SciTePress - Affiliation Profile

LORIA, France

Context (please use complete words only)

Field

Sort by

25 Papers

Towards an Online Incremental Approach to Predict Students Performance

Topics: Adaptive Educational Systems; Learning Analytics and Educational Data Mining

Authors: Chahrazed Labba and Anne Boyer

Affiliation: Loria, CNRS, University of Lorraine, Vandœuvre-lès-Nancy, France

Digital Object Identifier: 10.5220/0012620300003693

Proceedings of CSEDU 2024, ISBN: 978-989-758-697-2

Quick Abstract

Analytical models developed in offline settings with pre-prepared data are typically used to predict students’ performance. However, when data are available over time, this learning method is not suitable anymore. Online learning is increasingly used to update the online models from stream data. A rehearsal technique is typically used, which entails re-training the model on a small training set that is updated each time new data is received. The main challenge in this regard is the construction of the training set with appropriate data samples to maintain good model performance. Typically, a random selection of samples is made, which can deteriorate the model’s performance. In this paper, we propose a memory-based online incremental learning approach for updating an online classifier that predicts student performance using stream data. The approach is based on the use of the genetic algorithm heuristic while respecting the memory space constraints as well as the balance of class labels. In contrast to random selection, our approach improves the stability of the analytical model by promoting diversity when creating the training set. As a proof of concept, we applied it to the open dataset OULAD. Our approach achieves a notable improvement in model accuracy, with an enhancement of nearly 10% compared to the current state-of-the-art, while maintaining a relatively low standard deviation in accuracy, ranging from 1% to 2.1%.

How Far Can We Trust the Predictions of Learning Analytics Systems?

Topics: Machine Learning; Distance Education

Authors: Amal Ben Soussia and Anne Boyer

Affiliation: Université de Lorraine, LORIA, France

Digital Object Identifier: 10.5220/0012057800003470

Proceedings of CSEDU 2023, ISBN: 978-989-758-641-5

Quick Abstract

Prediction systems based on Machine Learning (ML) models for teachers are widely used in the Learning Analytics (LA) field to address the problem of high failure rates in online learning. One objective of these systems is to identify at-risk of failure learners so that teachers can intervene effectively with them. Therefore, teachers’ trust in the reliability of the predictive performance of these systems is of great importance. However, despite the relevance of this notion of trust, the literature does not propose particular methods to measure the trust to be granted to the system results. In this paper, we develop an approach to measure a teacher’s trust in the prediction accuracy of an LA system. For this aim, we define three trust granularities, including: the overall trust, trust per class label and trust per prediction. For each trust granularity, we proceed to the calculation of a Trust Index (TI) using the concepts of confidence level and confidence interval of statistics. As a proof of concept, we apply this approach on a system using the Random Forest (RF) model and real data of online k-12 learners.

A Dynamic Indicator to Model Students’ Digital Behavior

Topics: Educational Data Mining; Learning Analytics and Educational Data Mining; COVID-19 Educational Tools; COVID-19 Societal Impacts (on Education)

Authors: Oriane Dermy , Anne Boyer and Azim Roussanaly

Affiliation: Loria, Université de Lorraine, Nancy, France

Digital Object Identifier: 10.5220/0011039400003182

Proceedings of CSEDU 2022, ISBN: 978-989-758-562-3

Quick Abstract

During the first French Covid19 lockdown, students had to switch to a fully online learning mode. Therefore, understanding students’ digital behavior becomes crucial for analysts serving public institution policy. In particular, they want to determine and interpret the evolution of students’ digital behavior. This paper aims to offer them indicators. We propose to study generic student logs corresponding to standard digital workspace services. Therefore, this paper contributes to the scientific question: Can we give an easy-to-interpret and visual indicator to model students’ behavior changes from poor and generic data? We first verify that we can extract epidemic-specific temporal patterns on these logs using Contrast Mining. These patterns represent students’ behaviors and pace. Then, we propose a new method called Temporal Pattern Histories (TPH), representing the evolution of the temporal patterns’ over time. It is a dynamic representation of students’ digital behavior. Using this method, we present graphically abrupt changes during the Cov19 lockdown, and we give some hypotheses about these results. This case study proves the relevance of TPH to detect and analyze students’ behavioral changes in an interpretive way. This approach has the advantage of representing the global evolution of students’ behavior without giving students specific information.

