EVA: AN INTERACTIVE ENVIRONMENT FOR PERSONALIZED COLLABORATIVE LEARNING Gustavo Núñez, Leonid Sheremetov, Jesus Martinez Unidad Profesional “Adolfo López Mateos”, Zacatenco, Av. Juan de Dios Bátiz s/n, Esq. Av. Miguel Othón de Mendizabal, Col. Nueva Industrial Vallejo, CP 07738, México, D.F. Computer Science Research Center, National Technical University (CIC-IPN), Mexico
[email protected],
[email protected] ,
[email protected] Abstract This article reports on the results and future research work within the paradigm of Configurable Collaborative Distance Learning, which we have named EVA (states for Virtual Learning Spaces in Spanish). The article focuses on the description of the main concepts, architecture, software and LAN/ATM hardware platform developed for the EVA interactive environment. This environment is the result of the investigation and development of pedagogical models and information technologies to support learning activities of teams of students (planning, consulting, collaboration, etc.). ATM and SONET technologies have been used for the Experimental Laboratory of Virtual Learning based on LAN/ATM network. Key words: tele-teaching, learning environment, agents, ATM Introduction Since their conception, which dates back some 30 years, computer-supported and knowledge-based learning environments have offered significant potential for fundamentally changing the educational process [Mark & Greer, 95; Chan, 96; Youngblut, 94]. Nevertheless, despite of many expectations, real progress is slow and only few learning environments have made the difficult transition from the laboratory to the classroom. So, the development of pedagogically sound tools has been the time challenge. Recent trends in complex software systems development in general and learning environments, in particular, involve collaborative work as embodied in Computer Supported Cooperative Work (CSCW) for human-computer interaction and Distributed Artificial Intelligence (DAI) for learning. These resulted from the proliferation of networks ranging from local sites to the Internet, which stimulated a “virtual land rush” in distance education. Computer Supported Cooperative Learning (CSCL) systems and a relatively new type of systems called Course-Support Environments (CSE) have been built on the lessons learned from the past approaches such as Intelligent Tutoring Systems (ITS), Interactive Learning Environments (ILE) and Web-based distance education. Perhaps their most important significance is the ability to offer a collaborative environment that utilizes the experiences of other students, different teaching strategies, the group expertise recorded in knowledge base and the capabilities of software tools to arrive at a particular convergent solution in a group fashion. On the other hand, these systems are capable to configure and generate dynamically the learning environment in a personalized way according to the information stored in the database [Collis, 99].
The investigation project, which we have named EVA (Espacios Virtuales de Aprendizaje in Spanish - Virtual Learning Spaces), applies the methodology and tools of Distance Learning (DL), CSCW and ITS to obtain a new paradigm of the Configurable Collaborative Learning [Núñez et al., 98]. This project is dedicated to the research and development of pedagogic models and information technologies that provide spaces of knowledge, collaboration, consulting, and experimentation for supporting the learning activities of teams separated geographically who maintain common conversations, matters, and projects. The article focuses on the description of the architecture and technologies developed for the EVA interactive environment. EVA concepts EVA framework is based on the integration of artificial intelligence techniques, agent technology and interactivity in the virtual learning environment. It differs from the ITS and ILE paradigms, because, on the one hand, it does not pretend to substitute a human tutor by an intelligent software system with one-on-one tutoring model, and, on the other, doesn't permit students to investigate and learn topics absolutely free of external control. EVA framework is inspired in the works on the social-constructivist and shared cognition theories, making emphasis on the role of interactions and collaborations within the learning environment, giving a student all the necessary help, orientation and resources [Dillenbourg et al., 97]. The conceptual architecture of EVA is structured into the four essential learning elements, called Virtual Learning Spaces. These spaces are: knowledge - all the necessary information to learn, collaboration - real and virtual companions that get together to learn, consultation - the teachers or assessors (also real and virtual), who give the right direction for learning and consult doubts, and experimentation - the practical work of the students in virtual environment to obtain practical knowledge and abilities. The others main concepts of EVA are the following: 1. The "Multi-book" concept, a personalized electronic book, generated by concatenating selected units of learning material (ULM) along the learning trajectory for each knowledge domain. 