Conference Paper

Pensum, un système d’aide à la compréhension de cours àdistance

Conference Paper

Pensum, un système d’aide à la compréhension de cours àdistance

If you want to read the PDF, try requesting it from the authors.

Abstract

Cet article a pour but de présenter Pensum, un système permettant à des étudiants à distance de produire des synthèses de cours dans le but de les comprendre, et d’en avoir des retours automatiques fondés sur une analyse sémantique. Nous décrivons une première utilisation de ce logiciel, auprès de 17 étudiants de 2e année de Master en sciences de l’éducation, ayant soit le rôle d’apprenant soit le rôle de tuteur, et analysons trois types de données : traces, bugs et questionnaire d’opinion. Les résultats montrent que les participants parviennent rapidement à utiliser les principales fonctionnalités de Pensum et que, s’il reste encore à le tester en situation réelle, son utilisabilité permet d’en faire un outil pouvant promouvoir l’apprentissage autorégulé de cours à distance.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Pensum (Loiseau, Dupré & Dessus, 2011 ) est l'un des prototypes développés au sein du projet européen Language Technologies for Life-long Learning (LTfLL), ayant pour but de proposer des outils pour faciliter l'apprentissage tout au long de la vie 1 . Ce système vise à améliorer la production de synthèses de cours dans un contexte d'apprentissage à distance . ...
... Ces retours font suite à une analyse sémantique de la synthèse et du texte-source et sont de trois types : la mise en évidence des sauts de cohérence entre les phrases de la synthèse réalisée, des phrases de la synthèse sans lien avec le texte source et des phrases du texte source non reprises dans la synthèse . Ainsi, l'étudiant peut choisir de modifier le contenu de sa synthèse en accord avec le feedback fourni par Pensum ou bien de le remettre en question par un clic droit s'il estime que sa production satisfait ses intentions (Loiseau et al., 2011). ...
... Ces tutrices sont étudiantes en Master 2 de sciences de l'éducation de l'université Grenoble-2. Il est à noter que ces étudiantes avaient déjà participé à la validation d'une version précédente de Pensum (Loiseau et al., 2011). Après une formation sur l'utilisation du logiciel, chaque tutrice s'est vu attribuer un nombre équivalent d'étudiants et a été pouvait répondre à toute question concernant le logiciel et l'organisation de l'étude. ...
Article
Dans le cadre du projet européen Language Technologies for Lifelong Learning, nous avons développé Pensum, un système informatique d’aide à la rédaction de synthèse de textes pour les apprentissages à distance. L’étude de validation de ce logiciel auprès d’étudiants du CNED nous a permis d’examiner les modalités d’appropriation du système dans une utilisation à distance. L’étude était composée d’une période de test associée à un questionnaire et une série d’entretiens. Les différents indicateurs utilisés (ex : efficience perçue, charge cognitive) nous permettent de faire des propositions plus générales quant à l’usage et au non-usage des TIC.
... Des rétroactions peuvent également l'aider à réguler et évaluer son activité [3]. Nous avons conçu et implémenté deux EPA, Pensum [4] et ReaderBench [5, 6], partageant certaines caractéristiques : ils utilisent des techniques d'IA (comme LSA, Latent Semantic Analysis ou analyse sémantique latente, [7]) ou de TAL (comme l'étiquetage grammatical) ; ils proposent des activités de lecture et d'écriture plutôt génériques, pouvant s'intégrer à de multiples activités d'apprentissage ; enfin, leurs rétroactions sont générées automatiquement. ReaderBench peut être vu comme la généralisation et l'enrichissement de certaines fonctionnalités de Pensum— ReaderBench inclut également des fonctionnalités d'analyse de forums et chats, qui ne sont pas décrites ici [6]. ...
... Les méthodes complexes utilisées dans Pensum et ReaderBench rendent leur conception et leur réglage difficiles. De plus, leur utilisation est peu évidente sans information précise sur la manière dont ils fonctionnent, voire sans formation [4]. En effet, les utilisateurs (enseignants ou apprenants), peuvent difficilement faire des analogies avec des logiciels existants et d'usage courant, ce qui rend le comportement des logiciels peu prédictible. ...
... À partir de ce scénario d'assistance, nous avons intégré dans Pensum (Loiseau, Dupré & Dessus, 2011 ; Villiot-Leclercq, Mandin, Dessus, & Zampa, 2010, etArmitt et al., 2011 ; Hensgens et al., 2009Sadler, 1998). Le scénario d'assistance a donc été adapté pour que le système propose à l'étudiant une meilleure emprise sur le feedback : si l'apprenant considère que le système est trop sévère ou trop tolérant, il a la possibilité d'en modifier le niveau de tolérance, mais aussi de contester ses indications (par exemple, en signalant des liens sémantiques entre les phrases). ...
