Research Proposal for DAAD PhD scholarship Semantic Knowledge Management for E-learning

Research Proposal for DAAD PhD scholarship Semantic Knowledge Management for E-learning Darya Tarasova

1

Introduction

The Social Web is growing daily, in particular with regard to the number of users and applications. Users generate a significant part of Web content and traffic: they create, connect, comment, tag, rate, remix, upload/download new or existing resources in an architecture of participation, where user contribution and interaction add value. Users are also involved in a broad range of social activities like creating friendship relationships, recommending and sharing resources, suggesting friends, creating groups and communities, commenting friends activities and profiles and so on. But one of the most important activities, that became possible, is the collaborative e-learning. The emergence of new technologies and approaches provoked a huge increase in the use of Learning Management Systems (LMSs) for teaching and knowledge dissemination across the Web. The impact of e-learning is evident, shortly it can be formulated as ”convenience and portability (physical and temporal flexibility), cost and selection (wide range of courses and prices, different levels), individualization and a higher level of learner implication”( [1]). This changing reality poses new conditions and requirements for e-learning platforms. As it is mentioned in [1], ”the amount of knowledge that we deal with is much bigger than before, the interrelations between different forms of information are much more complex ”. Due to this, new approaches for learning process organisation have to be created. Thus, a model in which a teacher is the monopolising agent and the authorised representative of knowledge is no more adequate. Learners should be allowed to have their own knowledge and share it with teachers, in other words teachers and learners could switch their roles during the learning process. Such an opportunity is available in educational systems, based on the wiki [2] paradigm. Since its inception in the early 2000s, wiki technology became an ubiquitous pillar for enabling large-scale collaboration. Wiki approach enabled the creation of the largest encyclopedia of human-mankind edited by tens of thousands of volunteer editors – Wikipedia. Another successful example of largescale collaboration around textual educational content is the Wikibooks.org resource. However, Ward Cunningham’s wiki paradigm was mainly only applied to unstructured, textual content thus limiting the content structuring, re-purposing and reuse. Some steps in this direction are made by Semantic Web technologies and their combination with the wiki paradigm. Two kinds of semantic wikis exist: semantic text wikis, such as Semantic MediaWiki [3] are based on semantic annotations of the textual content; semantic data wikis, such as OntoWiki [4], [5] are directly based on the RDF data model. In many potential usage scenarios, however, the content to be managed by a wiki is neither purely textual nor fully semantic. Often (semi-)structured content (e.g. presentations, diagrams, etc.) should be managed and the collaboration of large user communities around such content should be effectively facilitated. Thus, a proper community collaboration, authoring, versioning, branching, reuse and re-purposing of educational content similarly as we know it from the open-source software community is currently not sup1

Semantic Knowledge Management for E-learning

ported. Besides, learners come from different environments, have different ages and educational backgrounds. That makes their integration into one single group more complicated or even impossible. That is why, new approaches should give the possibility to personalize the learning process. That means, that modern e-learning platforms should present only the information that is really relevant for the learner, in an appropriate manner, and at the appropriate time [6]. The research motivated in this proposal aims at providing new approaches for building a learning platform, that satisfy all the modern requirements: collaborative creation, editing and usage of educational material, enhanced possibilities to (re-)structuring, reuse and re-purpose these objects and personalization of the educational environment. Instead of dealing with large learning objects (often whole presentations or questionaires), we will decompose them into finegrained learning artifacts. Thus, instead of a large presentation a user will be able to edit, discuss and reuse individual slides; instead of a whole questionnaire she will be able to work on the level of individual questions. The result will comprise an architecture, methods and a prototype of such a platform. It will include some innovative, customized methods and approaches for knowledge structuring and personalization of the learning process.