A New Way to Characterize Learning Datasets

Topics: Machine Learning; Adaptive Educational Systems; Learning Analytics and Educational Data Mining

Authors: Célina Treuillier ^{1

;

2} and Anne Boyer ^{1

;

2}

Affiliations: ¹ LORIA, 615 Rue du Jardin Botanique, 54506 Vandoeuvre-lès-Nancy, France , ² Lorraine University, 34 Cours Léopold, 54000 Nancy, France

Digital Object Identifier: 10.5220/0010982500003182

Proceedings of CSEDU 2022, ISBN: 978-989-758-562-3

Quick Abstract

The student’s interaction with Virtual Learning Environments (VLE) produces a large amount of data, known as learning traces, which is commonly used by the Learning Analytics (LA) domain to enhance the learning experience. Digital learning systems are generally based on the processing of these traces and must be able to adapt to different student profiles. However, the information provided in raw traces is diversified and can’t be directly used for the profile identification task: it requires defining learning indicators pedagogically relevant, and measurable directly from learning traces, and then classify learners profiles according to these indicators. The paper’s main contribution remains on the characterization of LA datasets both in terms of groups sizes and observed digital behaviors. It answers the lack of clearly stated information for LA systems developers, who need to ensure that their algorithms do not introduce bias, especially by disfavoring specific categories of students, which would only worsen existing inequalities in the student population. To go further, the embodiment of these identified profiles by translating them into learner personas also participates in the improvement of the explicability of LA outcomes by providing easy-to-interpret descriptions of students. These personas consist of fictitious representative student profiles, expressing different needs and learning objectives to which the LA systems must respond.

Assess Performance Prediction Systems: Beyond Precision Indicators

Topics: Learning Analytics; Analytics in e-learning Environments; Analytics in e-assessment Applications

Authors: Amal Ben Soussia , Chahrazed Labba , Azim Roussanaly and Anne Boyer

Affiliation: Université de Lorraine, LORIA, France

Digital Object Identifier: 10.5220/0011124300003182

Proceedings of A2E 2022, ISBN: 978-989-758-562-3

Quick Abstract

The high failure rate is a major concern in distance online education. In recent years, Performance Prediction Systems (PPS) based on different analytical methods have been proposed to predict at-risk of failure learners. One of the main studied characteristics of these systems is its ability to provide accurate early predictions. However, these systems are usually assessed using a set of evaluation measures (e.g. accuracy, precision) that do not reflect the precocity, continuity and evolution of the predictions over time. In this paper, we propose to enrich the existing indicators with time-dependent ones including earliness and stability. Further, we use the Harmonic Mean to illustrate the trade-off between the predictions earliness and the accuracy. In order to validate the relevance of our indicators, we used them to compare four different PPS for predicting at-risk of failure learners. These systems are applied on real data of K-12 learners enrolled in an online physics-chemistry module.

Can Matrix Factorization Improve the Accuracy of Recommendations Provided to Grey Sheep Users?

Topics: Web Services and Web Engineering; User-Centric Systems; Recommendation Systems; Social Web Intelligence

Authors: Benjamin Gras , Armelle Brun and Anne Boyer

Affiliation: Université de Lorraine - LORIA, France

Digital Object Identifier: 10.5220/0006302700880096

Proceedings of WEBIST 2017, ISBN: 978-989-758-246-2

Quick Abstract

Matrix Factorization (MF)-based recommender systems provide on average accurate recommendations, they do consistently fail on some users. The literature has shown that this can be explained by the characteristics of the preferences of these users, who only partially agree with others. These users are referred to as Grey Sheep Users (GSU). This paper studies if it is possible to design a MF-based recommender that improves the accuracy of the recommendations provided to GSU. We introduce three MF-based models that have the characteristic to focus on original ways to exploit the ratings of GSU during the training phase (by selecting, weighting, etc.). The experiments conducted on a state-of-the-art dataset show that it is actually possible to design a MF-based model that significantly improves the accuracy of the recommendations, for most of GSU.

Related Ontology Subjects/Areas/Topics (6)

Enterprise Information Systems		Recommendation Systems		Web Information Systems and Technologies
Internet Technology		Software Agents and Internet Computing		Web Services and Web Engineering

Identifying Users with Atypical Preferences to Anticipate Inaccurate Recommendations

Topics: User Modeling; Recommendation Systems

Authors: Benjamin Gras , Armelle Brun and Anne Boyer

Affiliation: Université de Lorraine, France

Digital Object Identifier: 10.5220/0005412703810389

Proceedings of WEBIST 2015, ISBN: 978-989-758-106-9

Quick Abstract

The social approach in recommender systems relies on the hypothesis that users’ preferences are coherent between users. To recommend a user some items, it uses the preferences of other users, who have preferences similar to those of this user. Although this approach has shown to produce on average high quality recommendations, which makes it the most commonly used approach, some users are not satisfied. Being able to anticipate if a recommender will provide a given user with inaccurate recommendations, would be a major advantage. Nevertheless, little attention has been paid in the literature to studying this particular point. In this work, we assume that a part of the users who are not satisfied do not respect the assumption made by the social approach of recommendation: their preferences are not coherent with those of others; they have atypical preferences. We propose measures to identify these users, upstream of the recommendation process, based on their profile only (their preferences). The experiments conducted on a state of the art corpus show that these measures allow to identify reliably a subset of users with atypical preferences, who will get inaccurate recommendations.