2. A multilevel knowledge representation model of the problem domain based on concept graph formalism. 3. The intensive use of agent-based tools • as virtual tutors, virtual students or learning companions, virtual personal assistants that help students to learn, • for mining information, managing, planning and scheduling learning activities, information retrieval and filtering, student assistance and tracking for evaluation of their intentions and performance, and • to organize the workgroups and groups activities, and re-configure their work and knowledge spaces. All the applications are developed for the domain of the Master and Doctorate education in Computer Science. To represent the knowledge taxonomy we have proposed a model based on the hierarchy of knowledge domains and concept graph representation of knowledge and learning activities. Knowledge domains have a hierarchic structure that corresponds to different areas and levels of abstraction: common domain, specialty
domain, sub-specialty domain, knowledge units domain, etc. For example, common domain consists of Object Oriented Programming (OOP), Data Base Design (DBD), Discrete Mathematics (DM) courses, etc., domain of AI consists of AI, Logic Programming (LP), Mathematical Logic (ML) courses, etc., and contains several subdomains: Knowledge Based Systems (AI1), Automatic Learning(AI2), Natural Language(AI3), Vision and Robotics (AI4). Domains are connected by prerequisite relations (R1-R6). Usually this relation has hierarchic character, but sometimes, one domain is related with sub-domains of different specialty, which is also reflected at the level of knowledge units, e.g. the case of Distributed Systems (DS) and Distributed Intelligent Systems (DIS), related by R1. The model of learning knowledge for each domain is represented in the form of a concept graph (fig. 1). This model is used for learning trajectory planning, virtual space configuration and student tracking. Insert Fig. 1 here.
EVA Software Environment The software development in EVA is based mainly in the technologies and tools of AI, software agents, and intensive use of software which permits to create virtual environments of learning and collaborative work for students and tutors communicating via Internet (Fig. 2) [McCormack & Jones, 98]. Insert Fig. 2 here. This software is used to develop the key EVA systems of automatic configurable distance learning environment: • Virtual Classroom that permits to assist a traditional classroom from the Web site. It uses teleconferences, multimedia materials and the tools of interactive synchronous communication. • Virtual Didactic Support that permits the students to have an access to all the materials as multimedia-based virtual courses and Multibooks, the CIC library and virtual libraries, to have interactions with the students with the same specialization area and level of knowledge. • Virtual Evaluation that permits to evaluate the students the same way as in the conventional classroom practice, considering classroom assistance, participation in discussions, tasks and projects, and virtual exams. • Virtual Experimentation that permits to conduct experiments in Virtual Laboratories. EVA'S architecture is based on four servers (Java Web Server, Apache Web Server, Apache Web Server and Video and Sound Server), and a number of applications, implemented as servlets, applets and agents that offer the services to the final users. Usually, a perfect management of huge collections of web pages is a difficult problem that can be solved by storing the information about each link separately from the page itself. One of the solutions is to use Hyperwave Information Server having facilities to automatically linking the pages from page collections. To customize the learning spaces for each user, we have adopted the technology of dynamic page and link generation using servlets. Java Web Server is used to mount the servlets (they allow dynamic pages
generation), which constitute the concept of virtual classroom and virtual laboratory; since the servlets is the mechanism to communicate with the database manager; which manipulates all the information that is in the database. The same database is used to link student’s annotations and external links to the predefined context of the ULMs. Apache Web Server 1 offers the interaction services and collaboration to the users, through its discussion forums and of its chats. Video and Audio Server contains the material multimedia and also help distance communication. Agent Name Server contains a list of all the agents that they are working in EVA, so that the other agents can communicate among them, without caring the place where they are. These agents have as end to evaluate and modify study plan the one on the way to the users. Agent technology is the promising way to approach the problem of VLE development. The notion of agents is the central part of contemporary learning environments, where they act as virtual tutors, virtual students or learning companions, virtual personal assistants that help students to learn, mine information, manage and schedule their learning activities [Müller et al., 97; Barros & Perkusich 96; Gordon & Hall 98]. One of the main purposes of our project is to develop models, architectures