University of Granada

(Abstracto árabe de la Tesis - ملخص رسالة الدكتوراة)

Computer Engineering School

Software Engineering Department

(Languages and Information Systems)

FLEXIBLE LEARNING MANAGEMENT SYSTEM

TO SUPPORT LEARNING IN

THE TRADITIONAL AND OPEN UNIVERSITIES

Submitted by

Jamil A.S. Itmazi

Supervised by

Dr. Miguel Gea Megías

                                                                                IN Granada, September  2005

ABSTRACT

This work gives an introduction about E-learning and its: Challenges, requirements and prospective on educational organizations (universities). Explain the two Models of E-learning (Synchronous and Asynchronous). Different modes of learning and education are presented, focusing on the Open and traditional Learning.

Thus, we also give background about IMS, LMS, and knowledge, their definitions and their relationships with E-learning.

Our research focuses on a proposal for asynchronous E-learning called E-LMIS (E-learning Management Integrated System) to support of different modes of learning in universities.

We review the state of art; preview 28 famous LMS, present results of the analysis of ‘Comparison and evaluation Studies’.

We choose three packages; giving the reasons of this chose, studying and analysis those packages.

The studying and analyzing is covering some general information, Introduction, the package feathers then the weakness and evaluation.

The aims of this investigation are studying one famous LMS of OSS, analysis the weakness of this software especially the important KM and Cooperative learning feathers that does not exist in this package.

Our system (E-LMIS) is OS-LMS with some important knowledge-based and Cooperative learning feathers to present E-learning Management Integrated System rather than LMS due to using tools much more from E-learning solutions than LMS.

We present this new System “E-LMIS” which will be design especially to fulfill the requirements of open and traditional learning way by using E-learning tools in asynchronous way, and we give more detailed information about the tools of the system, which support the normal and open learning.

Finally, after defining basic tools and services, analyzing their features, the E-LMIS architecture is proposed.

Preface

This investigation covers the results in the field of E-Learning. Focusing in OS-LMS to present E-learning Management Integrated System (E-LMIS), which integrate some important knowledge-based and Cooperative learning feathers to one of OS-LMS´s.

 This work consists of:

Chapter 1 (Chapter I - introduction of E-learning and universities issues), consists of three parts:

PART I - starts with a general introduction about the field of E-Learning.

PART II - starts with a introduction of E-Learning and its Challenges, requirements, prospective and E-Learning Models (Synchronous and Asynchronous). Then it conducts a brief critical analysis of the advantages and disadvantages of E-Learning.

In addition, it contains a background about E-Learning relations and domains; IMS, KM, LMS, Cooperative learning, also it contain some information about E-learning and OSS.

PART III - starts with a general introduction about universities and “modes of learning”- then explain the kinds of universities: traditional, open and Virtual universities, and explain the modes of learning: traditional, Distance and Virtual learning, lighting the differences between those sectors.

The next chapter is Chapter 2 (The State Of the Art), it contains:

This chapter starts with a set of LMS´s (CMS or Course Website Software), listing their sources and sites, and mentioning their definition.

Then it present analysis of all Comparison and evaluation Studies of LMS´s, the chapter continues with explains of the reasons, which led to choose those software: Moodle, Ilias and BSCW, to study and analysis, end show general Features comparison between them.

Later the study and analysis of those packages start with general information about every package, and then list its Features and his weakness and evaluation.

The next chapter is Chapter 3 (The Results of the NECESSITY TOOLS) , it contains:

 Chapter 3 contains the results of the necessary feathers of KM and cooperative learning.

The next chapter is Chapter 4 (The Recommendation Systems –RS) , it contains:

The first part of the chapter is an introduction about recommendation systems, within the following elements:

Background. Somehow RS is a recent system that it spread within the internet, mainly to overcome the "Information Overload" problems, but it is not new in the real life, it is habitual behavior. RS needs; Background data, input data and an algorithm to be implement.

Identification of RM. RS can provide the user a set of candidate items that may be useful or preferable to the individual user, from a large amount of items. This process could be automatically or upon user request.

We declared three related concepts close to RS; about the IRS; the majority of RS activities are especial part of IRS techniques which differ form other IRS techniques. The personalization include how to find and filter the information that fits the user preferences and needs, how to represent it and how to give the user tools to reconfiguration the systems, in consequence, RS could be part of personalized environment in some systems. The user modeling is the process of constructing (often computer based) users models, while the User Model means all the information collected about a user that logs to a web site, in order to take into account her needs, wishes, and interests. In eLearning, every LMS has its techniques to modeling his users so as to construct the user model or profile.