... À partir de ce scénario d'assistance, nous avons intégré dans Pensum (Loiseau, Dupré & Dessus, 2011 ; Villiot-Leclercq, Mandin, Dessus, & Zampa, 2010, et le site http://webu2.upmf-grenoble.fr/sciedu/ltfll-pensum/), trois types de feedback évaluant : la cohérence « interphrase » du texte rédigé par l'apprenant ; le hors sujet, dans lequel les phrases de la synthèse ne semblent liées à aucune phrase des textes du cours ; la couverture, pour laquelle les phrases du cours ne semblent reprises dans aucune phrase de l'apprenant. ...
Article
Full-text available
This paper is intended for trainers interested in the implementation of a partially automated online human support scheme as part of distance learning. It is aimed to present an instructional design method based on scenarios and illustrate it through providing support to better understand text summaries. This situation was tested in an online application before being implemented in an academic course.
... Pensum (Loiseau, Dupré, & Dessus, 2011) est un système d'aide à la rédaction de synthèses. Il vise à laisser l'apprenant relativement libre dans son apprentissage, tout en lui permettant, à tout moment, de prendre une part active à son évaluation (Butler & Winne, 1995). ...
Article
Full-text available
Tout enseignement est à distance. De nouveaux environnements informatisés peuvent évaluer automatiquement diverses productions d’apprenants et leur procurer divers renseignements sur ces dernières. En révélant d’autres distances que physiques (épistémiques et sociales), et en s’agrégeant au sein de plates-formes sociales, ces environnements offrent de nouvelles perspectives dans l’enseignement à distance et en présence. Cet article présente trois widgets évaluant des aspects divers des processus d’apprentissage et délivrant leurs évaluations sous la forme de distances (distance sémantique, cognitive individuelle ou sociale). Ces trois widgets peuvent s’interconnecter pour former un environnement personnel d’apprentissage, environnement qui a fait l’objet d’une première étude exploratoire d’usage, avec des étudiants d’informatique. English abstract: Every teaching is distant. New computer environments can automatically assess various productions of learners and provide several feedback. By revealing non-physical distances (social and epistemic), and by joining in social platforms, these environments offer new opportunities in distance learning but also in face-to-face learning. This article presents three widgets evaluating various aspects of learning processes and delivering their evaluation as a form of distance (semantic distance, individual cognitive distance and social distance). All three widgets can be interconnected to build a personal learning environment, which is the subject of a first exploratory case-study with computer science students.
Article
Full-text available
Much computer-assisted language learning (CALL) software engenders a mixture of reactions upon initial exposure. For teachers, this may obscure a more objective view of the benefits to the learner. This article considers ways in which a languages teacher or a languages department might evaluate language learning software on a pedagogical basis. It emphasises that the classroom teacher is already in possession of much of the knowledge to do this successfully. It illustrates what might be considered good practice in various areas. This is done by matching a learning and teaching approach as currently practised within the U.K. secondary school environment against features of various CALL software packages.
Article
Full-text available
Self-regulated learning (SRL) is a pivot upon which students’ achievement turns. We explain how feedback is inherent in and a prime determiner of processes that constitute SRL, and review areas of research that elaborate contemporary models of how feedback functions in learning. Specifically, we begin by synthesizing a model of self-regulation based on contemporary educational and psychological literatures. Then we use that model as a structure for analyzing the cognitive processes involved in self-regulation, and for interpreting and integrating findings from disparate research traditions. We propose an elaborated model of SRL that can embrace these research findings and that spotlights the cognitive operation of monitoring as the hub of self-regulated cognitive engagement. The model is then used to reexamine (a) recent research on how feedback affects cognitive engagement with tasks and (b) the relation among forms of engagement and achievement. We conclude with a proposal that research on feedback and research on self-regulated learning should be tightly coupled, and that the facets of our model should be explicitly addressed in future research in both areas.
Article
Full-text available
The research on formative assessment and feedback is reinterpreted to show how these processes can help students take control of their own learning, i.e. become self-regulated learners. This refor-mulation is used to identify seven principles of good feedback practice that support self-regulation. A key argument is that students are already assessing their own work and generating their own feedback, and that higher education should build on this ability. The research underpinning each feedback principle is presented, and some examples of easy-to-implement feedback strategies are briefly described. This shift in focus, whereby students are seen as having a proactive rather than a reactive role in generating and using feedback, has profound implications for the way in which teachers organise assessments and support learning.