and multi-agent environment for collaborative learning and experimentation. Agents in EVA The core of the EVA environment consist of a number of components, composed of a set of deliberative and auxiliary agents, forming three multiagent systems (MAS): virtual learning community, composed of learning companions and personal learning assistants, Multiagent planning system and Multiagent Virtual Laboratories. Our prototype of learning community incorporates a monitor agent and a learning companions agent. Agents have mental states represented in terms of beliefs, knowledge, commitments, with their behaviour specified by rules. These agents have been implemented as JATLite agents with rule-based inferencing capabilities, programmed in Jess (JATLite 1998, Friedman-Hill, 1998). Each group has an associated Monitor Agent. He maintains the shared knowledge model of a group and compares it with a group problem model from the knowledge base that contains the objectives, concepts, activities, etc. that characterize a group. His behavior is guided by a set of domain independent conversation rules, which refers to the interactions between the group members. During the group session, the monitor agent maintains the current problem state (shared group knowledge space) and the history record of all contributions for each participant. The problem state in terms of shared beliefs and knowledge is used to change the interaction mode, choosing one of monitoring techniques from his rule base. This result in changing behavior of learning companion from group leader (strong companion) through a week companion to a passive observer. Companion Agents can differ in expertise level and role strategy, using 3 different pedagogical styles from supportive to non-supportive ones that can change their role in the group. Strong companions can adsorb the role of a group leaders, developing and
explaining problem solution, while the weak one is used to enhance student's role in a group and stimulate his learning abilities by enforcing him to teach his companion. The conversational categories are used to classify interactions and generate learning companion's messages and are mapped into a set of communicative actions, expressed in Knowledge Query and manipulation Language (KQML) (Finin, 1997). Emotions are simply visualized by means of companion's appearance. In his individual learning activities, students also receive support from assistant agents. Internet search and filtering agent: automatically searches for additional teaching material on the Internet, not yet located in the Knowledge Space, according to student's learning trajectory and through natural language searching tools (such as Clasitex [Guzmán, 98]). Synthesizes information compiled of several sources about the same topic and filter redundant and repetitive information. Performs connection functions with Newsgroup NNTP servers, Web servers, etc; and collaborates with other agents through the message passing mechanism in agent communication language (currently, KQML). Collaboration agent: compares student's academic development and profile to form collaboration groups, requests help or information from other assistants and make collaborative decisions on how to select, integrate, and order information to be shared between the tutor and students or inside the students group. Handles student and group models. Personal advisor: suggests a personalized study plan to student according to its academic formation, interests, abilities and advances, and modify study plan, if it is necessary. Selects, integrates, and orders the information to study; advices for the learning problem solving, tells where to find material (notes, books, magazine references, similar topics, etc) and tests. Helps the student to choose sources, suggests topics of thesis or projects, indicates to the student virtual labs, where it can make practices. This agent communicates with the search agent to find information according to the topic, it communicates with collaboration agent to collaborate with the rest of the group members. Handles the domain model. Evaluator agent: verifies periodically the learning advances, tries to find causes of misunderstanding, communicates with the advisor agent to re-organize the information to study. Handles the study model. Personal assistant: classifies student´s e-mails, reminds him about events, conferences, videoconferences, courses, tasks, tests, etc. Injector of agents: places demons in remote places, using the available network. It works under a variety of network types, communication types, protocols, and it supposes a hostile or disordered environment in the host, therefore needing to send several agents to become familiar with the host's environment [Martínez, 98]. The next MAS plans students learning trajectories. A learner navigates the virtual learning spaces by routes (study plans) suggested in an automatic manner by EVA. So, the purpose of the planning system is to design a particular learning trajectory for each student in the knowledge space, formed by ULM, which is organized in knowledge domains. At the next stage, personalized books, called Multibooks, are armed by