Objectives of RS. Mainly, to Overcome the problem of “Information Overload", but also Present information that fits the users  preferences, Facilitate organizations better managing their customer relationships, Protection and hide some material and give Feedbacks. Then five goals of e-commerce RS have been presented.

History of RS. After the Tapestry founded, a Varity techniques and technologies of RS have been produced and introduced. A Table within Examples of RS research has been presented.

The second part of the chapter is about the Current Usage of RS, some example of the actual uses of RS has been mentions within some main usage sectors. Those sectors include; E-commerce, Web Pages, News, Digital library, Censorship Systems and Other sectors that include E-learning RS.

The third part is Types of Recommender. RS consist of many approaches, every type of those methods has its advantages/disadvantages, and some of them are suitable to some domain while others are suitable to other domains. The RS approaches are content-based system, collaborative filtering, hybrid RS, economic filtering system, demographic-based system, rule-based filtering, knowledge-based systems.

We studied every one of those approaches in this mode; giving Background about it, presenting some Examples systems that used tit and finally giving its shortcomings.

The forth part is about the ability to use RS in LMS of eLearning. We have been declared that we could use it in LMS, because the reasons and motivations of using RS on other sectors are present in LMS.

Also, this part is covering the domains of LMS which could be using RS in university environment. This sectors or domains of RS in LMS which used in university environment are the following: Recommend Learning objects, Recommend Learning Subjects and Recommend teacher to student or teacher to course.

The next chapter is Chapter 5 (analysis and design the RS), contains:

 The first part of the chapter is an introduction about the implementation of RS in LMS within the following elements:

Background. It was mentioned the ability to implement RS in LMS and declared that the domain of the proposal of using RS in LMS will be the Learning objects recommendations

Related characteristics. It was cleared some LMS characteristics that can effect the implementing of RS. E.g. the log-in of registered users, manage courses, track learner activities, results and management reports. Also may be include additional functions such as: Authoring of content, Management of classroom training, instructors and resources…Learner collaboration tools (chat, discussion groups, etc.)

Consideration Factors. The user tastes and preferences while change in the RS in a LMS to the useful, interest and important objects to the students. Also, Every object Learning is a machine parsable form, or has attributes that assigned to them manually.

The second part of the chapter is about The suitability of RS approaches. This part was discussed every RS approaches, from some view points: the ability to use it to recommend learning objects in a LMS, the using method: alone or hybrid with others, the implementation methods and finally discussing the main shortcomings under the RS in a LMS.

Here we abstract the suitability of every approach:

·         The CBS can be used alone or as a primer approach with others.

·         The CFS could be used in LMS, alone or hybrid with others, but put practically; it is advice to use it with others as a part of HRS.

·         The EFS is not a preferable system to be using in a LMS.

·         The DBS can be used as a complementary approach in a LMS, but not alone.

·         RBF can be used as a complementary approach in a LMS, but not alone.

·         The KBRS can be used alone or as a primer approach with others.

The third part is study about the suitable approach. In this part two proposals have been presented, a very general structure of RS which useful with could be implemented in any LMS; while the other is a HRS algorithm which could be used in a LMS upon our consideration. Every proposal has the necessities explanations.

The system will get the recommendations automatically without the user request, e.g. when the student login a course.

The forth part is about how to integrate the HRS algorithm to the Moodle LMS. It give a background of the ingeneration process and give a declaration and modifications about some concepts and actions from view the HRS ingeneration to Moodle LMS.

It discuss and explanation the way to implement the selected approaches of our RHS algorithm in the Moodle. This way of such approach is a “stage”. The discussed stages were:

·         The Stage of Content-Based System.

·         The Stage of teacher recommendations.

·         The Stage of Collaborative Filtering.

·         The Stage of Demographic-Based Filtering.

·         The Stage of Rule-Based Filtering.

Every stage: has the explanation of its approach and the using method, beside list and discuss the related fields in Moodle, addition to the listing steps of the Sub-algorithm of that approach and finally a flowchart structure of this Sub-algorithm.

The next chapter is Chapter 6 (implementation of RS in LMS), it contains:

 It contains the programming steps, the implementation, the execution and finally the test.

CONCLUSIONS

The learning system is the soul of the human society، that always is looking for new tools and methods of education to adopt them to its system and to improve the learning process. One of those tools more outposts is the electronic learning.