2

User Scenario

Professor Wenzel wants to create a slide presentation, based on materials of the Mathematical Analysis course. He thinks this presentation will help his students to prepare for exams. He also wants to create a list of theoretical and practical questions for students to test themselves. Anna is a student of another university, she speaks another language. She needs to take an exam in the Mathematical Analysis course. She thinks, that she has not enough material for training, and would like to find more. • Case 1. Professor Wenzel creates his presentation locally and uploads it to the on-line storage. Then he gives his students a direct link to the presentation. In addition, he creates a list of questions manually in a separate document. Students could not learn this material collaboratively, it is very difficult to make and discuss any changes. In this case, Anna is unable to find this presentation, she will try to use some on-line books. It is also very difficult for her to test herself and get only the necessary material. • Case 2. Professor Wenzel creates his presentation locally and uploads it to an LMS. This helps him to share his presentation not only for his students, but also for all the people, using the same platform. Now it is possible to add metadata and a list of questions in a separate document (page). As a help for answers, professor can manually create links to his own materials. He can also manually create links for other educational materials, stored in the system. To do this, Prof. Wenzel firstly has to look through each document to be sure, that there is an explanation of the particular question. He can mention only the documents on the language he speaks. Anna now can find this presentation, she also can find many other materials. But, as she speaks another language, it will be very difficult or even impossible for her to manually translate this document. Thus, in this case, she will use another materials from the system in her own language. Prof. Wenzel’s students can also browse the list of questions and find out which themes they aware not well. Using the links, that were manually created by Prof. Wenzel, they can find the documents, where these questions are explained. But the particular place in the document they have to find by themselves. The presentation can be collaboratively discussed but not edited (if for example a student spots a mistake).

2

Darya Tarasova


• Case 3. Professor Wenzel uses Semantic Learning Management System (SLMS) to create his presentation. During the each slide editing process, the system asks him if he wants to create a list of questions. If he accepts, the system helps him to build this list and creates and maintains links between questions and answers on the slides semi-automatically. Also, the system semi-automatically creates some metadata and helps to structure the presentation accordingly to the semantics (e.g. in a topic hierarchy). Now each semantic structure (i.e. slide) has its own questions and metadata, so each slide could be considered as a separate learning artifact. With a built-in component the system can semi-automatically translate the presentation and questions into the language, Anna speaks. Anna could now start determining her deficiencies from the questions. Based on her answers, the system automatically offers her a personalized list of learning artifacts from different documents translated into her language. This could not only be the slides from Prof. Wenzel’s presentation, but all related material, stored in the system. She can sort this material in terms of level, date, university or other metadata, that was added manually or automatically. Each of these fine-grained learning artifacts could be edited and discussed collaboratively with other teachers and students.

Figure 1: Three scenarios of e-learning process: (a) e-learning without LMS; (b) e-learning in a usual LMS; (c) e-learning in the proposed SLMS.

3

State of the Art

Collaborative creation of e-learning content. The importance of creating reusable and re-purposable e-learning objects is meanwhile widely accepted by the e-learning community [7]. 3

