Philippe Dessus

A System To Assess the Semantic Content of Student Essays.

This paper presents Apex, a system that can automatically assess a student essay based on its content. It relies on Latent Semantic Analysis, a tool which is used to represent the meaning of words as vectors in a high-dimensional space. By comparing an essay and the text of a given course on a semantic basis, our system can measure how well the essay matches the text. Various assessments are presented to the student regarding the topic, the outline and the coherence of the essay. Our experiments yield promising results.

ReaderBench, an Environment for Analyzing Text Complexity and Reading Strategies

ReaderBench is a multi-purpose, multi-lingual and flexible environment that enables the assessment of a wide range of learners productions and their manipulation by the teacher. ReaderBench allows the assessment of three main textual features: cohesion-based assessment, reading strategies identification and textual complexity evaluation, which have been subject to empirical validations. ReaderBench covers a complete cycle, from the initial complexity assessment of reading materials, the assignment of texts to learners, the capture of metacognitions reflected in ones textual verbalizations and comprehension evaluation, therefore fostering learners self-regulation process.

Textual complexity and discourse structure in computer-supported collaborative learning

Computer-Supported Collaborative Learning (CSCL) technologies play an increasing role simultaneously with the appearance of the Social Web. The polyphonic analysis method based on Bakhtins dialogical model reflects the multi-voiced nature of a CSCL conversation and the related learning processes. We propose the extension of the model and the previous applications of the polyphonic method to both collaborative CSCL chats and individual metacognitive essays performed by the same learners. The model allows a tight correlation between collaboration and textual complexity, all integrated in an implemented system, which uses Natural Language Processing techniques.

ReaderBench: Automated evaluation of collaboration based on cohesion and dialogism

As Computer-Supported Collaborative Learning (CSCL) gains a broader usage, the need for automated tools capable of supporting tutors in the time-consuming process of analyzing conversations becomes more pressing. Moreover, collaboration, which presumes the intertwining of ideas or points of view among participants, is a central element of dialogue performed in CSCL environments. Therefore, starting from dialogism and a cohesion-based model of discourse, we propose and validate two computational models for assessing collaboration. The first model is based on a cohesion graph and can be perceived as a longitudinal analysis of the ongoing conversation, thus accounting for collaboration from a social knowledge-building perspective. In the second approach, collaboration is regarded from a dialogical perspective as the intertwining or synergy of voices pertaining to different speakers, therefore enabling a transversal analysis of subsequent discussion slices.

Using Production to Assess Learning: An ILE That Fosters Self-Regulated Learning

Current systems aiming at engaging students in Self-Regulated Learning processes are often prompt-based and domain-dependent. Such metacognitive prompts are either difficult to interpret for novices or ignored by experts. Although domain-dependence per se cannot be considered as a drawback, it is often due to a rigid structure which prevents from moving to another domain. We detail here Apex, a two-loop system which provides texts to be learned through summarization. In the first loop, called Reading, the student formulates a query and is provided with texts related to this query. Then the student judges whether each text presented could be summarized. In the second loop, called Writing, the student writes out a summary of the texts, then gets an assessment from the system. In order to automatically perform various comprehension-centered tasks (i.e., texts that match queries, assessment of summaries), our system uses LSA (Latent Semantic Analysis), a tool devised for the semantic comparison of texts.

Validating the Automated Assessment of Participation and of Collaboration in Chat Conversations

As Computer Supported Collaborative Learning (CSCL) gains a broader usage as a viable alternative to classic educational scenarios, the need for automated tools capable of supporting tutors in the time consuming process of analyzing conversations becomes more stringent. Moreover, in order to fully explore the benefits of such scenarios, a clear demarcation must be made between participation or active involvement, and collaboration that presumes the intertwining of ideas or points of view with other participants. Therefore, starting from a cohesion-based model of the discourse, we propose two computational models for assessing collaboration and participation. The first model is based on the cohesion graph and can be perceived as a longitudinal analysis of the ongoing conversation, thus accounting for participation from a social knowledge-building perspective. In the second approach, collaboration is regarded from a dialogical perspective as the intertwining or overlap of voices pertaining to different speakers, therefore enabling a transversal analysis of subsequent discussion slices.

Towards an integrated approach for evaluating textual complexity for learning purposes

Understanding a text in order to learn is subject to modeling and is partly dependent to the complexity of the read text. We transpose the evaluation process of textual complexity into measurable factors, identify linearly independent variables and combine multiple perspectives to obtain a holistic approach, addressing lexical, syntactic and semantic levels of textual analysis. Also, the proposed evaluation model combines statistical factors and traditional readability metrics with information theory, specific information retrieval techniques, probabilistic parsers, Latent Semantic Analysis and Support Vector Machines for best-matching all components of the analysis. First results show a promising overall precision (>50%) and near precision (>85%).

ReaderBench: An Integrated Tool Supporting both Individual and Collaborative Learning

The core of our ReaderBench software framework exposes a unified vision for predicting and assessing comprehension in both individual and collaborative learning scenarios. ReaderBench aims to improve both the quality and the classification of the analyzed documents by using an expanded range of criteria such as: morphology, semantics, discourse analysis with emphasis on polyphony and dialogism, thus providing reliable support for both tutors and students across a range of educational settings. ReaderBench uses a unitary cohesion-based representation of discourse applied into three major directions, all tightly connected by the underlying model and the Natural Language Processing (NLP) computations: reading strategies, textual complexity, and collaboration evaluation in Computer Supported Collaborative Learning (CSCL) conversations.

ReaderBench: An Integrated Cohesion-Centered Framework

Dascalu, M., Stavarache, L.L., Dessus, P., Trausan-Matu, S., McNamara, D.S., & Bianco, M. (2015). ReaderBench: An Integrated Cohesion-Centered Framework. In G. Conole, T. Klobucar, C. Rensing, J. Konert & E. Lavoue (Eds.), 10th European Conf. on Technology Enhanced Learning (pp. 505–508). Toledo, Spain: Springer.

An Overview of LSA-Based Systems for Supporting Learning and Teaching

We present an overview of LSA-based systems that have been used in instructional settings. Current research on this subject does not take into account the cognitive aspects of learning and teaching, and describes the systems at a technical level. We propose a cognitive-based classification of these systems that can lead to the design of novel LSA-based applications.