On WordNet Semantic Classes and Dependency Parsing
Bengoetxea, Kepa and Agirre, Eneko and Nivre, Joakim and Zhang, Yue and Gojenola, Koldo

Article Structure

Abstract

This paper presents experiments with WordNet semantic classes to improve dependency parsing.

Introduction

This work presents a set of experiments to investigate the use of lexical semantic information in dependency parsing of English.

Related work

Broadly speaking, we can classify the attempts to add external knowledge to a parser in two sets: using large semantic repositories such as WordNet and approaches that use information automatically acquired from corpora.

Experimental Framework

In this section we will briefly describe the PTB-based datasets (subsection 3.1), followed by the data-driven parsers used for the experiments (subsection 3.2).

Results

In all the experiments we employed a baseline feature set using word forms and parts of speech, and an enriched feature set (WordNet or clusters).

Conclusions

This work has tried to shed light on the contribution of semantic information to dependency parsing.

Topics

WordNet

Appears in 17 sentences as: WordNet (19)
In On WordNet Semantic Classes and Dependency Parsing
  1. This paper presents experiments with WordNet semantic classes to improve dependency parsing.
    Page 1, “Abstract”
  2. Broadly speaking, we can classify the methods to incorporate semantic information into parsers in two: systems using static lexical semantic repositories, such as WordNet or similar ontologies (Agirre et al., 2008; Agirre et al., 2011; Fujita et al., 2010), and systems using dynamic semantic clusters automatically acquired from corpora (Koo et al., 2008; Suzuki et al., 2009).
    Page 1, “Introduction”
  3. 0 Does semantic information in WordNet help
    Page 1, “Introduction”
  4. 0 How does WordNet compare to automatically obtained information?
    Page 1, “Introduction”
  5. Broadly speaking, we can classify the attempts to add external knowledge to a parser in two sets: using large semantic repositories such as WordNet and approaches that use information automatically acquired from corpora.
    Page 1, “Related work”
  6. The results showed a signi-cant improvement, giving the first results over both WordNet and the Penn Treebank (PTB) to show that semantics helps parsing.
    Page 2, “Related work”
  7. (201 1) successfully introduced WordNet classes in a dependency parser, obtaining improvements on the full PTB using gold POS tags, trying different combinations of semantic classes.
    Page 2, “Related work”
  8. Base WordNet WordNet Clusters
    Page 3, “Experimental Framework”
  9. WordNet .
    Page 3, “Experimental Framework”
  10. (2011), based on WordNet 2.1.
    Page 3, “Experimental Framework”
  11. WordNet is organized into sets of synonyms, called synsets (SS).
    Page 3, “Experimental Framework”

See all papers in Proc. ACL 2014 that mention WordNet.

See all papers in Proc. ACL that mention WordNet.

Back to top.

dependency parsing

Appears in 16 sentences as: dependency parser (2) dependency parsers (5) dependency parsing (9)
In On WordNet Semantic Classes and Dependency Parsing
  1. This paper presents experiments with WordNet semantic classes to improve dependency parsing .
    Page 1, “Abstract”
  2. We study the effect of semantic classes in three dependency parsers , using two types of constituency-to-dependency conversions of the English Penn Treebank.
    Page 1, “Abstract”
  3. This work presents a set of experiments to investigate the use of lexical semantic information in dependency parsing of English.
    Page 1, “Introduction”
  4. We will apply different types of semantic information to three dependency parsers .
    Page 1, “Introduction”
  5. dependency parsing ?
    Page 1, “Introduction”
  6. (2011) found improvements in dependency parsing
    Page 1, “Introduction”
  7. We will test three different parsers representative of successful paradigms in dependency parsing .
    Page 1, “Introduction”
  8. (201 1) successfully introduced WordNet classes in a dependency parser , obtaining improvements on the full PTB using gold POS tags, trying different combinations of semantic classes.
    Page 2, “Related work”
  9. (2008) presented a semisupervised method for training dependency parsers , introducing features that incorporate word clusters automatically acquired from a large unannotated corpus.
    Page 2, “Related work”
  10. They demonstrated its effectiveness in dependency parsing experiments on the PTB and the Prague Dependency Treebank.
    Page 2, “Related work”
  11. We have made use of three parsers representative of successful paradigms in dependency parsing .
    Page 2, “Experimental Framework”

See all papers in Proc. ACL 2014 that mention dependency parsing.

See all papers in Proc. ACL that mention dependency parsing.

Back to top.