Article
Full-text available
Writing from sources, or “discourse synthesis”, is a common but cognitively demanding reading-writing task requiring students to select, organize, and connect content from source texts as they compose their own new texts. The purpose of the present study was to investigate the effect of explicit instruction of relevant strategies and assessment criteria on the subjects' discourse synthesis processes and products. The subjects, 24 in-service teachers enrolled in the researcher's 1999–2000 course on “Reading and writing to learn”, were assigned the same discourse synthesis task – a review of the literature – at the beginning and at the end of the course. They were required to document their two performances of the task by means of a process log, to assess their pre and post instruction processes and products, and to discuss the differences between these. The findings obtained from a content analysis of the subjects' process log summaries and criterial self-assessments of products indicated significant improvement in the subjects' post instruction discourse synthesis processes and products.
Conference Paper
Full-text available
In a life-long learning context, learners have to perform a lot of writing-based activities, like document analysis, that are seldom assessed by teachers or tutors because this assessment is time-consuming. We introduce a service, Pensum, integrated in a PLE that helps students to understand a course content through writing a synthesis, as well as teachers and tutors to manage this activity. Based on natural language processing techniques, Pensum proposes just-in-time and reflexive feedback on a complex writing task.
Article
Full-text available
This paper presents a general computational cognitive model of the way a summary is assessed by teachers. It is based on models of two subprocesses: determining the importance of sentences and guessing the cognitive rules that the student may have used. All models are based on Latent Semantic Analysis, a computational model of the representation of the meaning of words and sentences. Models' performances are compared with data from an experiment conducted with 278 middle school students. The general model was implemented in a learning environment designed for helping students to write summaries.
Article
Extracts available on Google Books (see link below). For integral text, go to publisher's website : http://www.elsevierdirect.com/product.jsp?isbn=9780121098902
Article
Writing produces generally positive, but inconsistent, effects on learning. The reasons for this inconsistency are unknown. This review examines four hypotheses about writing-to-learn: Writers spontaneously generate knowledge “at the point of utterance” (Britton, 1980/1982); writers externalize ideas in text, then reread them to generate new inferences (Young and Sullivan, 1984); writers use genre structures to organize relationships among elements of text, and thereby among elements of knowledge (Newell, 1984); and writers set rhetorical goals, then solve content problems to achieve these goals (Bereiter and Scardamalia, 1987; Flower and Hayes, 1980a). These four hypotheses invoke different aspects of writing, and so are mutually compatible. The genre hypothesis has been supported by empirical research; the other three hypotheses have been tentatively supported by research concerning writing-to-learn, or indirectly supported by other research concerning learning or writing. Further investigation is needed concerning: The empirical validity of the four hypotheses, and interactions among the processes that they identify; the declarative and procedural knowledge that underpins writing-to-learn; and the educational effectiveness of applying cognitive strategy instruction to learning through writing.
Article
Les logiciels actuels d’apprentissage des langues sont, le plus souvent, peu performants didactiquement, car ils ne considèrent la langue que comme une suite de caractères. Si l’informatique peut être d’un grand secours pour l’apprentissage des langues, didacticiens et informaticiens n’admettent pas la même acception du terme « langue ». Pour les uns, il s’agit d’un système de concepts, pour les autres d’un système de formes. Cette différence s’explique lorsqu’on constate que l’informatique ne peut considérer que la forme de la langue alors que, pour la didactique, la forme n’est que la matérialisation des concepts qu’elle est censée représenter. Cette double approche est à l’origine de la plupart des imperfections des logiciels d’apprentissage des langues. Le Traitement Automatique de la Langue (TAL) est, à notre avis, la seule réponse à ce problème. Il permet de considérer les propriétés de la forme langagière et d’offrir aux didacticiens des outils de mise en œuvre de leur problématique. Plus qu’un élément de qualité, le TAL est le passage obligé pour que les logiciels d’apprentissage des langues enseignent la langue en tant que telle. Le système MIRTO tente de mettre en œuvre cette approche.
This article is a response to Black and Wiliam's 1998 review article on formative assessment. The response draws attention to a number of enduring issues in the theory and practice of formative assessment, including its often inadequate conceptualisation in research studies; the essential role of the teacher in mediating the curriculum and learning, and therefore feedback; the conditioning of students by dominant assessment practices that encourage survival habits rather than true learning; the intellectual resources teachers bring to the assessment process as a result of their extensive experience (which students do not possess); and the need to induct students explicitly into how to interpret feedback.
