Group Formation Algorithms in Collaborative ... - Semantic Scholar

Group Formation Algorithms in Collaborative Learning Contexts: A Systematic Mapping of the Literature Wilmax Marreiro Cruz and Seiji Isotani Department of Computer Systems University of Sao Paulo Sao Carlos, Brazil {wilmcruz,sisotani}@icmc.usp.br

Abstract. Group Formation is a complex and important step to design effective collaborative learning activities. Through the adequate selection of individuals to a group, it is possible to create environments that foster the occurrence of meaningful interactions, and thereby, increasing robust learning and intellectual growth. Many researchers indicate that the inadequate formation of groups can demotivate students and hinder the learning process. Thus, in the field of Computer-Supported Collaborative Learning (CSCL), there are several studies focusing on developing and testing group formation in collaborative learning contexts using best practices and other pedagogical approaches. Nevertheless, the CSCL community lacks a comprehensive understanding on which computational techniques (i.e. algorithms) has supported group formation. To the best of our knowledge, there is no study aimed at gathering and analyzing the research findings on this topic using a systematic method. To fill this gap, this research conducted a systematic mapping with the objective of summarizing the studies on algorithms for group formation in CSCL contexts. Initially, by searching on six digital libraries, we collected 256 studies. Then, after a careful analysis of each study, we verified that only 48 were related to group formation applied to collaborative learning contexts. Finally, we categorized the contributions of these studies to present an overview of the findings produced by the community. This overview shows that: (i) there is a gradual increase on research published in this topic; (ii) 41% of the algorithms for group formation area based on probabilistic models; (iii) most studies presented the evaluation of tools that implement these algorithms; but (iv) only 2% of the studies provide their source code; and finally, (v) there is no tool or guideline to compare the benefits, differences and specificities of group formation algorithms available to date. As a result of this work an infographic is also available at: http://infografico.caedlab.com/mapping/gf. Keywords: Group Formation, Algorithms, CSCL, Systematic Mapping.

1

Introduction

Computer-supported Collaborative Learning (CSCL) is a pedagogical approach in which knowledge construction occurs from social interactions among individuals with N. Baloian et al. (eds.): CRIWG 2014, LNCS 8658, pp. 207–222, 2014. © Springer International Publishing Switzerland 2014

208

W.M. Cruz and S. Isotani

explicit or implicit support from computers and its technologies [3][6][10][12]. Several researchers have highlighted the potential benefits of learning through collaboration [7][11][5]. Nevertheless, learning through interactions does not occur in any situation. According to Barkley and colleagues [2] the design of collaboration is fundamental to achieve desired learning goals. Among the studies in the design of well-thought-out collaborative learning scenarios, an important task to be conducted is the formation of groups. As indicated by Dillenbourg [3], forming groups without careful considerations (i.e. randomly) often causes problems such as disproportional participation of individuals, demotivation and resistance to group work in futures activities. Isotani et al. [5] also emphasizes that group formation is the first step to design a CSCL scenario where students can learn and participate more effectively. Through the process of selecting individuals to participate in a group, one can analyze and combine characteristics such as cultural background, knowledge, skills, learning styles, roles and so on, to create a positive synergy among participants that will lead to meaningful interactions and better learning situations. Yet, due to the possibility of using several learner’s characteristics and combine them in different ways to form learning groups, this task often requires computational support to be completed successfully. In this context, there are various studies in the literature on computer-supported group formation using different approaches, presenting new algorithms, frameworks, tools, techniques, experiments, and so on. For example, the work of Soh et al. [10], describes the development of an algorithm for group formation using a multi-agent approach and a pedagogical technique known as Jigsaw [1]. In another work, Moreno et al. [7] uses a genetic algorithm approach to consider an arbitrary number of student characteristics to create groups that are more effective. Finally, Ounnas et al. [8] proposed a semantic framework to improve the performance of some existing algorithms reflecting the diversity of approaches, algorithms, inputs and attributes (such as learning style, gender, personality, and so on). Although, there are several benefits from the use of computing techniques for group formation, to the best of our knowledge, there is not study that summarize and catalogue them in a comprehensive and systematic manner. Because of that, the CSCL community lack a better understanding about how many computational approaches have been developed and used to support group formation, where to find them, how good they are, what the difference between the various existing algorithms, and so on. To answer some of these questions this work carried out a systematic mapping of the literature to collect the research on group formation in CSCL contexts. We used the method proposed by Petersen [9] to conduct our research as described in Section 2. In summary, first, we defined the research protocol and selected the most important digital libraries in the field of computing and educational technologies. By searching on these digital libraries, we collected 256 studies (i.e. published articles). Then, through a first screening of each study, we verified that 48 of them match our needs (i.e. they are related to group formation in CSCL and met the inclusion and exclusion

