Acquiring Event Relation Knowledge by Learning Cooccurrence Patterns and Fertilizing Cooccurrence Samples

Abstract

Aiming at acquiring semantic relations between events from a large corpus, this paper proposes several extensions to a state-of-the-art method originally designed for entity relation extraction. First, expressions of events are defined to specify the class of the acquisition task. Second, the templates of co-occurrence patterns are extended so that they can capture semantic relations between event mentions. Experiments on a Japanese Web corpus show that (a) there are indeed specific co-occurrence patterns useful for event relation acquisition, and (b) For action-effect relation, at least five thousand relation instances are acquired from a 500M-sentence Web corpus with a precision of about 66%.

Capturing consistency between intra-clause and inter-clause relations in knowledge-rich dependency and case structure analysis

“…We automatically acquire large-scale transition knowledge of inter-clause relations from a raw corpus. The following two points are different from previous studies on the acquisition of interclause knowledge such as entailment/synonym knowledge (Lin and Pantel, 2001;Torisawa, 2006;Pekar, 2006;Zanzotto et al, 2006), verb relation knowledge (Chklovski and Pantel, 2004), causal knowledge (Inui et al, 2005) and event relation knowledge (Abe et al, 2008):…”

“…Inui et al (2005) classified the occurrences of the Japanese connective marker tame. Abe et al (2008) learned event relation knowledge for two semantic relations. They first gave seed pairs of verbs or verb phrases and extracted the patterns that matched these seed pairs.…”

Two-phased event relation acquisition

“…Inui et al (2003), for example, use a Japanese generic causal connective marker tame (because) and a supervised classifier learner to separately obtain four types of causal relations: cause, precondition, effect and means. More recently, Abe et al (2008) propose to extend Pantel and Pennac-chiotti (2006)'s Espresso algorithm, which induces specific reliable LSPs in a bootstrapping manner for entity-entity relation extraction, so that the extended algorithm can apply to event relations. Their method learns a large number of relatively specific patterns such as cannot find out (something) due to the lack of investigation in a boot-strapping fashion, which produces a remarkable improvement on precision.…”

“…Pattern-based methods, on the one hand, are designed to be capable of discriminating relatively fine-grained relation types. For example, the patterns used by Chklovski and Pantel (2005) identify six relation types, while Abe et al (2008) identify two of the four causal relation types defined by Inui et al (2003). However, these methods are severely limited for the purpose of shared argument identification because lexicosyntactic patterns are not a good indication of argument-shared structure in general.…”

“…For predicate pair acquisition, we can choose one from a range of state-of-the-art pattern-based methods. Among others, in our experiments, we adopted Abe et al (2008)'s method because it had an advantage in that it was capable of learning patterns as well as relation instances. Abe et al (2008)'s method is based on Pantel and Pennacchiotti (2006)'s Espresso algorithm, which is originally designed to acquire relations between entities.…”