concatenating of selected ULM along the learning trajectory for each knowledge domain. In the same way, groups of students with similar interests are arranged. Prerequisite relations are used to connect study plans, developed for each domain. Finally, student learning activities must be scheduled to satisfy temporal constraints. To perform planning and scheduling within the MAS paradigm, we have proposed to associate a planning agent with each domain model. These agents have goal and belief representation capabilities. Since local goals and decisions are interconnected, developed multistage negotiation algorithm provides means by which an agent can acquire enough knowledge to reason about the impact of his local decisions and modify its behavior accordingly to construct a globally consistent decision [Sheremetov & Nuñez, 1999]. A special coordinator agent facilitates communication and performs managing activities. Finally, agents in EVA form the substantial part of virtual experimentation space. Even though the virtual learning environments has been studied from a pedagogical perspective in the last years, few researchers has formulated clear methodologies how to create Virtual Laboratories for to cooperation in the WWW [Das, 94; Rodríguez et al., 98; Forbus & Kenneth, 96], even less based on Virtual Reality Modeling Language (VRML) scenarios [Lemays et al 96]. The EVA project contemplate in an integrated way, the following key elements, which have not been yet considered all together: • Tele-presence, including the monitoring and smart remote control of systems (telemonitoring and tele-control). • Computer aided experimentation (real and virtual). • Assistance in performing the experiments that enables configuration, performing and further explanation of results, as well as their group discussion with experts. • Distributed and collaborative work environment, based on virtual reality interfaces. Hence we propose the name "Virtual Space of Cooperative Experimentation", as a successor o extension of a Virtual Laboratory concept. We have developed several applications for the experimentation space of "Distributed Intelligent Systems" course, which allow the students to develop their agents, execute them in the multi-agent environment, and communicate with them through KQML. Examples include virtual bank system, virtual travel agency, and virtual manufacturing system. The visualization of agents behavior is performed in VRML allowing to generate 3D representations of scenes through a browser. The VRML allows to change the object and agents features of one scene (shape, color, position, size, orientation, etc) in function of determined events through the script languages, in other words to assign constraint behaviors [Mitra et al 95; Nadeau & Moreland 95]. We've developed a CASE tool called EasyVRML that permits to assign more complex behaviors to the VRML scenarios in a more simple form then conventional tools [Quintero et al., 98]. EVA platform Developments like those proposed in this paper would not be possible without nowadays achievements of computer networking and telecommunications. So, we
explore the advantages of the key communication technologies like Asynchronous Transfer Mode (ATM) and SONET (Synchronous Optical Network) as a transmission standard [Martinez et al., 98]. Then, to be able to test the EVA environment, the LAN/ATM based Experimental Distance Learning Laboratory (EDLL) was installed (fig. 3). The core of the system is an ATM switch with 8 gates at 155 Mbps, with the possibility to extend the capacity up to 622 Mbps, connected to the IPN ATM backbone and to an ATM LAN switch with 24 Ethernet inputs (10 Megabits each). Insert Fig. 3 here. The EDLL consists of two parts: the EVA development laboratory and experimental learning laboratory. The first one contains the equipment for servers enabling EVA environment, high performance PCs for the software development (PentiumII, 400Mhz and double processor PCs) and the equipment for ATM administration. The learning Laboratory is composed of a number camera-equipped PC with a 10BaseT Ethernet interface, one PC is hooked up to a TV set and/or a LCD projector, an ATM PC is connected to the electronic blackboard, supporting BrightLight server software. This Laboratory is the key element for the development of the ATM Experimental Center of Information Exchange, where the following on-going projects are being developed: • Development, implementation and testing of teaching/learning activities in EVA environment • Testing of performance of multimedia applications via assignment of virtual channels and bandwidth of the LAN/ATM network • Transmission of signals and images over ATM using different compression techniques. The EVA Internet front-end (versión 1.0) with the above mentioned functionality is installed on WEB UNIX and WindowsNT servers Apache, Personal Web Server and Java Web Server. User interface in the knowledge space is depicted in fig. 4. Prototypes of agents have been developed using Microsoft VC++, LALO, JAVA, and JATLite for Unix and Windows platforms. Personal learning assistants with information filtering capabilities and agents for individual and collaborative learning with artificial learning companions are under development. Insert Fig. 4 here.
