Hierarchical Grouping and Multivision of Events through Consistency Graphs

• Paulo H. L. de Paula UFMS
• Westerley S. Reis UFMS
• Solange O. Rezende USP
• Ricardo M. Marcacini UFMS

DOI: https://doi.org/10.5753/kdmile.2018.27397

Abstract

Event analysis has received attention recently due to the popularization of web platforms for publishing content, especially news portals, social networks, blogs and forums. These platforms store events through texts about different sectors of society and can be seen as a digital representation (virtual world) of the events that occur in our real world. Thus, event grouping is an important task to organize and map the events of this virtual world to our physical world, which allows the realization of several social, political and economic studies. This work presents an approach to hierarchical grouping and multivision of events extracted from texts. The different information about the events, such as textual information, temporal information and geographical information are considered different views during the grouping task. While the existing approaches require the user to define parameters on how to use temporal and geographical information in the grouping of events, the proposed approach allows to automatically learn time and location restrictions. To this end, a structure called a consistency graph has been proposed that represents the consensus of groupings between different views. An experimental evaluation with eight sets of benchmark events revealed that the proposed approach is superior to the approach traditionally used in the area, also presenting the differential of allowing the visualization of the relationships between events through the consistency graph.

Keywords: event analysis, multivision grouping, consistency graph

References

Aggarwal, C. C. Machine learning for text. Springer, 2018.

Allan, J. Topic detection and tracking: event-based information organization. Vol. 12. Springer, 2012.

Conrad, J. G. and Bender, M. Semi-supervised events clustering in news retrieval. In Recent Trends in News Information Retrieval Workshop. pp. 21–26, 2016.

Deza, M. M. Distances and similarities in data analysis. In Encyclopedia of Distances. Springer, pp. 323–339, 2014.

Florence, R., Nogueira, B., and Marcacini, R. Constrained hierarchical clustering for news events. In Proceedings of the 21st International Database Engineering & Applications Symposium. ACM, pp. 49–56, 2017.

Hogenboom, F., Frasincar, F., Kaymak, U., de Jong, F., and Caron, E. A survey of event extraction methods from text for decision support systems. Decision Support Systems vol. 85, pp. 12–22, 2016.

Horie, S., Kiritoshi, K., and Ma, Q. Abstract-concrete relationship analysis of news events based on a 5W representation model. In Int. Conference on Database and Expert Systems Applications. Springer, pp. 102–117, 2016.

Hou, L. and Li. Newsminer: multifaceted news analysis for event search. KBS Journal vol. 76, pp. 17–29, 2015.

Radinsky, K., Davidovich, S., and Markovitch, S. Learning causality for news events prediction. In Proceedings of the 21st International Conference on World Wide Web. ACM, pp. 909–918, 2012.

Radinsky, K. and Horvitz, E. Mining the web to predict future events. In Proceedings of the 6th ACM International Conference on Web Search and Data Mining. ACM, pp. 255–264, 2013.

Yang, Y., Pierce, T., and Carbonell, J. A study of retrospective and on-line event detection. In Proceedings of the 21st Annual International ACM SIGIR Conference. ACM, pp. 28–36, 1998.

Zhao, J., Xie, X., Xu, X., and Sun, S. Multi-view learning overview: Recent progress and new challenges. Information Fusion vol. 38, pp. 43–54, 2017.