Group Formation Algorithms in Collaborative Learning Contexts

209

criteria defined in this work). Finally, we carefully analyzed and categorized these studies as shown in Section 3 to answer the following questions: 1) what are their main research objectives and contributions? 2) What are the algorithms used to support group formation? 3) Is the source code available for analysis and reuse? 4) Is there any study or tool that compare group formation algorithms? We conclude this paper by discussing the results and practical applications of our findings in Section 4.

2

Method

Systematic mapping is a research method that provides guidelines to conduct literature reviews [9]. It consists of methodical steps to search, interpret, synthetize and analyze the information presented in published papers related to the target domain. The use of this method aims to provide an overview of the field of interest and minimize the chances of errors during the review process. Such a systematic process also gives better control to the review activity and remove possible mistakes that may cause misleading or imprecise conclusions. In this work, we used the steps proposed by Petersen [9] to carry out the systematic mapping. It consist of five sequential steps as follows: (i) definition of research questions; (ii) conduct search for primary studies; (iii) screening of papers for inclusion and exclusion; (iv) classification scheme; and (v) data extraction and mapping of studies. Each of these steps are presented in the next subsections. 2.1

Definition of Research Questions

The focus of this systematic mapping is to identify and classify computational techniques, particularly algorithms that have been used to assist the formation of groups in collaborative learning environments with computational support. To define our objectives and then search for evidences (i.e. primary studies1) on algorithms for group formation, the following research questions were defined: RQ1: What are the main research objectives and types of contributions from studies on group formation for CSCL? RQ2: What are the most common computational techniques (i.e. algorithms) used to support group formation in collaborative learning environments? RQ3: Is the source code of group formation algorithms available for analysis and reuse? RQ4: Is there any study or tool that compare group formation algorithms in the context of collaborative learning?

1

Individual studies that contribute to provide evidences to answer specific research questions.

210

2.2


Conduct Search for Primary Studies

The search for primary studies is composed of two steps. In the first step, we define the search string considering the most relevant terms related to our topic. In the second step, we select relevant electronic databases to conduct the search. To create the search string, we used keywords (i) contained in the research questions. (ii) extracted from well-known papers; and (iii) obtained from interview with experts in the field. The result is shown in Table 1. Three main keywords have been defined: "group formation", "collaborative learning" and "algorithms", each keyword form a category that contains their respective synonyms. Table 1. Categories of keywords and their synonyms

Reference C1

Category group formation

C2

collaborative learning

C3

Algorithms

Synonyms group creation group design group composition group organization team formation team creation team design team composition team organization cooperative learning cscl csgf social learning group learning team learning Approaches methods software technique

To create the final search string, the categories C1, C2 and C3 were combined by the Boolean operator "AND", and the keywords within each category were combined by the Boolean operator "OR", as shown below: (group formation OR group creation OR group design OR group composition OR group organization OR team formation OR team creation OR team design OR team composition OR team organization) AND (collaborative learning OR cooperative learning OR cscl OR csgf OR social learning OR group learning OR team learning) AND (algorithms OR approaches OR methods OR software OR technique). In the second step, to select relevant electronic databases to conduct our search, we started analyzing the results from Dyba et al. [4] that provided a list of important


211

databases in the field of Computer Science and Engineering. Then, to focus our search, we shrink this list to obtain the databases that cover the most important conferences and journals in the field of educational technology. Thus, the following electronic databases were selected: • • • • • •

ACM Digital Library IEEE Xplore ScienceDirect – Elsevier Scopus SpringerLink Web of Science

The search engine of each selected database uses different mechanisms and standards. Thus, we adapted the search string developed in this work to each database to conduct our search. After that we conduct the search on the titles, abstracts, and keywords of articles to collect the first set of primary studies. The results obtained are shown in Table 2. On the first column, we have the name of the database and, on the second columns, the number of returned papers. Table 2. Number of primary studies obtained

Database ACM Digital Library IEEE Xplore ScienceDirect – Elsevier Scopus SpringerLink Web of Science TOTAL 2.3