Generally، Internet is new means that have extended and popularized in the years 90، in which it emphasizes eLearning like a very recent and important tool. At the present time there are many educative centers (especially in the universities، great companies and organisms) that foment eLearning like means of flexible education and at a distance.

The first definitions use the term for the learning based on computer (CBT: Computer Based Training) and characterizes by the technique (multimedia) and the used support (Web or CD-ROM). Nevertheless، at the moment eLearning is conceived as the acquisition and use of knowledge in form of courses and objects for the learning distributed and obtained through electronic means.

eLearning is a concept that arose in the last decade of the century last، in the educative scope، like a new method for education، using didactic supports based on the new technologies، and that some are considered like an electronic version of education to distance.

Simply، we can consider eLearning in our Thesis like an umbrella that covers the learning almost at any time، anywhere، (asynchronous) in a computer، connected generally with a network.

At the moment، eLearning has become something fundamental at our time, especially for the universities, since with eLearning the education has a strategic importance in our society, especially for the universities in two slopes:

- Organization of the own University changing of traditional to virtual or Opened models

- Facilities in the process of learning of the student (pursuit, customized aid, adjustment to the schedule, etc.)

In the future the society it will use eLearning of natural form, and will stop being something specific and separated of the traditional education. ELearning wants to replace all the present system of learning, but to be a part of the same one.

In order to be successful, eLearning needs many requirements and conditions, some are requisite technicians and others are not technicians, for example: requirements of: organization, humans (expert, professors, and students), etc...

In the reality, eLearning is an interactive sector and an intersection with other sectors, for example IM, km, LMS/CMS, cooperative learning and has two ways: synchronous and asynchronous, each way has its advantages and disadvantages. The universities almost need the asynchronous way more than the synchronous one.

In fact, the complexity of the management of the information as much of the courses as of the students has demanded specific systems for the management of eLearning, denominated LMSs/CWSs (Learning/Course Management Systems) with commercial products (WebCT, Lotus, BlackBoard, etc...) and of license GPL (Moodle, Illias, etc...). Nevertheless, improvements are due to develop that foment the colaborativo learning (CSCL), the management of the knowledge and its semantic representation, the adaptation to the user, etc...

The LMSs/CMSs is the heart of the solutions of eLearning, are software that automate the administration of the events of the education. There are two concepts: vestibule of education and LCMS are very similar to the LMS concept, but they have some differences.

In general, all the LMSs they manage the connection of registered users, catalogues of the course, they follow activities of nascent and the results, and provide intelligence data for the administration, whereas some LMSs can include additional functions like for example: responsibility of contents, direction of education of classroom, instructors and tools of collaboration to the student (to char it, groups of discussion, etc...).

At the present time, there is a great amount of software that is used as it leaves from eLearning (there are almost 200 LMS), but most necessary they are studies than they allow to improve those systems, specially the based ones on OSS to favor its implantation in universities. Nevertheless, great part of the studies compared and evaluated between LMSs is only centered in analyzing services available, but it does not analyze the behavior and functionality.

After this analysis of LMSs/CMSs one has identified those platforms OSS that can be used candidates to their evaluation and like support to the development, since the use of platforms is considered more advisable that have a development community that supports it (to create a new one). Once identified, we have analyzed those tools (based on the management of the knowledge and colaborativo work) that would be necessary to give an effective support elearning to present/display a flexible LMS.

Previously, the objectives of the three evaluation of LMSs are leaving from a robust LMS based on Open Source to make an analysis of their characteristics and proposing the incorporation of characteristics that we considered specially important like they are the management of the knowledge and the colaborativo learning, that is as a rule not contemplated in these products. The proposal of a flexible system of management of eLearning is directed to incorporate these mechanisms on a platform OSS.

We presented/displayed this system that will be designed specially to satisfy the requirements with form of the open and traditional learning using the tools of eLearning in forms asynchronous. We presented/displayed detailed information more on the tools of the system, that support open and traditional learning.

Moodle is the LMS of OSS that reunites a series of interesting characteristics and flexible has been chosen to present/display ' a LMS that it has in addition to the MOODLE characteristics, some interesting tools, that is to say, the flexible system of management of eLearning is MOODLE Os-LMSs with (some characteristics of km and some tools of the cooperative learning).

These tools are:

o        Systems of recommendation.

o        Generation of structure of the knowledge.

o        Tools of standardization.

o        Services of groupware.

o        The tools of markers.

o        Private email (and mail).

o        Share / Reuse the content.

o        The construction of community.

o        Shared spaces.

o        Pursuit of the groups.

o        Tools of personalization.

o        System of evaluation.

