Amalia Todirascu

Multiword Expression Processing: A Survey

Multiword expressions (MWEs) are a class of linguistic forms spanning conventional word boundaries that are both idiosyncratic and pervasive across different languages. The structure of linguistic processing that depends on the clear distinction between words and phrases has to be re-thought to accommodate MWEs. The issue of MWE handling is crucial for NLP applications, where it raises a number of challenges. The emergence of solutions in the absence of guiding principles motivates this survey, whose aim is not only to provide a focused review of MWE processing, but also to clarify the nature of interactions between MWE processing and downstream applications. We propose a conceptual framework within which challenges and research contributions can be positioned. It offers a shared understanding of what is meant by “MWE processing,” distinguishing the subtasks of MWE discovery and identification. It also elucidates the interactions between MWE processing and two use cases: Parsing and machine translation. Many of the approaches in the literature can be differentiated according to how MWE processing is timed with respect to underlying use cases. We discuss how such orchestration choices affect the scope of MWE-aware systems. For each of the two MWE processing subtasks and for each of the two use cases, we conclude on open issues and research perspectives.

Using Semantics for Efficient Information Retrieval

The paper presents a system for querying a base of HTML documents in natural language for a limited domain. Different small French corpora (heart surgery, newspaper articles, papers on natural language processing) are used for testing the prototype. The domain knowledge, represented in description logics (DL), is used for filtering the results of the search and is extended dynamically as result of DL inference mechanisms. The system uses shallow natural language parsing techniques and DL reasoning mechanisms are used to handle incomplete or incorrect user queries.

Extracting Collocations in Contexts

The aim of this paper is to develop (i) a general framework for the analysis of verb-noun (VN) collocations in English and Romanian, and (ii) a system for the extraction of VN-collocations from large tagged and annotated corpora. We identify VN-collocations in two steps: (i) by calculation of the frequent lexical co-occurrences of each VN-pair, and (ii) the identification of the most typical lexico-grammatical constructions in which each VN-pair is involved in.

Using Cognates to Improve Lexical Alignment Systems

In this paper, we describe a cognate detection module integrated into a lexical alignment system for French and Romanian. Our cognate detection module uses lemmatized, tagged and sentence-aligned legal parallel corpora. As a first step, this module apply a set of orthographic adjustments based on orthographic and phonetic similarities between French - Romanian pairs of words. Then, statistical techniques and linguistic information (lemmas, POS tags) are combined to detect cognates from our corpora. We automatically align the set of obtained cognates and the multiword terms containing cognates. We study the impact of cognate detection on the results of a baseline lexical alignment system for French and Romanian. We show that the integration of cognates in the alignment process improves the results.

RefGen: Identifying Reference Chains to Detect Topics

In this paper, we present RefGen, the main module of a topic detection system used to improve a search engine by topic indexing. RefGen identifies reference chains and it uses genre specific properties of reference chains and (Ariel 1990)’s accessibility theory. RefGen checks several strong and weak constraints (lexical, morphosyntactic and semantic filters) to automatically identify coreference relations between referential expressions. We present the first results obtained by RefGen from a public reports corpus.

RefGen: A tool for reference chains identification

In this paper we present RefGen, a reference chain identification module for French. RefGen algorithm uses genre specific properties of reference chains and an accessibility measure to find the mentions of the referred entity. The module applies strong and weak constraints (lexical, morpho-syntactic, and semantic) to automatically identify coreference relations between referential expressions. We evaluate the results obtained by RefGen from a public reports corpus and we discuss the importance of the genre-dependent parameters to improve the reference chain identification.