Joint Learning Templates and Slots for Event Schema Induction

Proceedings of NAACL-HLT 2016, pages 428–434, San Diego, California, June 12-17, 2016. c⃝2016 Association for Computational Linguistics Joint Learning Templates and Slots for Event Schema Induction Lei Sha, Sujian Li, Baobao Chang, Zhifang Sui Key Laboratory of Computational Linguistics, Ministry of Education School of Electronics Engineering and Computer Science, Peking University Collaborative Innovation Center for Language Ability, Xuzhou 221009 China shalei, lisujian, chbb, szf@pku.edu.cn Abstract Automatic event schema induction (AESI) means to extract meta-event from raw text, in other words, to find out what types (tem- plates) of event may exist in the raw text and what roles (slots) may exist in each event type. In this paper, we propose a joint entity-driven model to learn templates and slots simultane- ously based on the constraints of templates and slots in the same sentence. In addition, the entities’ semantic information is also con- sidered for the inner connectivity of the enti- ties. We borrow the normalized cut criteria in image segmentation to divide the entities into more accurate template clusters and slot clus- ters. The experiment shows that our model gains a relatively higher result than previous work. 1 Introduction Event schema is a high-level representation of a bunch of similar events. It is very useful for the tra- ditional information extraction (IE)(Sagayam et al., 2012) task. An example of event schema is shown in Table 1. Given the bombing schema, we only need to find proper words to fill the slots when extracting a bombing event. There are two main approaches for AESI task. Both of them use the idea of clustering the poten- tial event arguments to find the event schema. One of them is probabilistic graphical model (Chamber- s, 2013; Cheung, 2013). By incorporating tem- plates and slots as latent topics, probabilistic graphi- cal models learns those templates and slots that best explains the text. However, the graphical models Bombing Template Perpetrator: person Victim: person Target: public Instrument: bomb Table 1: The event schema of bombing event in MUC-4, it has a bombing template and four main slots considers the entities independently and do not take the interrelationship between entities into account. Another method relies on ad-hoc clustering algo- rithms (Filatova et al., 2006; Sekine, 2006; Cham- bers and Jurafsky, 2011). (Chambers and Jurafsky, 2011) is a pipelined approach. In the first step, it us- es pointwise mutual information(PMI) between any two clauses in the same document to learn events, and then learns syntactic patterns as fillers. How- ever, the pipelined approach suffers from the error propagation problem, which means the errors in the template clustering can lead to more errors in the s- lot clustering. This paper proposes an entity-driven model which jointly learns templates and slots for event schema induction. The main contribution of this paper are as follows: • To better model the inner connectivity between entities, we borrow the normalized cut in image segmentation as the clustering criteria. • We use constraints between templates and be- tween slots in one sentence to improve AESI …

Content truncated to 500 words