Darya Tarasova


However, it seems that most of the works addresses the learning object reuse problem rather by means of semantic meta-data annotations, content tagging and packaging than by creating richly structured, reusable learning objects from the ground. The importance of creating learning objects already with reuse in mind was, for example, stated by [1]: ‘Content ... should be represented not as an object of study but rather as necessary elements towards a series of objectives that will be discovered in the course of various tests’. There are only few approaches for the direct authoring of reusable content, such as, for example, learning examples creation [8] or semantic structuring and annotation of video fragments [9]. In this proposal we aimed to create an educational content creation platform grounded on reusable content authoring and large-scale community collaboration (or crowd-sourcing). Wiki-based collaborative knowledge engineering. The importance of wikis for collaborative knowledge engineering is meanwhile widely acknowledged. In [10], for example, a knowledge engineering approach which offers wiki-style collaboration, is introduced aiming to facilitate the capture of knowledge-in-action which spans both explicit and tacit knowledge types. The approach extends a combined rule and case-based knowledge acquisition technique known as Multiple Classification Ripple Down Rules to allow multiple users to collaboratively view, define and refine a knowledge base over time and space. In a more applied context, [11] introduces the concept of wiki templates that allow end-users to define the structure and appearance of a wiki page in order to facilitate the authoring of structured wiki pages. Similarly the Hybrid wiki approach [12] aims to solve the problem of using (semi-)structured data in wikis by means of page attributes. We will apply the wiki paradigm and selected wiki-based knowledge engineering approaches for the creation and collaboration around e-learning artifacts and learning objects. Semantic wikis. Another approach to combine wiki technology with structured representation are semantic wikis [13]. There are two types of semantic wikis: semantic text wikis, such as Semantic MediaWiki [3] or KiWi [14] are based on semantic annotations of the textual content; semantic data wikis, such as OntoWiki [4], [5], are based on the RDF data model in the first place. Both types of semantic wikis, however, suffer from two disadvantages. Firstly, their performance and scalability are restricted by current triple store technology, which is still an order of magnitude slower when compared with relational data management, which is regularly confirmed by SPARQL benchmarks such as BSBM [15]. Secondly, semantic wikis are generic tools, which are not particularly adapted for a certain domain thus substantially increase the usage complexity for users. The latter problem was partially addressed by OntoWiki components such as Erfurt API, RDFauthor and Semantic Pingback, which evolved OntoWiki into a framework for Web Application development [4]. Semantic Web technologies for e-learning. Research regarding Semantic Web for e-learning has provided a wide variety of papers. Thus, [16] discusses how Semantic Web could be used in e-learning, and describes tools that could be developed to support context, socialization, discussions and conceptual modelling. In [17], it is described the architecture of a service-based e-learning system using metadata for dynamic contexts. The [18] overviews Semantic Web and its use for Web learning, and proposes a model for the development of semantic services for learning Web communities. Henze [19] developed a framework for workspace personalization using RDF/S and a service-oriented architecture . The works [20], [21] offer a system to provide teachers with feedback about the interaction between learners and learning resources. Recently, the approach in [22] builds an ontology with the language DAML + OIL and fulfills the integration within an e-learning platform to give semantics for the contents. After that, a multi-agent architecture is applied over the e-learning system and a reasoning engine provides learners with 4

Darya Tarasova


intelligent recommendations for their tasks. However, the Semantic Web technologies implementation, described in all these papers, is not an easy process. A lack of domain ontologies for learning courses is a barrier to the use of ontologies in e-learning systems and the ontology creation process is difficult and time consuming for instructors. It is not always clear for the average user, for what he has to do so much additional work. One of the important features of the proposed approach is the simplicity and clearness for both teachers and learners. And the proposed approach aims to achieve the level of simplicity, where end users do not experience additional difficulties, regardless of the underlying technologies.

4

Methodology

The methodology of the doctoral project described here is scheduled over a period of three years and will be conducted in two cycles. In the first cycle, we will start with a state-of-theart analysis to review existing literature related to Semantic Web technologies in context of e-learning. After that we will start to implement our approach with the prototype semantic wiki application, dealing with semi-structured e-learning content. When the basic functionality will be implemented, we will start to research and develop stuff for personalization, such as user model. After that we will work on the translation component. When all the components will be finished, we will work on the overall platform implementation due to increase its performance and usability. All phases have two horizontal activities directly integrated: Firstly we will develop a system demonstrating the usefulness of the developed research approaches and secondly we will evaluate the research results either based on our implementation or other suitable evaluation metrics. Each of the individual steps will result in a peer-reviewed conference or workshop paper thus documenting and publishing the results of the research. We will re-iterate all these steps in the second cycle to account for new or improved methods/techniques/services that may be introduced during that period by taking into account all the lessons learned from the first phase so as to avoid repeating mistakes. This iterative research and methodology also takes the dynamic character of the field into account, with many approaches being published and systems being developed every year. We will submit extended version of the conference and workshop articles incorporating improved results from the second iteration, as articles to suitable renowned journals. The last four months will be dedicated to completing the writing of the PhD thesis, where the previously published results will be compiled into a comprehensive monographic work. Table 1 outlines the proposed time plan for the project over a timespan of three years.