Article
1. Executive summary The adoption of Course Management Systems (CMSs) for web-based instruction continues to increase in today's higher education. A CMS is a software program or integrated platform that contains a series of web-based tools to support a number of activities and course management procedures (Severson, 2004). Examples of Course Management Systems are Blackboard, WebCT, eCollege, Moodle, Desire2Learn, Angel, etc. An argument for the adoption of e-learning environments using CMSs is the flexibility of such environments when reaching out to potential learners in remote areas where brick and mortar institutions are non-existent. It is also believed that e-learning environments can have potential added learning benefits and can improve students' and educators' self-regulation skills, in particular their metacognitive skills. In spite of this potential to improve learning by means of using a CMS for the delivery of e-learning, the features and functionalities that have been built into these systems are often underutilized. As a consequence, the created learning environments in CMSs do not adequately scaffold learners to improve their self-regulation skills. In order to support the improvement of both the learners' subject matter knowledge and learning strategy application, the e-learning environments within CMSs should be designed to address learners' diversity in terms of learning styles, prior knowledge, culture, and self-regulation skills. Self-regulative learners are learners who can demonstrate 'personal initiative, perseverance and adaptive skill in pursuing learning' (Zimmerman, 2002). Self-regulation requires adequate monitoring strategies and metacognitive skills. The created e-learning environments should encourage the application of learners' metacognitive skills by prompting learners to plan, attend to relevant content, and monitor and evaluate their learning. This position paper sets out to inform policy makers, educators, researchers, and others of the importance of a metacognitive e-learning approach when designing instruction using Course Management Systems. Such a metacognitive approach will improve the utilization of CMSs to support learners on their path to self-regulation. We argue that a powerful CMS incorporates features and functionalities that can provide extensive scaffolding to learners and support them in becoming self-regulated learners. Finally, we believe that extensive training and support is essential if educators are expected to develop and implement CMSs as powerful learning tools.
Article
Designing metacognitive activities that focus on both cognitive and social development is a theoretical and practical challenge. This balanced approach to metacognition concerns itself with many aspects of student development, ranging from academic competence to knowledge about the self-as-learner. In this article, I examine two basic approaches to supporting metacognition: (a) strategy training and (b) creation of a supprotive social environment for metacognition. There are also two kinds of content that are taught using these two approaches: (a) knowledge about a specific domain and (b) knowledge about the self-as-learner. These approaches and contents have been used frequently in metacognitive interventions over the past two decades. Each offers unique contributions to metacognitive development. However, programs that address these approaches and contents simultaneously are rare. Maintaining the coordination, on one hand, between strategy training and creating social supports, and on the other hand, between knowledge about the subject domain and knowledge about the self-as-learner, is a challenge for most design efforts in metacognition. Future design issues include: (a) developing a system approach to promote coordination among these approaches and contents; and (b) finding ways to build knowledge about the self-as-learner.
Article
How do people know as much as they do with as little information as they get? The problem takes many forms; learning vocabulary from text is an especially dramatic and convenient case for research. A new general theory of acquired similarity and knowledge representation, latent semantic analysis (LSA), is presented and used to successfully simulate such learning and several other psycholinguistic phenomena. By inducing global knowledge indirectly from local co-occurrence data in a large body of representative text, LSA acquired knowledge about the full vocabulary of English at a comparable rate to schoolchildren. LSA uses no prior linguistic or perceptual similarity knowledge; it is based solely on a general mathematical learning method that achieves powerful inductive effects by extracting the right number of dimensions (e.g., 300) to represent objects and contexts. Relations to other theories, phenomena, and problems are sketched.
« Research in writing instruction. What we know and what we need to know
  • G A Troia
[TROIA 07] Troia, G. A., « Research in writing instruction. What we know and what we need to know », In M. Pressley, A. K. Billman, K. H. Perry, K. E. Reffitt, J. Moorhead Reynolds (Eds.), Shaping literacy achievement. Research we have, research we need, Guilford, New York, 2007, p. 129-156.
Situation models and levels of coherence
  • I Tapiero
[TAPIERO 07] Tapiero, I., Situation models and levels of coherence, Mahwah, Erlbaum, 2007.
« The effects of new technologies on writing and writing processes
  • C A Macarthur
[MACARTHUR 06] MacArthur, C. A., « The effects of new technologies on writing and writing processes », In C. A. MacArthur, S. Graham, J. Fitzgerald (Eds.), Handbook of writing research, Guilford Press, New York, 2006, p. 248-262.
« Computational cognitive models of summarization assessment skills
  • Lemaire
[LEMAIRE et al. 05] Lemaire, B., Mandin, S., Dessus, P., Denhière, G., « Computational cognitive models of summarization assessment skills », In B. G. Bara, L. Barsalou, M. Bucciarelli (Eds.), Proceedings of the 27th Annual Conference of the Cognitive Science Society (CogSci' 2005), Erlbaum, Mahwah, 2005, p. 1266-1271.
« Fostering the cycle of self-regulation in writing learning protocols Avoiding simplicity, confronting complexity Advances in studying and designing (computer-based) powerful learning environments
  • Hübner
[HÜBNER et al. 06] Hübner, S., Nückles, M., Renkl, A., « Fostering the cycle of self-regulation in writing learning protocols », In G. Clarebout, J. Elen (Eds.), Avoiding simplicity, confronting complexity. Advances in studying and designing (computer-based) powerful learning environments, Sense Publishers, Rotterdam, 2006, p. 155-164.