Knowledge Rrich Context for Terminology Translation – CRISTAL

The globalization of commercial trade has made it necessary for all companies to communicate with their partners, clients, and employees in their native language. Often, employees must also be capable of easily communicating in a foreign language, particularly in their field of expertise. Terminology competency (either multilingual or monolingual) is then essential, as company documents must be created in a coherent and time-efficient manner. However, linguistic resources adapted to companies’ specific technical fields are exceedingly rare. Furthermore, technical knowledge, and the terms used to describe it, evolve very rapidly, which makes the manual creation of such resources cost-heavy and susceptible to being quickly obsolete.

Computer-assisted translation tools, as well as other innovative terminology management tools, address this problem, by meeting the needs of private companies and public bodies for resources to aid and manage multilingual and monolingual writings. Currently, these tools are able to extract a veritable dictionary of the company, which gives information about, translates, and gives access to terms that are unique to the company as well as technical terms that are related to the business sector in question.

While access to expressions, technical terms and their translations is crucial to communication, these terms must still be put into context to “maximize their potential”. Simply having access to a term or its translation is not enough; one must be able to (1) use the term correctly (2) understand its exact meaning. The contextualization of terms is a two-fold process: (1) in documents: the user must have access to information regarding the use of the term. This requires the automatic extraction of Linguistic Knowledge-Rich Contexts; (2) from a conceptual point of view: the user must have access to the semantic or conceptual relationships between terms in order to have a better understanding of their meanings. This requires the automatic extraction of Conceptual Knowledge-Rich Contexts.

The goal of the CRISTAL project is to develop a method of extracting Knowledge-rich Contexts (KRC) in order to create new and innovative types of dictionaries. These will feature, for each term and its possible translations, a terminological index listing the KRCs and the knowledge that they contain. The extraction of such contexts requires the creation of advanced and robust algorithms that are able to extract knowledge and to automatically and efficiently analyze linguistic and semantic content in unstructured plain texts, no matter the language. Our goal is to develop a new generation of tools to assist translation and terminology management, which will facilitate multilingual communication.