Conclusions and future work The main characteristics of the EVA learning environment are the following: • It considers in an integrated manner all the methods and channels, by means of which students acquire information and communicate to learn, to exchange knowledge, to solve problems, to experiment and, in general, to realize academic activities. • It also contemplates the development of the environment capable of delivering to the user in an integrated and virtual way the main services, work and entertainment spaces usually offered by the institutions of high education, including the interactions outside of the classroom. • The environment is dynamically configured, generated and delivered through the user’s WEB browser based on the models stored in the database.
• This environment makes intensive use of agent technology, where agents plan, configure, evaluate, help and administer the learning process. • The knowledge is represented in a multimodal way, i.e. logic, linguistic, graphic and/or animated, audio, to facilitate it’s communication and assimilation by students with different learning styles. • The technological platform is developed based on the recent results on the high speed communications to efficiently realize the integration of data, audio y video. We argue, that that technological infrastructure, which will be necessary to enable telelearning applications in the first quarter of the next century, depends more and more upon the fact of having in disposal the high bandwidth networks, so we try to make intensive use of the ATM technology. We also consider as the future work the integration of the ATM Experimental Center of Information Exchange into the Internet2 Mexican initiative, which is now at the planning stage and will be operative in 2001. A number of Multibooks have been developed: Languages, Automata and Complexity, Mathematical Logic, Artificial Intelligence, Intelligent Distributed Systems, Virtual Reality Modeling Language, Topological Design of X.25 Networks, to mention just a few ones. The work on virtual learning community, composed of learning companions and personal learning assistants is in process. We are also working on animation assignation to learning companions and tutors (especially in the experimentation space) to obtain their strong visual presence in a 3D learning environment. We have developed a number of prototypes but a lot of work is still to be done to convert them into a real VLE. At the current stage of the experiment we are also developing the course on "The Potential of Virtual Reality in the Teaching of Engineering" to be delivered at the University level for about 200 teachers of engineering. We are convinced that a strict evaluation of the improvements brought by the mentioned techniques is necessary. Acknowledgments Support for this work has been provided by the CONACyT within the REDII Program and IPN, Mexico. The EVA project is the effort of many people. We would particularly like to thank A. Guzman and R. Menchaca for their support of the project, as well as our colleagues for their determination and contribution to the system development, especially J. Su. We also thank the students of Doctorate and Master Programs for contributing ideas and talent to EVA programming, especially R. Peredo, L. Balladares, R. Quintero, A. Armenta and J. L. Rodriguez. References Barros Costa, E., & Perkusich, A.(1996). Modeling the Cooperative Interaction in a Teaching/Learning Situation. In Intelligent Tutoring Systems: 168-176. Chan, T.W. (1996). Learning companion Systems, Social Learning Systems, and Intelligent Virtual Classroom, In Proc. of the World Congress on AI in Education, 1996, Vol. 7, No. 2, 125-159. Collis, B. (1999), Applications of Computer Communications in Education: An Overview. IEEE Communications Magazine, March 1999, 82-86.