POS tags

Appears in 16 sentences as: POS tagger (1) POS tagging (1) POS tags (18)
In On WordNet Semantic Classes and Dependency Parsing
  1. Overall, we can say that the improvements are small and not significant using automatic POS tags, contrary to previously published results using gold POS tags (Agirre et al., 2011).
    Page 1, “Abstract”
  2. using MaltParser on gold POS tags .
    Page 1, “Introduction”
  3. In this work, we will investigate the effect of semantic information using predicted POS tags .
    Page 1, “Introduction”
  4. (201 1) successfully introduced WordNet classes in a dependency parser, obtaining improvements on the full PTB using gold POS tags , trying different combinations of semantic classes.
    Page 2, “Related work”
  5. We modified the system in order to add semantic features, combining them with wordforms and POS tags , on the parent and child nodes of each arc.
    Page 2, “Experimental Framework”
  6. For all the tests, we used a perceptron POS-tagger (Collins, 2002), trained on WSJ sections 2—21, to assign POS tags automatically to both the training (using 10—way jackknifing) and test data, obtaining a POS tagging accuracy of 97.32% on the test data.
    Page 3, “Results”
  7. Overall, we see that the small improvements do not confirm the previous results on Penn2Malt, MaltParser and gold POS tags .
    Page 4, “Results”
  8. One of the obstacles of automatic parsers is the presence of incorrect POS tags due to auto-
    Page 4, “Results”
  9. For example, ZPar’s LAS score on the LTH conversion drops from 90.45% with gold POS tags to 89.12% with automatic POS tags .
    Page 5, “Results”
  10. We will examine the influence of each type of semantic information on sentences that contain or not POS errors, and this will clarify whether the increments obtained when using semantic information are useful for correcting the negative influence of POS errors or they are orthogonal and constitute a source of new information independent of POS tags .
    Page 5, “Results”
  11. With this objective in mind, we analyzed the performance on the subset of the test corpus containing the sentences which had POS errors (1,025 sentences and 27,300 tokens) and the subset where the sentences had (automatically assigned) correct POS tags (1,391 sentences and 29,386 tokens).
    Page 5, “Results”

See all papers in Proc. ACL 2014 that mention POS tags.

See all papers in Proc. ACL that mention POS tags.

Back to top.

Treebank

Appears in 12 sentences as: Treebank (7) treebank (6)
In On WordNet Semantic Classes and Dependency Parsing
  1. We study the effect of semantic classes in three dependency parsers, using two types of constituency-to-dependency conversions of the English Penn Treebank .
    Page 1, “Abstract”
  2. In addition, we explore parser combinations, showing that the semantically enhanced parsers yield a small significant gain only on the more semantically oriented LTH treebank conversion.
    Page 1, “Abstract”
  3. Most experiments for English were evaluated on the Penn2Malt conversion of the constituency-based Penn Treebank .
    Page 1, “Introduction”
  4. tion 3 describes the treebank conversions, parsers and semantic features.
    Page 1, “Introduction”
  5. The results showed a signi-cant improvement, giving the first results over both WordNet and the Penn Treebank (PTB) to show that semantics helps parsing.
    Page 2, “Related work”
  6. They demonstrated its effectiveness in dependency parsing experiments on the PTB and the Prague Dependency Treebank .
    Page 2, “Related work”
  7. 3.1 Treebank conversions
    Page 2, “Experimental Framework”
  8. PermZMalt1 performs a simple and direct conversion from the constituency-based PTB to a dependency treebank .
    Page 2, “Experimental Framework”
  9. supervised approach that makes use of cluster features induced from unlabeled data, providing significant performance improvements for supervised dependency parsers on the Penn Treebank for English and the Prague Dependency Treebank for Czech.
    Page 3, “Experimental Framework”
  10. Looking at table 2, we can say that the differences in baseline parser performance are accentuated when using the LTH treebank conversion, as ZPar clearly outperforms the other two parsers by more than 4 absolute points.
    Page 4, “Results”
  11. We can also conclude that automatically acquired clusters are specially effective with the MST parser in both treebank conversions, which suggests that the type of semantic information has a direct relation to the parsing algorithm.
    Page 4, “Results”

See all papers in Proc. ACL 2014 that mention Treebank.

See all papers in Proc. ACL that mention Treebank.

