Multilingual collocation extraction with a syntactic parser

Abstract

An impressive amount of work was devoted over the past few decades to collocation extraction. The state of the art shows that there is a sustained interest in the morphosyntactic preprocessing of texts in order to better identify candidate expressions; however, the treatment performed is, in most cases, limited (lemmatization, POS-tagging, or shallow parsing). This article presents a collocation extraction system based on the full parsing of source corpora, which supports four languages: English, French, Spanish, and Italian. The performance of the system is compared against that of the standard mobile-window method. The evaluation experiment investigates several levels of the significance lists, uses a fine-grained annotation schema, and covers all the languages supported. Consistent results were obtained for these languages: parsing, even if imperfect, leads to a significant improvement in the quality of results, in terms of collocational precision (between 16.4 and 29.7%, depending on the language; 20.1% overall), MWE precision (between 19.9 and 35.8%; 26.1% overall), and grammatical precision (between 47.3 and 67.4%; 55.6% overall). This positive result bears a high importance, especially in the perspective of the subsequent integration of extraction results in other NLP applications.

Syntactic concordancing and multi-word expression detection

“…Once the target file has been found, the sentence that is likely to be the translation of the source sentence is identified using an in-house sentence alignment method (Nerima et al, 2003;Seretan, 2008). This method consists of comparing the relative lengths of paragraphs in the source and target documents in order to find a paragraph alignment, and then assuming a 1:1 match for sentences inside a paragraph.…”

“…To evaluate the performance of our extractor based on deep parsing, we performed several cross-language evaluation experiments in which we compared the two approaches, i.e., the syntax-based approach vs. the syntax-free approach (Seretan, 2008). For example, in an experiment conducted on English, French, Italian and Spanish data consisting of 2,000 pairs sampled from different levels of the significance list -ranging from top to 10% of the list -we measured the extraction precision by taking into account reference annotations produced by expert linguists.…”

On-line and off-line translation aids for non-native readers

“…The ability of Twic/TwicPen to handle expressions comes from the quality of the linguistic analysis provided by the multilingual Fips parser and of the collocation knowledge base (Seretan et al, 2004(Seretan et al, , 2008. A sample analysis is given in (7b), showing how extraposed elements are connected with canonical empty positions, as assumed by generative linguists.…”

A tool for multi-word expression extraction in modern Greek using syntactic parsing

“…As we already mentioned, the Greek MWE extractor is part of FipsCo, a larger extraction system based on a symbolic parsing technology (Seretan, 2008) which we previously applied on text corpora in other languages. The recent development of the Greek parser enabled us to extend it and apply it to Greek.…”

“…The tool relies on a symbolic parsing technology, and is part of FipsCo, a larger extraction system (Seretan, 2008) which has previously been used to build MWE resources for other languages, including English, French, Spanish, and Italian. Its extension to Greek will ultimately enable the inclusion of this language in the list of languages supported by an in-house translation system.…”