SciELO - Scientific Electronic Library Online

vol.18 issue3SIMTEX: An Approach for Detecting and Measuring Textual Similarity based on Discourse and SemanticsParaphrase and Textual Entailment Generation in Czech author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • Have no similar articlesSimilars in SciELO


Computación y Sistemas

Print version ISSN 1405-5546


CALVO, Hiram; SEGURA-OLIVARES, Andrea  and  GARCIA, Alejandro. Dependency vs. Constituent Based Syntactic N-Grams in Text Similarity Measures for Paraphrase Recognition. Comp. y Sist. [online]. 2014, vol.18, n.3, pp.517-554. ISSN 1405-5546.

Paraphrase recognition consists in detecting if an expression restated as another expression contains the same information. Traditionally, for solving this problem, several lexical, syntactic and semantic based techniques are used. For measuring word overlapping, most of the works use n-grams; however syntactic n-grams have been scantily explored. We propose using syntactic dependency and constituent n-grams combined with common NLP techniques such as stemming, synonym detection, similarity measures, and linear combination and a similarity matrix built in turn from syntactic n-grams. We measure and compare the performance of our system by using the Microsoft Research Paraphrase Corpus. An in-depth research is presented in order to present the strengths and weaknesses of each approach, as well as a common error analysis section. Our main motivation was to determine which syntactic approach had a better performance for this task: syntactic dependency n-grams, or syntactic constituent n-grams. We compare too both approaches with traditional n-grams and state-of-the-art systems.

Keywords : Paraphrase recognition; Microsoft Research paraphrase corpus; similarity measures; syntactic n-grams; constituent analysis; dependency analysis.

        · text in English     · English ( pdf )


Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License