Back to top.

semantic representation

Appears in 5 sentences as: semantic representation (4) semantic representations (1)
In On WordNet Semantic Classes and Dependency Parsing
  1. Finally, we will describe the different types of semantic representation that were used.
    Page 2, “Experimental Framework”
  2. We will experiment with the semantic representations used in Agirre et a1.
    Page 3, “Experimental Framework”
  3. We experiment with both full 88s and SFs as instances of fine-grained and coarse-grained semantic representation , respectively.
    Page 3, “Experimental Framework”
  4. For each semantic representation , we need to determine the semantics of each occurrence of a target word.
    Page 3, “Experimental Framework”
  5. (2011) used i) gold-standard annotations from SemCor, a subset of the PTB, to give an upper bound performance of the semantic representation , ii) first sense, where all instances of a word were tagged with their most frequent sense, and iii) automatic sense ranking, predicting the most frequent sense for each word (McCarthy et a1., 2004).
    Page 3, “Experimental Framework”

See all papers in Proc. ACL 2014 that mention semantic representation.

See all papers in Proc. ACL that mention semantic representation.

Back to top.

Penn Treebank

Appears in 4 sentences as: Penn Treebank (4)
In On WordNet Semantic Classes and Dependency Parsing
  1. We study the effect of semantic classes in three dependency parsers, using two types of constituency-to-dependency conversions of the English Penn Treebank .
    Page 1, “Abstract”
  2. Most experiments for English were evaluated on the Penn2Malt conversion of the constituency-based Penn Treebank .
    Page 1, “Introduction”
  3. The results showed a signi-cant improvement, giving the first results over both WordNet and the Penn Treebank (PTB) to show that semantics helps parsing.
    Page 2, “Related work”
  4. supervised approach that makes use of cluster features induced from unlabeled data, providing significant performance improvements for supervised dependency parsers on the Penn Treebank for English and the Prague Dependency Treebank for Czech.
    Page 3, “Experimental Framework”

See all papers in Proc. ACL 2014 that mention Penn Treebank.

See all papers in Proc. ACL that mention Penn Treebank.

Back to top.

lexical semantic

Appears in 3 sentences as: lexical semantic (2) lexical semantics (1)
In On WordNet Semantic Classes and Dependency Parsing
  1. This work presents a set of experiments to investigate the use of lexical semantic information in dependency parsing of English.
    Page 1, “Introduction”
  2. Whether semantics improve parsing is one interesting research topic both on parsing and lexical semantics .
    Page 1, “Introduction”
  3. Broadly speaking, we can classify the methods to incorporate semantic information into parsers in two: systems using static lexical semantic repositories, such as WordNet or similar ontologies (Agirre et al., 2008; Agirre et al., 2011; Fujita et al., 2010), and systems using dynamic semantic clusters automatically acquired from corpora (Koo et al., 2008; Suzuki et al., 2009).
    Page 1, “Introduction”

See all papers in Proc. ACL 2014 that mention lexical semantic.

See all papers in Proc. ACL that mention lexical semantic.

Back to top.

parsing algorithms

Appears in 3 sentences as: parsing algorithm (1) parsing algorithms (2)
In On WordNet Semantic Classes and Dependency Parsing
  1. Table 1: LAS results with several parsing algorithms , Penn2Ma1t conversion (T: p <0.05, 1;: p <0.005).
    Page 3, “Experimental Framework”
  2. Table 2: LAS results with several parsing algorithms in the LTH conversion (T: p <0.05, 1;: p <0.005).
    Page 3, “Experimental Framework”
  3. We can also conclude that automatically acquired clusters are specially effective with the MST parser in both treebank conversions, which suggests that the type of semantic information has a direct relation to the parsing algorithm .
    Page 4, “Results”

See all papers in Proc. ACL 2014 that mention parsing algorithms.

See all papers in Proc. ACL that mention parsing algorithms.

Back to top.

synset

Appears in 3 sentences as: synset (2) synsets (1)
In On WordNet Semantic Classes and Dependency Parsing
  1. WordNet is organized into sets of synonyms, called synsets (SS).
    Page 3, “Experimental Framework”
  2. Each synset in turn belongs to a unique semantic file (SF).
    Page 3, “Experimental Framework”
  3. As an example, knife in its tool sense is in the EDGE TOOL USED AS A CUTTING INSTRUMENT singleton synset , and also in the ARTIFACT SF along with thousands of words including cutter.
    Page 3, “Experimental Framework”

See all papers in Proc. ACL 2014 that mention synset.

See all papers in Proc. ACL that mention synset.

Back to top.

PDF available from www.aclweb.org.
Send us an email