Das H. (1994) Telemanipulator and telepresence technologies, SPIE 2351. Dillenbourg, P., Jermann, P., Schneider, D., Traum, D. & Buiu, C. (1997), The design of MOO Agents: Implementations from an Empirical CSCW Study, In B. Du Boulay and R. Mizoguchi (Eds.) Artificial Intelligence in Education, IOS Press, 15-22. Finin, T., Labrou, Ya., & Mayfield J. (1997) KQML as an agent communication language, in J. Bradshaw, ed., Software Agents, AAAI/MIT Press, 1997. Forbus, D. & Kenneth, (1996). Articulate Virtual Laboratories for Science and Engineering Education; http://www.qrg.ils.nwu.edu/projects/NSF/avl.htm Friedman-Hill, E. J. (1998), Jess, The Java Expert System Shell, Version 4.3, Technical report SAND98-8206 Sandia National Laboratories, Livermore, CA, http://herzberg.ca.sandia.gov/jess. Gordon, A. & Hall, L. (1998).Collaboration with Agents in a Virtual World. In Proc. of the Workshop on Current Trends and Artificial Intelligence in Education, 4 World Congress on Expert Systems, Mexico, 1998. Guzmán A. (1998). Finding the main themes in a Spanish document. Journal Expert Systems with Applications, Vol. 14, No. 1/2, Jan/Feb. 139-148. JATLite Beta Complete Documentation (1998), Stanford University, http://java.stanford.edu Lemays, L., Murdock, K., & Couch J. (1996). 3D Graphics and VRML 2, Sams Net. Martínez G.(1998) Automatic installer of systems: mobile agents with blackboard architecture. Master of Science thesis, Electrical Engineering Department (Computer Section), Centro de Investigación y Estudios Avanzados del I.P.N. Mexico City. (In Spanish). Martínez, J. Su, J., Cano, C. & Arredondo, B. (1998), Análisis y Diseño de Redes de Conmutación de Celdas. En Conferencia Internacional del CIC, México. Revista Instalaciones, N 18/19. McCormack, C. & Jones, D. (1998), Building a Web-Based Education System, Wiley. Müller, J., Wooldridge, M. J. and Jennings, N. R., eds.(1997), Intelligent Agents III: Proc. of the 3rd Int. Workshop on Agent Theories, Architectures, and Languages, Lecture Notes in Artificial Intelligence, Vol. 1193. Nuñez, G., Sheremetov, L., Martínez, J., Guzmán, A., & Albornoz, A. (1998). The Eva Teleteaching Project - The Concept And The First Experience In The Development Of Virtual Learning Spaces. In Proceedings of the 15th IFIP World Computer Congress, Teleteaching’98, Part II, 769-778. Mark, M.A., and Greer, J.E. (1995), The VCR tutor: Effective Instruction for Device Operation. Journal of the Learning Sciences, 4(2):209-246. Mitra, Yasuaki Honda, Kouichi Matsuda, Gavin Bell, Chee Yu, & Chris Marrin (1995). Moving Worlds: behaviors for VRML, http://www.vrml.sgi.com/spec.html Nadeau R. D., Moreland L. J., (1995). The Virtual Reality Behavior system (VRBS): A Behavior Language Protocol for VRML, San Diego SuperComputer Center. http://www.sdsc.edu/projects/vrml/vrbs_proto.html. Quintero, R., Armenta, A., Balladares, L. & Nuñez G. (1998) EasyVRML Herramienta CASE para la asignación de comportamiento a componentes de mundo VRML. XI Congreso Nacional ANIEI, Mexico, 1998 (in Spanish) Rodrigues de Queiroz, L. et al. (1998). A Robotics and Computer Vision Virtual Laboratory, IEEE´98 In Proceedings of AMC´98, Coimbra, June29-July 1, 694-699. Sheremetov, L. & Núñez, G. (1999). Multi-Stage Cooperation Algorithm and Tools for AgentBased Planning and Scheduling in Virtual Learning Environment, The 1st International Workshop of Central and Eastern Europe on Multi-agent Systems (CEEMAS'99), 30th May-3rd June, St. Petersburg, Russia. (to appear) Youngblut, Ch. (1994) Government - Sponsored Research and Development Efforts in the Area of Intelligent Tutoring Systems. Institute for Defense Analyses, USA.
