This paper presents a method for the syntactic simplification of French texts. Syntactic simplification aims at making texts easier to understand by simplifying complex syntactic structures that hinder reading. Our approach is based on the study of two parallel corpora (encyclopaedia articles and tales). It aims to identify the linguistic phenomena involved in the manual simplification of French texts and organise them within a typology. We then propose a syntactic simplification system that relies on this typology to generate simplified sentences. The module starts by generating all possible variants before selecting the best subset. The evaluation shows that about 80% of the simplified sentences produced by our system are accurate.

This chapter presents the results of a systematic review of the literature on dialogue-based computer-assisted language learning (CALL), resulting in a conceptual framework for research on the matter. Applications allowing a learner to have a conversation in a foreign language with a computer have been studied from various perspectives and under different names (dialogue systems, conversational agents, chatbots...). Considering the fragmentation of what we identify under the term dialogue-based CALL, we attempt to offer a structured overview of these efforts into a conceptual framework. Through a methodical search strategy, we collected a corpus of 343 publications. From this corpus, we formalized an operational definition of dialogue-based CALL, which allowed us to identify 96 relevant systems. Analysing the type of dialogue they offer, on a continuum of constraints on form and meaning, we propose to classify those systems into four groups. We have called these branching, form-focused, goal-oriented and reactive systems, and we describe their corresponding interactional, instructional and technological traits. We summarise the main results from empirical studies on such systems, distinguishing observational, survey and experimental studies, and discuss the impact of dialogue-based CALL on motivation and L2 development, identifying positive evidence on both outcomes. Finally, we propose two main avenues for future research: relative effectiveness of dialogue-based CALL approaches, and dialogue systems as an environment for testing second language acquisition (SLA) hypotheses.

No abstract

Reading is known to be an essential task in language learning, but finding the appropriate text for every learner is far from easy. In this context, automatic procedures can support the teacher's work. Some tools exist for English, but at present there are none for French as a foreign language (FFL). In this paper, we present an original approach to assessing the readability of FFL texts using NLP techniques and extracts from FFL textbooks as our corpus. Two logistic regression models based on lexical and grammatical features are explored and give quite good predictions on new texts. The results shows a slight superiority for multinomial logistic regression over the proportional odds model.

Nous étudions les indicateurs de la cohésion (chaînes de référence et anaphoriques) par rapport à la complexité des textes au sein de deux corpus français qui s’adressent à plusieurs types de public (enfants/adultes pour le premier, apprenants de FLE pour le second). Cette étude a comme objectif la modélisation des aspects de cohésion textuelle des textes en vue de faciliter leur accessibilité pour les lecteurs. Pour ce faire, nous avons annoté les chaînes de référence dans les deux corpus. Nous comparons les propriétés de ces chaînes (longueur, nature des maillons, fonctions syntaxiques), en relation avec le niveau de difficulté, dans des textes informatifs et narratifs.