Multiword Expression Processing: A Survey

Abstract

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.

A Lattice Based Algebraic Model for Verb Centered Constructions

“…verb subcategorization, TAG or FrameNet), main difference can be phrased as follows. Firstly, our model puts great emphasis on filled places, and accordingly on complex proper VCCs (which have filled places and possibly free places as well), connecting our approach to multiword expression processing [2]. Secondly, the aim of our model is to represent not just one VCC but all VCCs of a corpus together including their relationships to each other, in order to be able to tackle proper VCCs based on this combined model.…”

ER documents categorization and explanation

“…Many errors were caused by typos: even the trivial lack of a space between two words may prevent the tokenizer from correctly recognizing the terms involved in the linguistic expression, and tools could be adopted that have been designed for the interactive correction and semantic annotation, also with special focus on narrative clinical reports [32,33]. Additionally, one desideratum would be individuating multiword expressions such as 'neck of the bottle' or 'lacerated bruised wound' that need to be handled as a whole (and that, conversely, cannot be dealt with in a token by token mode) [34]. Unfortunately, in the considered domain and for the considered text excerpts, standard approaches such as mwetoolkit [35] are frequently mislead to such an extent that their adoption does not ensure substantial processing advantage.…”

A Neural Graph-based Approach to Verbal MWE Identification

“…They tested different network architectures, e.g., a layered feed-forward network and a recurrent neural network, and all of them outperformed more traditional MWE identification methods. The approaches based on deep learning have the advantage that they can easily leverage pre-trained word vectors as features (Constant et al, 2017;Taslimipoor and Rohanian, 2018;Ehren et al, 2018). The method described in this work also relies on pretrained word vectors.…”

“…One of the main tasks that constitute MWE processing is the automatic identification of MWEs in running text which can be used as a preprocessing step for parsing or machine translation. MWE identification can be seen as a sequence labeling task similar to named entity recognition (NER): A system receives sequences of tokens as input and outputs the same sequences with annotation labels added to it (Constant et al, 2017). As in NER, most parts of the sequence will belong to the negative class, that is, the majority of words is not part of an MWE.…”