Quantity 3 23 10 138 15 67 256

Screening of Papers for Inclusion and Exclusion

The first screening of the returned papers (Table 2) consist of applying a set of inclusion (I) and exclusion (E) criteria to add or remove papers from our analysis: I1. If several papers are related to the same study, only the most recent paper is selected; I2. If the paper describes more than one study, each study is assessed individually; I3. If there are versions of the same study, a short and a full, the full version must be included. E1. Papers that do not present studies relating to education; E2. Papers that do not present studies relating to group formation algorithms; E3. Papers in languages other than English or Portuguese;

212


E4. Technical reports and documents that are available in the form of summaries or presentations (gray literature) and secondary studies (i.e., systematic reviews and mapping studies). After defining the criteria for inclusion and exclusion, we read the titles and abstracts (and sometimes the introduction and conclusion) of each paper in order to identify those considered irrelevant to our work. Then, after the application of these criteria, we carefully read in full the final selection of papers and the data contained in these papers were extracted, analyzed and categorized. The final set of paper and the data analysis is presented in section 3. 2.4

Classification Scheme

To have a better understand of the contributions from each analyzed paper, we used the categories suggested by Wieringa and colleagues [13] to analyze, classify and categorize the types of studies described in the papers. The categories of study types are: • Validation Research: novel techniques that have not yet been implemented in practice. Usually used in experimental settings in laboratory. • Evaluation Research: techniques that are implemented in practice and an evaluation is conducted. This includes analysis of their implementation, benefits and drawbacks. • Solution Proposal: A solution for a problem is proposed, which can be a new solution or an extension of an existing technique. The potential benefits of the solution is presented using case studies (small examples) or other argumentations. • Philosophical Papers: papers that present a new look or direction to the field, often using taxonomies and conceptual frameworks. • Opinion Papers: Studies that express a personal opinion about whether a technique is good or bad and/or how it should be used or implemented. • Experience Papers: Contain the personal experience of the author explaining what and how something has been done in practice. 2.5

Data Extraction and Mapping of Studies

The papers were analyzed and classified according to the steps and categories presented in previous sections. We also created other categories to separate the research contributions of each paper (see Section 3). The data extracted from papers were stored and subjected to qualitative and quantitative analysis. This analysis aimed at finding evidences to answer the questions defined in Section 2.1. To organize the findings, we used a spreadsheet to document the data extraction process, which allowed us to also carry out other statistical analysis such as the number of publications per year, their venue, type, and so on.


3

213

Results and Analysis

In this Section, we present the results of our systematic mapping. The main purpose is to give an overview on how algorithms for group formation are being developed and applied in collaborative learning environments. This work was conducted over the period of four months between November 2013 to February 2014. The Figure 1 shows the execution of the steps presented in sections 2.1, 2.2 and 2.3. Initially, by conducting the search in the selected databases 256 papers returned. From these, we identified that 91 of them were stored in multiple databases, thus, we eliminated the duplications leaving only a copy of each paper in our records. Thus, 165 papers remained to be analyzed in the next step. We then applied the inclusion and exclusion criteria in all 165 papers by reading their title, abstract, introduction and conclusion leaving only 65 papers. In the last stage of this process, we carefully read the 65 papers in full, and again applied the criteria for inclusion and exclusion, where 17 papers were eliminated. The final result of this process left 48 papers that were used as evidence to answer our research questions. The list of these papers appears in the end of this document.

Fig. 1. Overview of the process of filtering papers

Before answering the main research questions of this work, it is also important to give an overview about where and when the 48 papers were published. Figure 2 shows on the x-axis the number of papers by publication type (i.e. Journal article, conference proceedings or book chapter) and on the y-axis the database where the papers were retrieved. According to Figure 2, the number of papers published in conference proceedings (26 papers) are the most common, followed by Journal articles (19 papers) and book chapters (3 papers). This result highlights the importance of conferences for dissemination of research on the topic of group formation in collaborative learning environments. It is worth to point out that none of the 48 papers is

214


associated with the Scopus database because their primary index came from other databases (e.g. papers published by IEEE are primarily indexed in IEEE Xplorer but also appear in Scopus). Nevertheless, it is also important to emphasize the importance of Scopus database during the validation and calibration of the search string. Observing the frequency of publications, we found out that 78% of the studies on this topic were published in the last seven years as shown in Figure 3. This trend of gradual increase indicates the growing importance and potential of the area. Regarding the year 2013, in which only two papers were found, it is possible that by the time we run the extraction process some studies had not yet been indexed in the databases.

