CMU at SemEval-2016 Task 8: Graph-based AMR Parsing with Infinite Ramp Loss

Proceedings of SemEval-2016, pages 1202–1206, San Diego, California, June 16-17, 2016. c⃝2016 Association for Computational Linguistics CMU at SemEval-2016 Task 8: Graph-based AMR Parsing with Infinite Ramp Loss Jeffrey Flanigan♠ Chris Dyer♠ Noah A. Smith♥ Jaime Carbonell♠ ♠School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA ♥Computer Science & Engineering, University of Washington, Seattle, WA, USA {jflanigan,cdyer,jgc}@cs.cmu.edu, nasmith@cs.washington.edu Abstract We present improvements to the JAMR parser as part of the SemEval 2016 Shared Task 8 on AMR parsing. The major contributions are: improved concept coverage using external re- sources and features, an improved aligner, and a novel loss function for structured prediction called infinite ramp, which is a generalization of the structured SVM to problems with un- reachable training instances. 1 Introduction Our entry to the SemEval 2016 Shared Task 8 is a set of improvements to the system presented in Flanigan et al. (2014). The improvements are: a novel train- ing loss function for structured prediction, which we call “infinite ramp,” new sources for concepts, im- proved features, and improvements to the rule-based aligner in Flanigan et al. (2014). The overall archi- tecture of the system and the decoding algorithms for concept identification and relation identification are unchanged from Flanigan et al. (2014), and we refer readers seeking a complete understanding of the system to that paper. 2 New Concept Fragment Sources and Features The concept identification stage relies on a function called clex in Section 3 of Flanigan et al. (2014) to provide candidate concept fragments. In that work, clex has three sources of concept fragments: a lexicon extracted from the training data, rules for named entities identified by the named entity tagger, and rules for time expressions. We augment these sources with five additional sources: • Frame file lookup: for every word in the input sentence, if the lemma matches the name of a frame in the AMR frame files (with sense tag removed), we add the lemma concatenated with “-01” as a candidate concept fragment. • Lemma: for every word in the input sentence, we add the lemma of the word as a candidate concept fragment. • Verb pass-through: for every word in the in- put sentence, if the word is a verb, we add the lemma concatenated with “-00” as a candidate concept fragment. • Named entity pass-through: for every span of words of length 1 until 7 in the input, we add the concept fragment “(thing :name (name :op1 word1 ... :opn wordn)” as a candidate concept fragment, where n is the length of the span, and “word1” and “wordn” are the first and last words in the fragment. We use the following features for concept identi- fication: • Fragment given words: Relative frequency estimates of the probability of a concept frag- ment given the sequence of words in the span. • Length of the matching span (number of to- kens). • Bias: 1 for any concept graph fragment. 1202 • First match: 1 if this is the first place in the sentence …

Content truncated to 500 words