4.1 4.1.1

Individual Work Package Descriptions WP1: State-of-the-Art analysis

Motivation: In order to be aware of the current trends in the field of Semantic knowledge management for e-learning, we need to conduct a state-of-the-art analysis. All existing literature related to the research subject needs to be reviewed. Also, all tools and user interfaces available to perform the task of semantic knowledge management in context of the educational content need to be evaluated. Based on the literature review, the state-of-the-art specification will end with a detailed technological and scientific road map for the doctoral project. Methodology: We will follow a formal systematic literature review process for this study based on the guidelines proposed in [23]. A systematic literature review is an evidence-based approach to thoroughly search studies relevant to some pre-defined research questions and critically select, appraise, and synthesize findings to answer such research questions. Systematic reviews maximize

5

Darya Tarasova


Work Package 1 2 3 4 5 6 7

Phase

Time Span (Months)

State-of-the-Art analysis Using semantic wikis for e-learning content creation User adaptation of semantically represented e-learning content Semi-automatic translation of semantically structured learning objects Implementation and evaluation Phase II - Re-iteration of WP 1 - 5 Writing PhD thesis

1-2 3-6 6 - 10 11 - 14 15 - 17 18 - 31 32 - 36

Table 1: Proposed time plan. the chance to retrieve complete data sets and minimize the chance of bias. As a part of the review process, we will develop a protocol that provides a plan for the review in terms of the method to be followed, including the research questions and the data to be extracted. Results: The result of this work package will be a list of related studies that show the current trend in the area of Semantic knowledge management for e-learning. It will provide a set of guidelines to design suitable and effective systems for e-learning. An article on the state-of-theart review will also be written, which will provide an overview and accumulate existing similar approaches for further exploitation and incorporation into this project. Evaluation: To evaluate the results of this work package, we will consult experts in the area of Semantic Web and will ask them to assess the list of related studies derived from our systematic literature review for completeness and review our categorization scheme. 4.1.2

WP2: Using semantic wikis for e-learning content creation

Motivation: As a semantic wiki paradigm has a great potential for building an effective elearning platform, we need to find the best solutions for its disadvantages, mentioned above. Additional features should be added for the test assessment mechanism. Also, we have more complex data structure, as is used in existed semantic wikis, so the problem of versioning and evolution also is important field of research. Methodology: We will implement our approach with the development of the wiki system dealing with slide presentations. We will consider the full presentation to be a learning object and each slide to be a leaning artifact. In that case a presentation can be viewed as a treestructure, that can include other slides and presentations inside. We will generally follow the approach, that content objects are not updated or deleted, but instead, in accordance with the data model, new revisions will be created. For that, we need the algorithm for determining the cases, in which new revision of a learning object should be created. This algorithm should prevent uncontrolled proliferation of the revisions. Also we need to develop algorithms for automatically tests creation and assessment. Each slide can have one or several questions that it answers. All these questions are connected both with the slide and with the full presentation. The system will randomly create a questionnaire for a student across the presentation topic and then dynamically update it depending on the answers. During the work on this package we will develop and evaluate algorithms for this task. This will not only be useful feature itself, but also, based on the results, the system in the future will provide to the user further personalized content. As we have complexed data model, we also have to focus on the creating effective user interface.

6

Darya Tarasova


Results: An approach and a system prototype for collaborative learning artifacts creation, discussing and recombining; algorithm for dynamical tests creation and its implementation. Evaluation: In order to determine whether we succeeded to effectively hide the system’s data model complexity, we will perform a usability user study. Subjects will be drawn from the members of AKSW research group, the computer science department at the university of Leipzig. We will use the System Usability Scale (SUS) [24] to grade the usability of the platform. In addition to quantitative results, we also will collect a number of user suggestions, that will help us in the next phase. 4.1.3