Fig. 2. Number of papers by type of publication (x axis) and databases (y axis)

Fig. 3. Number of papers per year


215

In Section 2.4, we mentioned the use of the categories propose by [13] to cluster papers by type of study. Nevertheless, to answer RQ1 (what are the main research objectives and types of contributions from studies on group formation for CSCL?) we also needed to classify the papers by their research objectives. Thus, after reading all papers we proposed seven categories specific to this work as follows: • • • • • • •

Tool: the main objective of the study is to develop or extend a tool that implement a specific algorithm for group formation. Framework: the objective is to propose a model (technical foundations) that support the creation, or facilitate the use, of group formation algorithms. Investigation: the purpose of the study is to inspect an existing group formation algorithm. Improvement: the study focus on propose and implement better solutions for existing group formation algorithms. Methodology: the study aims to present best practices to use collaborative learning as well as group formation. Script: the objective of the study is propose guidelines to design CSCL activities, including steps related to group formation. Technical: the study focus on presenting a computational technique to implement group formation.

We used these categories and the ones present in Section 2.4 to create a bubble chart (Figure 4) to show the distribution of studies in each category. In the x-axis we have the study types categories and in the y-axis the research objectives. The size and number of each bubble represents the number of studies that fall into a specific x-y situation.

Fig. 4. Distribution of papers by study type (x axis) and research objectives (y axis)

216


By adding the numbers in the bubbles related to each row of Figure 4, it is possible to verify that the majority of studies, total of 17, fall in the category “Tool” and had as their main research objective to present an application that provides mechanisms to support group formation in CSCL contexts. Furthermore, 7 of these studies evaluated their results through use cases and 9 of them used data from real learning scenarios. The research objective category “Technique” is also well explored in the literature with 12 studies. 4 studies in this category aims at presenting a novel algorithm (solution) for group formation; other 2 studies validate these computational techniques in laboratory; and finally 6 studies evaluate these techniques with students in real scenarios. From another viewpoint, by adding the numbers in the bubbles related to each column of Figure 4, it is also possible to verify that most studies concentrate their effort on proposing a solution, total of 17 studies, or evaluating their results in real scenarios, total of 20 studies. This results show the maturity of the field since most studies were conducted with learners and their learning environment. To answer RQ2 (what are the most common algorithms used to support group formation in collaborative learning environments?), we analyzed each study and found out that 44 from the 48 selected studies proposed or implemented algorithms as a solution to the problem of forming groups in collaborative learning environments. Figure 5 shows the type and number of algorithms utilized by these studies. About 41% of them (18 studies) are based on probabilistic algorithms. 8 are Genetic Algorithms (GA), demonstrating the great interest of researchers in using this technique as a solution to group formation in CSCL due to their applicability to deal with a large number of variables and the possibility to rapidly generate useful (semi-)optimal solutions (i.e. groups). Moreover, 5 studies presented algorithms based on Swarm Intelligence such as PSO (Particle swarm optimization) and ACO (Ant colony optimization) to form groups. In another algorithm category, a data mining approach known as kmeans is utilized (4 studies). Among the algorithms listed as “Others” there are many different computational techniques such as the use of semantic web, ontologies, Bayesian Network, machine learning techniques, and so on. Finally the category “Unspecified” include studies that did not specify the computational technique utilized (e.g. ad-hoc group formation algorithms based on authors’ knowledge). To tackle RQ3 (is the source code of group formation algorithms available for analysis and reuse?) we extracted from the studies information about the implementation of the algorithms. We verified that 82% actually developed and implemented the algorithms for group formation in a specific CSCL environment (Figure 6). Nevertheless, although there is a high number of studies that implemented and tested their algorithms, few of them published their source code. In fact, according to our analysis only 2% of the studies did present the source code on the paper or put the code available on the Web. 23% of the studies presented only the pseudo-code and the majority, 75% of the studies, did not present any code of the implemented algorithms (Figure 7). This result is critical and disturbing, because the community do not have access to valuable data that would enable researchers to reuse the achievements (i.e. developed and tested algorithms) obtained in previous studies and build better tools and algorithms on top of that. It is also not possible to check and compare the existing group


217

Fig. 5. Amount of algorithms for each type of approach

Fig. 6. Percentage of studies developed and implemented

Fig. 7. Percentage of availability of codes used in studies

formation algorithms in several educational settings to acquire more and better knowledge about their benefits for learning; which answers RQ4 (is there any study or tool that compare group formation algorithms in the context of collaborative learning?) – No.

