<<Back to RESEARCH

Online communities: Wikis & E-learning

Online communities, like wikis and e-learning groups, base their success on the massive participation of users for non-monetary incentives (contribution to common good, reputation, learning etc.). 

Because of their voluntary nature, improving these communities strongly relies on understanding their underlying mechanisms, and developing algorithms that will sustain user participation.

This line of my work has implications for improving large-scale asynchronous collaboration.

Related Publications: Wikis

Wikis in enterprise settings: A survey​ 

Lykourentzou I., Dagka F., Papadaki K., Lepouras G., Vassilakis C. (2012), Enterprise Information Systems, 6(1), 1-53.

​In this survey we examine the use of wikis on a variety of organisational functions, from broader ones like the codification of organisational knowledge and the formulation of corporate communities of practice, to more specific ones such as software development, interactions with third parties, management activities and organisational response in crisis situations. For each of these functions, we examine related research findings to highlight the advantages and concerns raised by the wiki usage and to identify specific solutions addressing them. Based on the above findings, we discuss various aspects of the wiki usage in the enterprise and identify trends and future research directions on the field.

Planning for a successful corporate wiki​

Lykourentzou I., Djaghloul Y., Papadaki K., Dagka F., Latour T. (2011), Communications in Computer and Information Science, Digital Enterprise and Information Systems, 194, Springer Berlin Heidelberg.

In this work we provide an overview of the key factors affecting the successful implementation of corporate wikis, based on a meta-analysis of thirty application cases reported in the literature. The result of this meta-analysis is a core set of common best practices to help stakeholders plan and realize the successful integration of a wiki within a given enterprise context. 

Improving wiki article quality through crowd coordination: a resource allocation approach

Lykourentzou, I., Naudet, Y., Vergados, D. J. (2013), International Journal on Semantic Web and Information Systems, 9 (3), 105-125.

In this work we propose a crowd coordination mechanism to increase the quality of articles produced in wikis. The mechanism recommends articles to wiki users according to their skills, based on a resource allocation approach of article-to-user matching. To examine the proposed mechanism we build a simulated model of the English Wikipedia, which we parametrize and validate using longitudinal field studies. Experimental results on a series of user behavior scenarios indicate that the proposed mechanism can lead to the production of wiki articles of higher quality, compared to the respective results achieved by a fully self-coordinated wiki community.

CorpWiki: A self-regulating wiki to promote corporate collective intelligence through expert peer matching

Lykourentzou, I., Papadaki, K., Vergados, D. J., Polemi, D., Loumos, V. (2010), Information Sciences, Special Issue: Collective Intelligence, 180(1), 18-38.

One of the main challenges that organizations face is how to efficiently harness individual employee know-how to build a sustainable corpus of corporate knowledge. In this study we propose CorpWiki, a self-regulating wiki system for effective acquisition of high-quality knowledge content. Inserted articles undergo a quality assessment control by a large number of corporate peer employees. In case the quality is inadequate, CorpWiki uses a novel expert peer matching algorithm (EPM), based on feed-forward neural networks, that searches the human network of the organization to select the most appropriate peer employee who will improve the quality of the article. Performance evaluation results indicate that CorpWiki improves the final quality levels of the inserted articles as well as the time and effort required to reach them. 

Related Publications: E-learning

Dropout prediction in e-learning courses through the combination of machine learning techniques

Lykourentzou, I., Giannoukos, I., Mpardis, G., Nikolopoulos, V., Loumos, V. (2009), Computers & Education , 53 (3), 950-965. 

We propose a dropout prediction method for e-learning, based on three popular machine learning techniques and detailed student data. The machine learning techniques used are feedforward neural networks, support vector machines and probabilistic ensemble simplified fuzzy ARTMAP. Since a single technique may fail to accurately classify certain e-learning students, whereas another may succeed, we test three decision schemes, which combine in different ways the results of the three machine learning techniques. Examining the method in terms of overall accuracy, sensitivity and precision we find that it produces significantly better results than those reported in relevant literature.

Early and Dynamic Student Achievement Prediction in E-Learning Courses Using Neural Networks

Lykourentzou, I., Giannoukos, I., Mpardis, G., Nikolopoulos, V., Loumos, V. (2009), Journal of the American Society for Information Science and Technology, 60(2), 372-380.

The increasing popularity of e-learning has created the need for automated student achievement prediction mechanisms, which allow instructors to improve their courses by addressing student needs at an early stage. In this paper, we present a student achievement prediction method that is based on multiple feed-forward neural networks to dynamically predict the students' final achievement and cluster them in two virtual performance groups. Results of applying the method on a 10-week introductory level e-learning course, show that it accurately predicted student performance early on in the process, specifically at the third week, with very low misplacement rates. The method was further found to be more effective than the typically used approach of linear regression, in all prediction stages. The proposed methodology is expected to support instructors in providing better educational services and customized assistance according to the students’ predicted level of performance.