Section: Overall Objectives

Overall Objectives

The Alpage team is specialised in Language modelling, Computational linguistics and Natural Language Processing (NLP). These fields are of crucial importance for the new information society. Applications of this domain of research include the numerous technologies grouped under the term of `language engineering'. This includes domains such as machine translation, question answering, information retrieval, information extraction, data mining, text simplification, automatic or computer-aided translation, automatic summarisation, foreign language reading and writing aid. From a more research-oriented point of view, experimental linguistics can be also viewed as an `application' of NLP.

NLP, the domain of Alpage, is a transdisciplinary domain: it requires an expertise in formal and descriptive linguistics (to develop linguistic models of human languages), in computer science and algorithmics (to design and develop efficient programs that can deal with such models) and in applied mathematics (to automatically acquire linguistic or general knowledge). It is one of the specificities of Alpage to put together both researchers with a background in computer science (Inria members) and researchers with a background more oriented towards linguistics, all of them working on a single topic: simulation on computers of human understanding and production of language.

Natural language understanding systems convert samples of human language into more formal representations that are easier for computer programs to manipulate. Natural language generation systems convert information from computer databases into human language. Alpage focuses on text understanding and generation (by opposition to speech processing and generation).

One specificity of NLP is the diversity of human languages it has to deal with. Alpage focuses mostly on French. One of the main objectives of the team is to develop generic linguistically relevant and computationally efficient tools and resources for French which are freely distributed. These products are dedicated to the francophone community so as to help French to be part of the new information society. However, Alpage does not ignore other languages, through collaborations, in particular with those that are already studied by its members or by long-standing collaborators (e.g., English, Spanish, Polish, Persian and others). This is of course of high relevance, among others, for language-independant modelling and multi-lingual tools and applications.

Alpage covers all linguistics domains, although not at the same level. At the creation of the team, the morphological and syntactic levels was the most developed and led to a number of applications, especially with industrial partners. However, the importance of the semantic and discourse levels has increased during the evaluation period and the interface between syntax and semantics has been better worked on. Our goal is also to apply our knowledge, tools and resources in various contexts such as research in experimental linguistics, operational applications and prototypes as well as standardisation of linguistic resources and annotations.

Our four main objectives, as reworded and updated while writing the 2015 Inria evaluation report, are the following:

  • Objective i: Towards large scale natural language understanding at the sentence level This objective covers all the work carried out on shallow processing, tagging, syntactic parsing, deep-syntactic parsing and shallow semantic parsing.

  • Objective ii : Language resource development, evaluation and use This objective covers all language resource development efforts that range from morphology to semantics including syntax, but not including supra-sentential (discourse) resources.

  • Objective iii : Modelling and parsing supra-sentential phenomena This objectives covers all efforts, including language resource development efforts, regarding discourse and other phenomena that cross sentence boundaries (e.g. anaphora).

  • Objective iv : Application domains This objectives regroups the three main application domains for Alpage: empirical linguistics, academic downstream NLP applications and industrial applications.