4

Conclusion

Group formation is an important research topic in the field of CSCL due to its potential application to increase learning benefits of individuals when working in groups. Due to the complexity of forming effective groups manually, there are several

218


algorithms to (semi-) automate such a process. In this work, we used a systematic mapping method (Section 2) to collect, analyze and summarize the research achievements on this topic. We initiated analyzing 256 papers and, after a careful inspection, we discarded 208 papers that did not fulfil the defined inclusion and exclusion criteria. The remaining 48 papers were considered as studies that produced evidences to answer the four research questions introduced in section 2.1 - RQ1: What are the main research objectives and types of contributions from studies on group formation for CSCL? RQ2: What are the most common computational techniques (i.e. algorithms) used to support group formation in collaborative learning environments? RQ3: Is the source code of group formation algorithms available for analysis and reuse? RQ4: Is there any study or tool that compare group formation algorithms in the context of collaborative learning? According to the results presented in Section 3, development of tools and conduction of evaluation research are, respectively, the main research objectives and type of studies identified in the literature (see figure 4) – answering RQ1. Furthermore, the most common computational techniques implemented and used to form groups are probabilistic algorithms (e.g. genetic and swarm intelligence algorithms), followed by data mining techniques (e.g. k-means) and multi-agent approaches (see figure 5) – answering RQ2. Unfortunately, we verified that although many studies implemented group formation algorithms, only 2% provided their source code (see Figure 7) – answering RQ3. This is problematic, since there is a lack of source codes and pseudocodes to replicate and reuse group formation algorithms. As a result, the community do not have instruments to compare, evaluate and better understand the different approaches to form groups in CSCL contexts. In fact, in our review we did not find any study or tool where the goal is to compare existing group formation algorithms - answering RQ4. We believe that through the conduction of this systematic mapping, it was possible to provide to the CSCL community an overview of the research on group formation algorithms applied to collaborative learning contexts. Besides showing the increasing number of publications and a variety of computational approaches to deal with the topic, we also identified a critical problem in existing works, namely, the lack of presenting and sharing the source code of developed algorithms. This problem can be exploited to open new and important opportunities for future research. Finally, to provide a quick visual overview of the findings presented in this paper an infographic was created and it is available at: http://infografico.caed-lab. com/mapping/gf . It also contains information that where not fully covered in this paper due to scope and space limitation. Acknowledgment. We thank FAPESP (Process: 2013/13056-4) and CNPq (Processes: 310204/2011-9; 400481/2013-8 and 470757/2013-2) for providing support for this research.


219

References 1. Aronson, E., Patnoe, S.: The jigsaw classroom: building cooperation in the classroom, 2nd edn. Addison Wesley Longman, New York (1997) 2. Barkley, E., Cross, K.P., Major, C.H.: Collaborative Learning Techniques: A Practical Guide to Promoting Learning in Groups. Jossey Bass, San Francisco (2005) 3. Dillenbourg, P.: Over-scripting CSCL: The risks of blending collaborative learning with instructional design. In: Three Worlds of CSCL. Can we support CSCL?, pp. 61–91. Open University Nederland, Heerlen (2002) 4. Dyba, T., Dingsoyr, T., Hanssen, G.K.: Applying Systematic Reviews to Diverse Study Types: An Experience Report, 225-234 (2007) 5. Isotani, S., Inaba, A., Ikeda, M., Mizoguchi, R.: An Ontology Engineering Approach to the Realization of Theory-Driven Group Formation. International Journal on ComputerSupported Collaborative Learning 4(4), 445–478 (2009) 6. Isotani, S., Mizoguchi, R., Isotani, S., Capeli, O.M., Isotani, N., de Albuquerque, A.R.P.L., Bittencourt, I.I., Jaques, P.A.: A Semantic Web-based authoring tool to facilitate the planning of collaborative learning scenarios compliant with learning theories. Computers & Education 63, 267–284 (2013) 7. Moreno, J., Ovalle, D.A., Viccari, R.M.: A genetic algorithm approach for group formation in collaborative learning considering multiple student characteristics. Computers and Education 58, 560–569 (2012) 8. Ounnas, A., Davis, H.C., Millard, D.E.: A Framework for Semantic Group Formation in Education. Educational Technology and Society 12(4), 43–55 (2009) 9. Petersen, K., Feldt, R., Shahid, M., Mattsson, M.: Systematic Mapping Studies in Software Engineering. In: Proceedings of the Evaluation and Assessment in Software Engineering, pp. 1–10 (2008) 10. Soh, L.-K., Khandaker, N., Jiang, H.: I-MINDS: A Multiagent System for Intelligent Computer- Supported Collaborative Learning and Classroom Management. International Journal on Artificial Intelligence in Education 18, 119–151 (2008) 11. Strijbos, J., Martens, R.L., Jochems, W.M.G., Broers, N.J.: The effect of functional roles on perceived group efficiency during computer-supported collaborative learning: a matter of triangulation. Computers in Human Behavior 23(1), 353–380 (2007) 12. Wang, D.-Y., Liu, Y.-C., Sun, C.-T.: A grouping system used to form teams full of thinking styles for highly debating, pp. 725–730. World Scientific and Engineering Academy and Society (WSEAS), Stevens Point, Wisconsin, USA (2006) 13. Wieringa, R., Maiden, N.A.M., Mead, N.R., Rolland, C.: Requirements engineering paper classification and evaluation criteria: a proposal and a discussion. Requirements Engineering 11(1), 102–107 (2006)

