Twitter is used for various applications such as disaster monitoring and news material gathering. In these applications, each Tweet is classified into pre-defined classes. These classes have a semantic relationship with each other and can be classified into a hierarchical structure, which is regarded as important information. Label texts of pre-defined classes themselves also include important clues for classification. Therefore, we propose a method that can consider the hierarchical structure of labels and label texts themselves. We conducted evaluation over the Text REtrieval Conference (TREC) 2018 Incident Streams (IS) track dataset, and we found that our method outperformed the methods of the conference participants.

Citation (Short format):
Taro Miyazaki, Kiminobu Makino, Yuka Takei, Hiroki Okamoto and Jun Goto (2019). Label Embedding using Hierarchical Structure of Labels for Twitter Classification In: Proceedings of EMNLP-IJCNLP 2019.

Paper PDF: Hosted by the ACL

Conference: 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

This is the primary overview paper for the TREC-IS 2018 edition. It is an extended version of the original notebook that was published in ISCRAM 2019. It covers the main motivation and aims for the track, as well as overall track data analysis and participant performance summaries.

Please cite this paper when refering to the track

Citation (Short format):
Richard McCreadie, Cody Buntain and Ian Soboroff (2019)
TREC Incident Streams: Finding Actionable Information on Social Media
In: Proceedings of ISCRAM 2019.

Paper PDF: Hosted by the University of Glasgow

Conference: 16th International Conference on Information Systems for Crisis Response and Management