WP3: User adaptation of semantically represented e-learning content

Motivation: The behaviour of a personalized e-learning system varies accordingly to the data from the user model and the user profile. Users are different, they have different background, different knowledge about a subject, different preferences, goals and interests. To individualise, personalize or customise actions in a semantic e-learning system, we need a user model that allows the selection of individualised interaction with the user. Methodology: Figure 2 shows the structure of the envisioned e-learning adaptation system. The user model contains all information that the system knows about the user. In order to create a user model, we first need to collect some information about the target user. The typical attributes maintained in the user model are user preferences, interests, attitudes and goals, proficiencies (e.g. task domain knowledge, proficiency with system), interaction history (e.g., interface features used, tasks performed/in progress, goals attempted/achieved, number of requests for help) and user classification (i.e. stereotype). In our case we will also extensively use the user’s answers on tests questions, as it’s very natural way that does not require additional actions from the user. Thereafter, it is maintained by the system, although the user may be able to review and edit her profile. User actions and events at various conceptual levels, such as mouse clicks, task completion and requests for help, are reported by the user interface or core application to the user profile. An analysis engine combines the user profile with other models of the system to derive new ”facts” about the user. The analysis engine can update the user profile with the derived facts or initiate an action in the application (such as interrupting the user with a suggestion). The analysis engine also responds to queries from the application. For constructing the user model, we have to recognize the behavior of E-learners by choosing and employing appropriate modeling techniques [25] like Social Network Analysis, Bayesian (probabilistic), Logic-based (e.g.inference techniques or algorithms), Machine learning (e.g. neural networks), Stereotype-based or Inference rules techniques. Results: User modeling for an personalized semantic e-learning system yields a set of models and rules for generating the educational content at run time. This user model gives the possibility to distinguish between users and provides the system with the ability to tailor its reaction depending on the model of the user. Evaluation: In order to evaluate the result of this workpackage we compare the resulting adaptive system with its analogous non-adaptive instance. We are then able to see whether an adaptive interface is measurably better than a non-adaptive interface, and under what conditions the benefit outweighs the perceived loss of control due to unannounced adaptations or the interruption from the interface component as it interacts with the user to adapt or offer advice. 4.1.4

WP4: Semi-automatic translation of semantically structured learning objects

Motivaton: Since all the content is versioned and semantically structured, it is possible to translate it and to keep track of changes in various multi-lingual versions of the same content object. This feature will greatly enhance the efficiency of the proposed platform, as it will 7

Darya Tarasova


Figure 2: The structure of the envisioned e-learning adaptation system. decrease the language barrier between teachers and learners and increase the re-using and repurposing of the content. Methodology: To implement semi-automatic translation of the learning objects we will use existing automatic translation services such as Google translate. This task has at least two difficulties to be solved. The first research should be conducted to find the right granularity for automatic translation. The larger the granule, the better the translation quality due to the context. But the large granule increases the complexity of keeping different language versions synchronized. We will determine the right granularity of content to be submitted for translation - we have to balance between large content chunks, which will allow the automatic translation to employ context sensitive translation. The other problem is to combine manual and automatic translation. A feedback loop between automatic translation and manual refinement should be created. To organise this we will use personalized thesauri, that allows us to consider background knowledge for individual user communities and domains. Results: An approach, methodology and a prototype of a component, that allows to translate educational content in context of multi-versioned system. Evaluation: We will compare different automatic learning artifact translation strategies with a manually created gold standard and thus assess which granularity of content provides the best ratio between translation quality and re-usability. 4.1.5

WP5: Implementation and evaluation

Motivation: After the main research approaches (from WP2-4) are implemented, we have to integrate these into a coherent system and to optimize its scalability. We have to optimize all the program components, build on the previous steps, and create ergonomic user-friendly interface. Also we need to implement the functionality for importing and exporting presentations in/from the system. Without such possibility (i.e. making backups and transferring data to other data 8

