noticed an increment of daily tweets from an average of 3.3M per day during May to 4.6M during August.

  22. www.foreignpolicy.com/articles/2014/09/26/why_big_data_missed_the_early_warning_signs_of_ebola.

  23. Here we show only one tweet for the sake of space, whereas in our testing dataset we retrieve 10–20 tweets.

  24. These events can easily be filtered out by a classifier, however teen events could be of interest.

  25. http://jgibblda.sourceforge.net/.

  26. http://code.google.com/p/btm/.

  27. Requests must be addressed to the authors.

References

  • Blei DM, Ng AY, Jordan MI (2003) Latent Dirichlet allocation. J. Mach. Learn. Res. 3:993–1022

    MATH  Google Scholar 

  • Chae J, Thom D, Bosch H, Jang Y, Maciejewski R, Ebert D, Ertl T (2013) Spatiotemporal social media analytics for abnormal event detection and examination using seasonal-trend decomposition. IEEE symposium on visual analytics science and technology, Seattle

  • Cha M, Haddadi H, Benvenuto F, Gummadi K (2010) Measuring user influence in twitter: the million followers fallacy. In: Proceedings of conference on artificial intelligence AAAI

  • Cheng T, Wicks T (2014) Event detection using Twitter: a spatio-temporal approach. PLoS One 9(6):e97807. doi:10.1371/journal.pone.0097807

    Article  Google Scholar 

  • Dao Q, Jiang J, Zhu F, Lim WP (2012) Finding bursty topics from microblogs. In: Proceedings of conference association of computational linguistics ACL 2012

  • Dou W, Wang X, Ribarsky W, Zhou M (2012) Event detection in social media data. In: IEEE VisWeek workshop on interactive visual text analytics. Seattle, WA

  • Dredze M (2012) How social media will change public health. IEEE Intell Syst 27(4):81–84. doi:10.1109/MIS.2012.76

    Article  Google Scholar 

  • Hong L, Davison B (2010) Empirical study of topic modeling in twitter. In: Proceedings of the first workshop on social media analytics, pp. 80–88. ACM

  • Hong L, Dom B, Gurumurthy S, Tsioutsioulikis K (2011) Time-dependent topic model for multiple text streams. In: ACM conference on knowledge discovery and data mining KDD 2011, San Diego

  • Huang B, Yang Y, Mahmood A, Wang H (2012) Microblog topic detection based on LDA model and single-pass clustering RSCTC 2012, LNAI 7413, pp. 166–171

  • Ifrim G, Shi B, Brigadir I (2014) Event detection in Twitter using aggressive filtering and hierarchical tweet clustering proceedings of SNOW-WWW workshop, Korea

  • Jain A (2010) Data clustering: 50 years beyond K-means. Patt Recogn Lett 31:651–666

    Article  Google Scholar 

  • Keogh E, Chakrabarti K, Pazzani M (2001) Locally adaptive dimensionality reduction for indexing large time series databases. In: Proceedings Of ACM special interest group on management of data SIGMOD, pp. 151–162

  • Kovacs F, Legany C, Babos A (2005) Cluster validity measurement techniques. In: Proceedings of 6th international symposium of Hungarian researchers on computational intelligence, Budapest

  • Lee R, Sumiya K (2010) Measuring geographical regularities of crowd behaviors for twitter-based geo-social event detection. Proceedings of the 2nd ACM international workshop on location based social networks SIGSPATIAL, LBSN ’10. ACM, New York, pp. 1–10

  • Lehmann J, Goncalves B, Ramasco JJ, Cattuto C (2012) Dynamical classes of collective attention in Twitter. Proceedings of World Wide Web Conference WWW2012

  • Lin J, Keogh E, Li W, Lonardi S (2007) Experiencing SAX: a novel symbolic representation of time series. Data Mining Knowl Discov 15(2):107–144