Further Reading [RP1] Abnar, S., Orooji, F., Taghiyareh, F.: An evolutionary algorithm for forming mixed groups of learners in web based collaborative learning environments. In: 2012 IEEE International Conference on Technology Enhanced Education (ICTEE), pp. 1–6 (2012) [RP2] Adán-Coello, J.M., Tobar, C.M., de Faria, E.S.J., de Menezes, W.S., de Freitas, R.L.: Forming Groups for Collaborative Learning of Introductory Computer Programming Based on Students’ Programming Skills and Learning Styles. International Journal of Information and Communication Technology Education 7, 34–46 (2011)

220


[RP3] Ardaiz-Villanueva, O., Nicuesa-Chacon, X., Brene-Artazcoz, O., Sanz de Acedo Lizarraga, M.L., Sanz de Acedo Baquedano, M.T.: Evaluation of Computer Tools for Idea Generation and Team Formation in Project-Based Learning. Computers & Education 56(3), 700–711 (2011) [RP4] Brauer, S., Schmidt, T.C.: Group formation in elearning-enabled online social networks. In: International Conference on Interactive Collaborative Learning (ICL), pp. 1–8 (2012) [RP5] Cadavid, J.M., Ovalle, D.A., Vicari, R.M.: A genetic algorithm approach for group formation in collaborative learning considering multiple student characteristics. Computers & Education 58, 560–569 (2012) [RP6] Cavanaugh, R., Ellis, M., Layton, R., Ardis, M.: Automating the Process of Assigning Students to Cooperative-Learning Teams. In: Proceedings of the American Society for Engineering Education Annual Conference & Exposition (2004), http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.65.682 [RP7] Christodoulopoulos, C.E., Papanikolaou, K.A.: A Group Formation Tool in a ELearning Context. In: 19th IEEE International Conference on Tools with Artificial Intelligence, ICTAI, pp. 117–123 (2007) [RP8] Cocea, M., Magoulas, G.D.: User behaviour-driven group formation through casebased reasoning and clustering. Expert Systems with Applications 39, 8756–8768 (2012) [RP9] Craig, M., Horton, D., Pitt, F.: Forming reasonably optimal groups (FROG). In: Proceedings of the 16th ACM International Conference on Supporting Group Work (GROUP 2010), pp. 141–150 (2010) [RP10] Daradoumis, T., Xhafa, F., Marques, J.M.: A methodological framework for projectbased collaborative learning in a networked environment. International Journal of Continuing Engineering Education and Lifelong Learning 12(5/6), 389–402 (2002) [RP11] Filho, J.A.B.L., Quarto, C.C., França, R.M.: Clustering Algorithm for the Socioaffective Groups Formation in Aid of Computer Supported Collaborative Learning. In: Collaborative Systems II - Simposio Brasileiro de Sistemas Colaborativos, pp. 24–27 (2010) [RP12] Fukś, H., Raja Gabaglia Mitchell, L.H., Gerosa, M.A., de Lucena, C.J.P.: Competency Management for Group Formation on the AulaNet Learning Environment. In: Favela, J., Decouchant, D. (eds.) CRIWG 2003. LNCS, vol. 2806, pp. 183–190. Springer, Heidelberg (2003) [RP13] Gogoulou, A., Gouli, E., Boas, G., Liakou, E., Grigoriadou, M.: Forming Homogeneous, Heterogeneous and Mixed Groups of Learners. In: Proceedings of the Workshop on Personalisation in Learning Environments at Individual and Group Level, pp. 33–40 (2007) [RP14] Graf, S., Bekele, R.: Forming Heterogeneous Groups for Intelligent Collaborative Learning Systems with Ant Colony Optimization. In: Ikeda, M., Ashley, K.D., Chan, T.-W. (eds.) ITS 2006. LNCS, vol. 4053, pp. 217–226. Springer, Heidelberg (2006) [RP15] Haake, J.M., Haake, A., Schümmer, T., Bourimi, M., Landgraf, B.: End-User Controlled Group Formation and Access Rights Management in a Shared Workspace System. In: Proceedings of the ACM International Conference on Computer-Supported Collaborative Work (CSCW), pp. 554–563 (2004) [RP16] Ho, T.-F., Shyu, S.-J., Wang, F.-H., Li, C.T.-J.: Composing High-Heterogeneous and High-Interaction Groups in Collaborative Learning with Particle Swarm Optimization. In: World Congress on Computer Science and Information Engineering (CSIE), pp. 607–611 (2009)