From all these tools, we have taken the systems of recommendation for its exhaustive analysis, explaining its characteristics within a system of eLearning and method of implantation. Later, it has been made a design and implementation of a module in Moodle to give support to RS.

RS recommends the most appropriate content to the students; the objects of learning can be enumerated nearer available which describes the instructor like the content of the module. They provide to a user a list of points (articles) candidates who can be to them useful or preferable to the user, of between a great amount of articles.

Generally, RS, has relation with the personalization concept, that tries to also orient the content Web and the applications to specific users and, have relation with UM (the user modeled one, User Modeling), which is a process in an interactive system that helps to construct a user model-profile; in order to have access to this profile adding, updating and suppressing its entrances; and to provide to other parts of the system about the user.

Relatively, RS is a new technique for computer and Internet. First RS is "Tapestry" that appeared in 1992. We can summarize the objects of RS like: to solve problems of information overload, to present/display information that is interesting to the users, to hide some types of materials and to give information to the directors to heighten and to improve its knowledge about the users.

At the moment, RS is used in many sectors: and-I deal, pages Web, the news (and messages of the email), digital library, systems of censorship, etc... At the moment, there are types of the recomendadores (methods, approaches), that are used in those sectors, most famous they are: based on content colaborativo filtrate system of recomendador economic filtrate system knowledge based system, hybrid systems, system demographic system based filtrate system based on rules and system based on utility.

Theoretically, RS could be used in LMS because the reasons and the motivations to use RS in other sectors are present in LMS. Hay some investigators mention the capacities and the necessities to use RS in systems of eLearning in general and LMS in individual.

The sectors/dominions where RS could be used in LMS in a university atmosphere are: to recommend educative objects, to recommend educative subjects, to recommend professor to the student or professor to the course and to recommend books, resources of the library in line.

This Thesis studies and investigates the first sector; (to recommend educative objects: courses /materials /contents) to implement it like RS in a LMS that is used in university atmosphere that has registered students.

The truth, later to study and to investigate the plus convenience of each method-approach to make RS: to recommend educative objects, indicates that the appropriate method-approach but is the hybrid system of recomendador, which means that the appropriate method is mixture of these method-approaches.

Therefore, in this Thesis, we decided to use the system based on content like main approach, next to certain approaches like complementary: colaborativo filtrate system filtrate system, demographic system based on rules.

In this Thesis, we presented/displayed a new algorithm with the purpose of making RS in a LMS to recommend courses to students, where the system automatically prepares the recommendations without the request of the user when the student accedes to the system to a course; the chosen LMS to implement RS is Moodle OSS.

Later the presentation of this algorithm; we studied how to integrate the algorithm of HRS to MOODLE. One occurs to an antecedent of the integration process and a explanation-modification on some concepts and actions of the vision of the HRS. In detailed way, we discussed and we explained the form to implement the selected approaches of this algorithm in the MOODLE.

To study this algorithm, it splits to a few stages, in such a way that every stage, there is had the explanation of his approach and of the method that it uses, next to the list and discussion of the fields related in MOODLE, in addition to the list of steps of the Sub-algorithm of this approach and finally of a structure of flowchart of this sub-algorithm.

The controversial stages are: the stage of the system based on content, the stage of the recommendations of the teacher, the stage of the system of leaked colaborativo, the stage of the system based in demographic and the stage of the system of leaked based on rules.

To realize the stage of the system of leaked colaborativo, it is necessary to add central system of evaluation (voting or grade), which is allowed to itself that the students evaluate or they vote to the courses. This system improves the process of RS when the courses recommended to a student appear firstly that his neighbors (his partners of the school) find these interesting courses.

Generally, we add two activities (modules) to the system Moodle: RS who might be considered to be a module of course (activity), whereas another activity is the System of evaluation that might be considered to be a fixed block that appears below the screen of the course; also, we add some necessary fields to some profiles (files) that have been detailed mentioned; the said profiles are: of course, of user and of system.

Also, we mention the programming of RS, where we have mentioned all the necessary codes of programming to implement our algorithm of RS in Moodle LMS.

All the codes of programming that we have developed or modified are listed in the appendices, where we classify those codes of programming to some appendices and sub-appendices as the relations with the previous classification.

Finally, we present the test of RS, later we add courses of samples, where some of them have been evaluated by some students, so that the manager and some students put his criteria of filtration, this way, we have a series of suitable, such results as a list of recommended courses, which appear on the screen of the course, working automatically once the user gains access.