    Article  Google Scholar 

  • Lin J, Khade R, Li Y (2012) Rotation-invariant similarity in time series using bag-of-patterns representation. J Intell Inf Syst 39:287–315

    Article  Google Scholar 

  • Li C, Sun A, Datta A (2012) Twevent: segment-based event detection from tweets. In: Proceedings of ACM international conference on information and knowledge management CIKM

  • Maynard D, Funk A (2012) Challenges in developing opinion mining tools for social media. In: Proceedings Of @NLP cann u tag #usergenartedcontent? Workshop at LREC 2012, Istanbul

  • McMinn A, Moshfeghi Y, Jose JM (2013) Building a large scale corpus for evaluating event detection in twitter, ACM international conference on information and knowledge management CIKM’13, San Francisco

  • Mei Q, Zhai C (2005) Discovering evolutionary theme patterns from text—an exploration of temporal text mining. In: Proceedings of conference of knowledge discovery and data mining KDD’05, Chigago

  • Oncina J, Garcıa P (1992) Inferring regular languages in polynomial updated time. In: 4th Spanish symposium on pattern recognition and image analysis, MPAI. vol. 1. World Scientific, pp. 49–61

  • Petrovic S, Osborne M, Lavrenko V (2010) Streaming first story detection with application to Twitter. In: Proceedings of national American conference of the association of computational linguistics NAACL

  • Petrovic S, Osborne M, Mc Creadie R (2013) Can Twitter replace Newswire for breaking news?. In: Proceedings of the 7th international AAAI conference on weblogs and social media, ICWSM

  • Pohl D, Bouchachia A, Hellwagner H (2012) Automatic sub-event detection in Emergency management using social media (2012), WWW2012-SWDM’12 Workshop, Lyon

  • Popescu AM, Pennacchiotti M, Paranjpe D (2011) Extracting events and event descriptions from twitter. In: Worls Wide Web Conference WWW2011, pp. 105–106, 2011

  • Rui L, Kin L, Ravi K, Kevin C (2012) TEDAS: a Twitter-based event detection and analysis system. In: IEEE 28th international conference on data engineering (ICDE), pp. 1273–1276

  • Wang X, Zhu F, Jing J, Li S (2013) Real time event detection in Twitter, conference on web age information management WAIM, Spinger

  • Weng J, Lim E, Jiang J, He Q (2010) Twitterrank: finding topic-sensitive influential twitterers. In: Proceedings of the third ACM international conference on Web Search and data mining WSDM, ACM, pp. 261–270

  • Weng J, Yao Y, Leonardi E, Lee B (2011) Event detection in Twitter. In: International AAAI conference on weblogs and social media ICWSM

  • Xie W, Zhu F, Jang J, Lim E, Wang K (2013) TopicSketch: real-time bursty topic detection from Twitter, IEEE 13th international conference on data mining (ICDM)

  • Yang J, Leskovec J (2011) Patterns of temporal variation in online media. In Proceedings of the fourth ACM international conference on web search and data mining (WSDM), pp. 177–186

  • Yan X, Guo J, Lan Y, Cheng X (2013) A biterm topic model for short texts. In: World Wide Web conference WWW 2013, Rio de Janeiro

Download references

Author information

Authors and Affiliations

  1. Department of Computer Science, Sapienza University of Roma, Via Salaria 113, Rome, Italy

    Giovanni Stilo & Paola Velardi

Authors
  1. Giovanni Stilo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Paola Velardi
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Paola Velardi.

Additional information

Responsible editor: Eamonn Keogh.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 196 KB)

Appendix

Appendix

See Table 6.

Table 6 “Summer 2014” experiments
Full size table

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Stilo, G., Velardi, P. Efficient temporal mining of micro-blog texts and its application to event discovery. Data Min Knowl Disc 30, 372–402 (2016). https://doi.org/10.1007/s10618-015-0412-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10618-015-0412-3

Keywords

Search

Navigation