221

[RP17] Sánchez Hórreo, V., Carro, R.M.: Studying the Impact of Personality and Group Formation on Learner Performance. In: Haake, J.M., Ochoa, S.F., Cechich, A. (eds.) CRIWG 2007. LNCS, vol. 4715, pp. 287–294. Springer, Heidelberg (2007) [RP18] Huang, Y.-M., Wu, T.-T.: A Systematic Approach for Learner Group Composition Utilizing U-Learning Portfolio. Educational Technology & Society 14, 102–117 (2011) [RP19] Inaba, A., Supnithi, T., Ikeda, M., Mizoguchi, R., Toyoda, J.: Learning goal ontology for structuring a collaborative learning group supported by learning theories. Electronics Communications in Japan, Part III 86(8), 79–90 (2003) [RP20] Isotani, S., Mizoguchi, R.: Theory-Driven Group Formation through Ontologies. In: Woolf, B.P., Aïmeur, E., Nkambou, R., Lajoie, S. (eds.) ITS 2008. LNCS, vol. 5091, pp. 646–655. Springer, Heidelberg (2008) [RP21] Isotani, S., Inaba, A., Ikeda, M., Mizoguchi, R.: An Ontology Engineering Approach to the Realization of Theory-Driven Group Formation. International Journal on Computer-Supported Collaborative Learning 4(4), 445–478 (2009) [RP22] Khandaker, N., Soh, L.-K.: A Wiki with Multiagent Tracking, Modeling, and Coalition Formation. In: Proceedings of the 22th Innovative Applications of Artificial Intelligence (IAAI), pp. 1799–1806 (2010) [RP23] Khandaker, N., Soh, L.-K.: SimCoL: A Simulation Tool for Computer-Supported Collaborative Learning. IEEE Transactions on Systems, Man, and Cybernetics, Part C 41, 533–543 (2011) [RP24] Kobbe, L., Weinberger, A., Dillenbourg, P., Harrer, A., Hämäläinen, R., Häkkinen, P., Fischer, F.: Specifying computer-supported collaboration scripts. International Journal on Computer-Supported Collaborative Learning 2, 211–224 (2007) [RP25] Kyprianidou, M., Demetriadis, S., Tsiatsos, T., Pombortsis, A.: Group Formation Based on Learning Styles: Can It Improve Students’ Teamwork? Educational Technology Research and Development 60(1), 83–110 (2012) [RP26] Layton, R.A., Loughry, M.L., Ohland, M.W., Rico, G.D.: Design and validation of a web-based system for assigning members to teams using instructor-specified criteria. Advances in Engineering Education 2(1), 1–28 (2010) [RP27] Li, Z., Zhao, X.: The Design of Web-Based Personal Collaborative Learning Sys-tem (WBPCLS) for Computer Science Courses. In: Li, F., Zhao, J., Shih, T.K., Lau, R., Li, Q., McLeod, D. (eds.) ICWL 2008. LNCS, vol. 5145, pp. 434–445. Springer, Heidelberg (2008) [RP28] Lin, Y.-T., Huang, Y.-M., Cheng, S.-C.: An automatic group composition system for composing collaborative learning groups using enhanced particle swarm optimization. Computers & Education 55(4), 1483–1493 (2010) [RP29] Liu, S., Joy, M.S., Griffiths, N.: Incorporating Learning Styles in a ComputerSupported Collaborative Learning Model. In: International Workshop on Cognitive Aspects in Intelligent and Adaptive Web-based Education Systems (2008), http://eprints.dcs.warwick.ac.uk/119/1/liu_joy_ griffiths_ciawes_08.pdf [RP30] Liu, S., Joy, M.S., Griffiths, N.: iGLS: Intelligent Grouping for Online Collaborative Learning. In: IEEE International Conference on Advanced Learning Technologies (ICALT), pp. 364–368 (2009) [RP31] Liu, S., Joy, M.S., Griffiths, N.: An Exploratory Study on Group Formation Based on Learning Styles. In: IEEE International Conference on Advanced Learning Technologies (ICALT), pp. 95–99 (2013)