Darya Tarasova


mediums or applications) user will be discouraged from contributing and maintaining data on the platform. Methodology: As an approach for implementation of the proposed platform we will use modelview-controller (MVC) architecture pattern. The MVC architecture enables the decoupling of the user interface, program logic and database controllers and thus allows developers to maintain each of these components separately. Since our application will be implemented as a typical Web application the view layer is accessed by users via a Web browser using the HTTP protocol, AJAX and HTML, CSS documents. In both the front-end and back-end part of the application we will extensively use existing open-source components. This will allow to reduce the time of implementation. A crucial feature of the proposed platform is an ability to import and export data into different formats. The main data format that will be used is HTML. However, there are other popular formats commonly used by desktop application users, such as PowerPoint presentations and Excel Diagrams. Thus we will implement importing and exporting functionality for above-mentioned formats. Results: A prototype of the e-learning platform, where all the research approaches from the previous work packages are implemented in an optimal way. Evaluation: To measure the system performance, we will conduct benchmarking. We will measure the performance with different datasets of educational objects and compare it with potentially similar systems. To compare the scalability we also will plot the average execution time for each operation for our system and possible alternative implementations. Based on the measurement results, we will be able to determine whether our system performance is sufficient, in what operations, how the system scales and what should be improved in another development iteration. 4.1.6

WP 6: Phase II - Re-iteration of WP 1 - 5

Motivation: We expect WP 1 to WP 6 to require approximately 18 months. In order to account for new or improved methods/techniques/services that may be introduced during that time period, we will re-iterate WP 1 to WP 5 in the second phase of our project by incorporating the lessons learned from the first phase. This will enhance our observatory with an additional information from new resources and also open up possibilities to provide new dimensions for semantic knowledge management in e-learning. Moreover, this work package will further improve the base architecture and even the user interface thus increasing the quality of the system as a whole. This incremental methodology also ensures that we obtain first results early in the project, which in turn will allow us to adapt corrective measures for the second cycle, in case of problems. Methodology: We will first review the new literature published after our first systematic literature review. From that, we might get new challenges or new areas of research in the context of adaptive authoring for the Social Web. We will then revisit the WP 2, 3, 4 and 5 for the new methods/techniques/services. Using the new results, we will try to enhance the performance of our proposed reference architecture for SLMS. Results: Identification and addition of new methods/techniques/services to enrich the existing ones with additional information to answer even more challenging questions. Enhancement and refinement of pre-defined methods and approaches for adaptive learning interfaces. Evaluation: The results obtained in the second phase will be compared with those from the first phase to ensure that the additionally included results are relevant, indeed go beyond previous work and provide useful information.

9

Darya Tarasova


4.1.7

WP 7: Writing PhD thesis

The last four months will be spent in writing the doctoral thesis including organizing the material resources and completing the work. The doctoral thesis will contain detailed information on each of the work packages and also on the reference architecture for an SLMS. Since we aim to publish result of each of the individual steps as workshop, conference or journal articles, large parts of the material required for the thesis will be already available in good quality and have gone through a thorough peer-review process. This will simplify the integration of the work into a comprehensive monographic work as required for a doctorate thesis.

5

Result: Semantic Knowledge Management Approach and System Prototype for E-learning

This project will result in an intelligent platform for collaborative learning which is an important prerequisite to achieve the Social Semantic Web vision. On the one hand, the proposed platform will provide the possibility to structure educational material semi-automatically. On the another hand, the platform will be able to produce personalized learning artifacts on the user’s language from the amount of semantically structured educational materials. Such objects could be created directly within the system, and then edited and re-combined collaboratively. As a part of my internship at University of Leipzig a first feasibility study of a semanticsbased e-learning system called SlideWiki was already created. A screenshot of the user interface mock-up is shown in Figure 3

Figure 3: The interface for slide editing in the SlideWiki proof-of-concept application.

