A study of retrospective and on-line event detection
Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval
Flexible intrinsic evaluation of hierarchical clustering for TDT
CIKM '03 Proceedings of the twelfth international conference on Information and knowledge management
Tracking dynamics of topic trends using a finite mixture model
Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Event threading within news topics
Proceedings of the thirteenth ACM international conference on Information and knowledge management
Discovering evolutionary theme patterns from text: an exploration of temporal text mining
Proceedings of the eleventh ACM SIGKDD international conference on Knowledge discovery in data mining
ICML '06 Proceedings of the 23rd international conference on Machine learning
Topics over time: a non-Markov continuous-time model of topical trends
Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining
Discovering event evolution graphs from news corpora
IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans
Discovering emerging topics in unlabelled text collections
ADBIS'06 Proceedings of the 10th East European conference on Advances in Databases and Information Systems
A content and social network approach of bibliometrics analysis across domains
Proceedings of the 2012 iConference
Expert Systems with Applications: An International Journal
Learning to explore spatio-temporal impacts for event evaluation on social media
ISNN'12 Proceedings of the 9th international conference on Advances in Neural Networks - Volume Part II
| Hi-index | 0.01 |
Due to the advance of Internet and Web 2.0 technologies, it is easy to extract thousands of threads about a topic of interest from an online forum but it is nontrivial to capture the blueprint of different aspects (i.e., subtopic, or facet) associated with the topic. To better understand and analyze a forum discussion given topic, it is important to uncover the evolution relationships (temporal dependencies) between different topic aspects (i.e. how the discussion topic is evolving). Traditional Topic Detection and Tracking (TDT) techniques usually organize topics as a flat structure but it does not present the evolution relationships between topic aspects. In addition, the properties of short and sparse messages make the content-based TDT techniques difficult to perform well in identifying evolution relationships. The contributions in this paper are two-folded. We formally define a topic aspect evolution graph modeling framework and propose to utilize social network information, content similarity and temporal proximity to model evolution relationships between topic aspects. The experimental results showed that, by incorporating social network information, our technique significantly outperformed content-based technique in the task of extracting evolution relationships between topic aspects.