222


[RP32] Mavrommatis, G., Tsimaras, D.: Forming cliques of collaborating distance learners. In: Proceedings of the 7th Conference on Applied Informatics and Communications, pp. 309–312 (2007) [RP33] Mehta, D., Kouri, T., Polycarpou, I.: Forming project groups while learning about matching and network flows in algorithms. In: Annual Conference on Innovation and Technology in Computer Science Education (ITiCSE), pp. 40–45 (2012) [RP34] Messeguer, R., Medina, E., Royo, D., Navarro, L., Juárez, J.P.: Group Prediction in Collaborative Learning. In: International Conference on Intelligent Environments, pp. 350–355 (2010) [RP35] Muehlenbrock, M.: Formation of Learning Groups by using Learner Profiles and Context Information. In: International Conference on Artificial Intelligence in Education (AIED), pp. 507–514 (2005) [RP36] Mujkanovic, A., Lowe, D., Willey, K.: Unsupervised Learning Algorithm for Adaptive Group Formation: Collaborative Learning Support in Remotely Accessible Laboratories. In: IEEE International Conference on Information Society, pp. 59–66 (2012) [RP37] Ounnas, A., Davis, H.C., Millard, D.E.: Towards Semantic Group Formation. In: IEEE International Conference on Advanced Learning Technologies (ICALT), pp. 825–827 (2007) [RP38] Ounnas, A., Davis, H.C., Millard, D.E.: A Framework for Semantic Group Formation in Education. Educational Technology and Society 12(4), 43–55 (2009) [RP39] Redmond, M.A.: A computer program to aid assignment of student project groups. In: ACM SIGCSE Conference, pp. 134–138 (2001) [RP40] Rubens, N., Vilenius, M., Okamoto, T.: Automatic Group Formation for Informal Collaborative Learning. In: Web Intelligence Workshops, pp. 231–234 (2009) [RP41] Spoelstra, H., van Rosmalen, P., van de Vrie, E., Obreza, M., Sloep, P.B.: A Team Formation and Project-based Learning Support Service for Social Learning Networks. Journal of Universal Computer Science 19, 1474–1495 (2013) [RP42] Tobar, C.M., de Freitas, R.L.: A support tool for student group definition. IEEE Frontiers in Education Conference (FIE),T3J-7,T3J-8 (2007) [RP43] Wang, D.-Y., Liu, Y.-C., Sun, C.-T.: A grouping system used to form teams full of thinking styles for highly debating. World Scientific and Engineering Academy and Society (WSEAS), pp. 725–730 (2006) [RP44] Wang, D.-Y., Lin, S.S.J., Sun, C.-T.: DIANA: A computer-supported heterogeneous grouping system for teachers to conduct successful small learning groups. Computers in Human Behavior 23(4), 1997–2010 (2007) [RP45] Wessner, M., Pfister, H.-R.: Group formation in computer-supported collaborative learning. In: ACM Conference on Supporting Groupwork, pp. 24–31 (2001) [RP46] Yannibelli, V.D., Amandi, A.: Forming well-balanced collaborative learning teams according to the roles of their members: An evolutionary approach. In: IEEE 12th International Symposium on Computational Intelligence and Informatics (CINTI), pp. 265–270 (2011) [RP47] Yannibelli, V., Amandi, A.: A Memetic Algorithm for Collaborative Learning Team Formation in the Context of Software Engineering Courses. In: Cipolla-Ficarra, F., Veltman, K., Verber, D., Cipolla-Ficarra, M., Kammüller, F. (eds.) ADNTIIC 2011. LNCS, vol. 7547, pp. 92–103. Springer, Heidelberg (2012) [RP48] Yannibelli, V.D., Amandi, A.: A deterministic crowding evolutionary algorithm to form learning teams in a collaborative learning context. Expert Systems with Applications 39, 8584–8592 (2012)