10

Darya Tarasova


6

Potential Impact

For learners: The proposed approach is aimed to improve existing e-learning technologies with the recommendations of the specialists in the field of human education. The wiki-based implementation will allow users to be active participants in the educational process instead of passive learners. The content personalization mechanisms will increase the effectiveness of the educational process, as learners will spend less time for searching the necessary material and more time directly on the learning. The synchronized content translation algorithms will increase the quantity of available material for each student. Also, learners will have a possibility to compare the views on the problem in the different countries. It also increases the overall quality of the educational material. We expect that our approaches and methods for organising a semi-automatic tests assessment will be useful feature for self-verification. For instructors: We expect, that the proposed approach will decrease the time-consuming of the educational content creation process. It is expected because of possibility of truly collaborative work process organisation. Our approach implies, that both teachers and learners could participate in the learning objects creation. This promotes the destruction of the border between teachers and learners and allows to engage in direct dialogue. The possibility of re-use and repurpose of the content also decreases the time spent on the object creation and allows to show the relations between different topics. As learning objects are personalized semi-automatically, instructors do not need to create different versions for different groups of learners. As a content will be translated semi-automatically, instructors could get feedbacks and contributions from users from all over the world, that will help to improve the quality of the material. The proposed approach of semi-automatic tests assessment allows teachers to create effective tests easily and quickly. For academical community: The proposed approach could be implemented for building an e-learning platform, where potentially large communities of teachers, lecturers, academics are empowered to create sophisticated educational content in a truly collaborative way. For newly emerging research fields, the proposed facility could help to disseminate content and educate PhD students and peer-researchers more rapidly, since the burden of creating and structuring the new field can be distributed among a large community. Specialists for individual aspects of the new field can focus on creating educational content in their particular area of expertise and still this content can be easily integrated with other content, re-structured and re-purposed. A particular aspect, which is facilitated by our approach implementations is multi-linguality. As a consequence, we expect that our approach will have a substantial impact with regard to disseminating educational content in many languages.

7

My Prior Research

The proposed doctorate thesis builds on my previously acquired competences and research interests in an ideal way. I have got a BSc in Applied Computer Science and my MSc was in the same field. During my education I did a lot of research on Semantic Web and knowledge management. In particular, I worked on semantic knowledge management as an intersection of these two fields which aims to bring Semantic Web techniques to collaborative knowledge engineering. That was very similar to the proposed research but in a different context. My master thesis was entitled “Development

11

Darya Tarasova


of the knowledge management system for virtual enterprises based on Semantic Web technologies”. The research mainly aimed to find the best solutions and approaches of using semantic technologies for effective collaborative knowledge engineering. My research was truly new for the certifying committee, I was asked a lot of questions, and as a result, I have got an excellent grade. Based on this project I have published two research papers, this work also allowed me to start a scientific research at the University of Leipzig. As a research intern in the Agile Knowledge and Semantic Web research group at the University of Leipzig I was involved in the development of the application prototype called “’SlideWiki’. This application represents a feasibility study for the research proposed in this application. Thus, participation in this application’s development could be considered as my first step to the future e-learning platform. Based on the work done, we have written a research paper, describing our approach for (semi-)structured data management. This paper is submitted to the World Wide Web conference (WWW’2012) and is now under the review.

References [1] Nieves Pedreira, José Ram´ on Méndez Salgueiro, and Manuel Mart´ınez Carballo. E-learning in new technologies. In Encyclopedia of Artificial Intelligence, pages 532–535. IGI Global, 2009. [2] Bo Leuf and Ward Cunningham. The Wiki way: quick collaboration on the Web. AddisonWesley, London, 2001. [3] Markus Kr¨ otzsch, Denny Vrandeˇcić, Max Völkel, Heiko Haller, and Rudi Studer. Semantic Wikipedia. Journal of Web Semantics, 5(4):251–261, 2007. [4] Norman Heino, Sebastian Dietzold, Michael Martin, and Sören Auer. Developing semantic web applications with the ontowiki framework. In Networked Knowledge - Networked Media, volume 221 of Studies in Computational Intelligence, pages 61–77. Springer, Berlin / Heidelberg, 2009. [5] S¨ oren Auer, Sebestian Dietzold, and Thomas Riechert. Ontowiki - a tool for social, semantic collaboration. In The Semantic Web - ISWC 2006, 5th International Semantic Web Conference, pages 736 – 749. Springer, 2006. [6] Robin D. Morris. Web 3.0: Implications for online learning. Techtrends, 55(1):42–46, 2006. [7] Vladan Devedzic. Semantic Web and Education (Integrated Series in Information Systems). Springer-Verlag New York, Inc., Secaucus, NJ, USA, 2006. [8] Yen-Hung Kuo, Qing Tan Kinshuk, Yueh-Min Huang, Tzu-Chien Liu, and Maiga Chang. Collaborative creation of authentic examples with location for u-learning. In IADIS International Conference e-Learning 2008, pages 16 – 20. IADIS, 2008. ´ [9] Elena Garc´ıa Barriocanal, Miguel-Angel Sicilia, Salvador Sánchez Alonso, and Miltiadis D. Lytras. Semantic annotation of video fragments as learning objects. Interactive Learning Environments, 19(1):25–44, 2011. [10] Debbie Richards. A social software/web 2.0 approach to collaborative knowledge engineering. Inf. Sci., 179(15):2515–2523, 2009.