Related Ontology Subjects/Areas/Topics (6)

Enterprise Information Systems		Software Agents and Internet Computing		Web Information Systems and Technologies
Recommendation Systems		User Modeling		Web Interfaces and Applications

Studying Relations Between E-learning Resources to Improve the Quality of Searching and Recommendation

Topics: e-Learning Hardware and Software; Distance and E-Learning in a Global Context; Distance Education; Tools for Educational Communication and Collaboration

Authors: Nguyen Ngoc Chan , Azim Roussanaly and Anne Boyer

Affiliation: Université de Lorraine, France

Digital Object Identifier: 10.5220/0005454301190129

Proceedings of CSEDU 2015, ISBN: 978-989-758-107-6

Quick Abstract

Searching and recommendation are basic functions that effectively assist learners to approach their favorite learning resources. Several searching and recommendation techniques in the Information Retrieval (IR) domain have been proposed to apply in the Technology Enhanced Learning (TEL) domain. However, few of them pay attention on particular properties of e-learning resources, which potentially improve the quality of searching and recommendation. In this paper, we propose an approach that studies relations between e-learning resources, which is a particular property existing in online educational systems, to support resource searching and recommendation. Concretely, we rank e-learning resources based on their relations by adapting the Google's PageRank algorithm. We integrate this ranking into a text-matching search engine to refine the search results. We also combine it with a content-based recommendation technique to compute the similarity between user profile and e-learning resources. Experimental results on a shared dataset showed the efficiency of our approach.

Related Ontology Subjects/Areas/Topics (8)

Computer-Supported Education	e-Learning	Social Context and Learning Environments
Distance and e-Learning in a Global Context	e-Learning Hardware and Software	Ubiquitous Learning
Distance Education	Information Technologies Supporting Learning

Episode Rules Mining Algorithm for Distant Event Prediction

Topics: Web Mining; Information Extraction

Authors: Lina Fahed , Armelle Brun and Anne Boyer

Affiliation: Université de Lorraine and LORIA, France

Digital Object Identifier: 10.5220/0005027600050013

Proceedings of KDIR 2014/IC3K, ISBN: 978-989-758-048-2

Quick Abstract

This paper focuses on event prediction in an event sequence, where we aim at predicting distant events. We propose an algorithm that mines episode rules, which are minimal and have a consequent temporally distant from the antecedent. As traditional algorithms are not able to mine directly rules with such characteristics, we propose an original way to mine these rules. Our algorithm, which has a complexity similar to that of state of the art algorithms, determines the consequent of an episode rule at an early stage in the mining process, it applies a span constraint on the antecedent and a gap constraint between the antecedent and the consequent. A new confidence measure, the temporal confidence, is proposed, which evaluates the confidence of a rule in relation to the predefined gap. The algorithm is validated on an event sequence of social networks messages. We show that minimal rules with a distant consequent are actually formed and that they can be used to accurately predict distant events.

Related Ontology Subjects/Areas/Topics (7)

Artificial Intelligence	Knowledge-Based Systems	Symbolic Systems
Information Extraction	Soft Computing	Web Mining
Knowledge Discovery and Information Retrieval

User Semantic Model for Dependent Attributes to Enhance Collaborative Filtering

Topics: Learning User Profiles; Recommendation Systems

Authors: Sonia Ben Ticha ¹ , Azim Roussanaly ² , Anne Boyer ² and Khaled Bsaïes ³

Affiliations: ¹ Lorraine university and Tunis El Manar University, France , ² Lorraine University, France , ³ Tunis El Manar University, Tunisia

Digital Object Identifier: 10.5220/0004951102050212

Proceedings of WEBIST 2014, ISBN: 978-989-758-024-6

Quick Abstract

Recommender system provides relevant items to users from huge catalogue. Collaborative filtering and content-based filtering are the most widely used techniques in personalized recommender systems. Collaborative filtering uses only the user-ratings data to make predictions, while content-based filtering relies on semantic information of items for recommendation. Hybrid recommendation system combines the two techniques. The aim of this work is to introduce a new approach for semantically enhanced collaborative filtering. Many works have addressed this problem by proposing hybrid solutions. In this paper, we present another hybridization technique that predicts users preferences for items based on their inferred preferences for semantic information of items. For this, we design a new user semantic model by using Rocchio algorithm and we apply a latent semantic analysis to reduce the dimension of data. Applying our approach to real data, the MoviesLens 1M dataset, significant improvement can be noticed compared to usage only approach, and hybrid algorithm.

Related Ontology Subjects/Areas/Topics (3)

Enterprise Information Systems

Recommendation Systems

Software Agents and Internet Computing

Topic(s):

Title:
ISBN:
Year:
Acronym:
Subject:

Title:
Author:
Affiliation:
Subject:

Name:
Affiliation:
Country:
Conference:
Subject:

Name:
Country:
Subject:

Affiliation

Filter listed results, by