HLT-FBK: a Complete Temporal Processing System for QA TempEval

Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), pages 801–805, Denver, Colorado, June 4-5, 2015. c⃝2015 Association for Computational Linguistics HLT-FBK: a Complete Temporal Processing System for QA TempEval Paramita Mirza FBK, Trento, Italy University of Trento paramita@fbk.eu Anne-Lyse Minard FBK, Trento, Italy minard@fbk.eu Abstract The HLT-FBK system is a suite of SVMs- based classification models for extracting time expressions, events and temporal relations, each with a set of features obtained with the NewsReader NLP pipeline. HLT-FBK’s best system runs ranked 1st in all three domains, with a recall of 0.30 over all domains. Our at- tempts on increasing recall by considering all SRL predicates as events as well as utilizing event co-reference information in extracting temporal links result in significant improve- ments. 1 Introduction QA TempEval is a continuation of the TempEval task series (Verhagen et al., 2007; Verhagen et al., 2010; UzZaman et al., 2013), which shifts its eval- uation methodology from temporal information ex- traction accuracy to temporal question-answering (QA) accuracy. However, the main task is the same as its predecessor tasks, which is to automatically annotate texts with temporal information following TimeML specification (Pustejovsky et al., 2003a). This paper describes the HLT-FBK system sub- mitted to QA TempEval. The system decomposes the task into three sub-tasks, i.e. temporal expres- sion (timex) extraction, event extraction and tempo- ral relation extraction. Each sub-task is formulated as a supervised classification problem using SVMs- based classifiers, which make use of the information acquired from the NewsReader1 NLP pipeline. 1http://www.newsreader-project.eu 2 Data, Resources and Tools The training data set is the TimeML annotated data released by the task organizers, which includes TBAQ-cleaned and TE3-Platinum corpora reused from the TempEval-3 task (UzZaman et al., 2013). We extended the training corpus for the timex ex- traction system with the TempEval-3 silver corpus. The test data are 30 plain texts of News, Wikipedia and Blogs domains (10 documents each). For evalu- ating the system, 294 temporal-based questions and the test data annotated with entities relevant for the questions are used. The resources used by the system to extract some features are lists of temporal signals extracted from the TimeBank corpus (Pustejovsky et al., 2003b) and a list of nominalizations extracted from the SPE- CIALIST Lexicon2 distributed by the U.S. National Library of Medicine, which contains commonly oc- curring English words in addition to biomedical terms, with syntactic and morphological informa- tion. We extracted all nouns resulting from a nom- inalization. Other features come from the annota- tion of the addDiscourse tool (Pitler and Nenkova, 2009), which identifies discourse connectives and assigns them to one of the four semantic classes: Temporal, Expansion, Contingency and Compari- son. The MorphoPro module, part of the TextPro tool suite3, is used to get the morphological analysis of each token in a text. The time expression nor- 2http://www.nlm.nih.gov/research/umls/ new_users/online_learning/LEX_001.html 3http://textpro.fbk.eu/ 801 malization sub-task is carried out by TimeNorm4 (Bethard, 2013), a library for converting natural lan- guage expressions of dates and times into their nor- malized form. …

Content truncated to 500 words