12

Darya Tarasova


[11] Anja Haake, Stephan Lukosch, and Till Sch¨ ummer. Wiki-templates: adding structure support to wikis on demand. In Proceedings of the 2005 International Symposium on Wikis, pages 41–51. ACM, 2005. [12] Florian Matthes, Christian Neubert, and Alexander Steinhoff. Hybrid wikis: Empowering users to collaboratively structure information. In ICSOFT (1), pages 250–259, 2011. [13] Sebastian Schaffert, Fran¸cois Bry, Joachim Baumeister, and Malte Kiesel. Semantic wikis. IEEE Software, 25(4):8–11, 2008. [14] Sebastian Schaffert, Julia Eder, Szaby Gr¨ unwald, Thomas Kurz, Mihai Radulescu, Rolf Sint, and Stephanie Stroka. Kiwi - a platform for semantic social software. In SemWiki, volume 464 of CEUR Workshop Proceedings. CEUR-WS.org, 2009. [15] Christian Bizer and Andreas Schultz. The berlin sparql benchmark. Int. J. Semantic Web Inf. Syst., 5(2):1–24, 2009. [16] V. P. Yli-Luoma and A. Naeve. Towards a semantic e-learning theory by using a modelling approach. British Journal of Educational Technology, 37(3):445–459, 2006. [17] S. Dietze, A. Gugliotta, and J. Domingue. Towards adaptive e-learning applications based on semantic web services. Proceedings of workshop TENCompetence open workshop on service oriented approaches and lifelong competence development infrastructures, pages 75–83, 2007. [18] Martin Dzbor, Arthur Stutt, Enrico Motta, and T. Collins. Representations for semantic learning webs: Semantic web technology in learning support. J. Comp. Assisted Learning, 23(1):69–82, 2007. [19] Nicola Henze. Personal readers: Personalized learning object readers for the semantic web. In AIED, volume 125 of Frontiers in Artificial Intelligence and Applications, pages 274–281. IOS Press, 2005. [20] Jelena Jovanovic, Dragan Gasevic, Christopher A. Brooks, Vladan Devedzic, Marek Hatala, Timmy Eap, and Griff Richards. Using semantic web technologies to analyze learning content. IEEE Internet Computing, 11(5):45–53, 2007. [21] Carlo Torniai, Jelena Jovanovic, Dragan Gasevic, Scott Bateman, and Marek Hatala. Elearning meets the social semantic web. In ICALT, pages 389–393. IEEE, 2008. [22] J¨ urgen Dunkel, Ralf Bruns, and Sascha Ossowski. Semantic e-learning agents - supporting e-learning by semantic web and agents technologies. In ICEIS (4), pages 271–278, 2004. [23] B. Kitchenham. Procedures for performing systematic reviews. Technical report, Keele University and NICTA, 2004. [24] James Lewis and Jeff Sauro. The Factor Structure of the System Usability Scale. In Human Centered Design, volume 5619 of LNCS, chapter 12, pages 94–103. Springer Berlin / Heidelberg, Berlin, Heidelberg, 2009. [25] User Modeling, Adaption and Personalization - 19th International Conference, UMAP 2011, Girona, Spain, July 11-15, 2011. Proceedings, volume 6787 of Lecture Notes in Computer Science. Springer, 2011.

13

Darya Tarasova

Research Proposal for DAAD PhD scholarship Semantic Knowledge Management for E-learning

Recommend Documents