Knowledge-based generation of educational web pages
1 like755 views
This document discusses generating personalized web pages for tutoring systems using knowledge-based approaches. It covers key topics like ontologies, student modeling, cognitive psychology, and hypertext. Personalized web pages can be adapted based on a student's knowledge, learning style, goals, preferences and other factors inferred from their interactions. The document argues that web pages should be designed following principles of cognitive ergonomics and rhetoric to facilitate understanding and avoid issues like high cognitive load.
Knowledge-based generation of educational web pages
- 1. Knowledge-Based Contents Generation of Personalized Web Pages Introduction for Tutoring Web resources for learning Stefan Trausan-Matu Web page generation Computer Science Department, Knowledge Bucharest "Politehnica" University, and Computer-Human Interaction Romanian Academy Center for Artificial Intelligence Web page generation ROMANIA trausan@cs.pub.ro http://www.racai.ro/~trausan Stefan Trausan-Matu, ITS 2002, Biarritz 2 Intelligent Tutoring Systems Knowledge based systems Student modeling Reasoning for: Introduction Student diagnosis Explanations generation Lesson planning Intelligent interfaces Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 3 Biarritz 4 Implied CS domains for ITS on the web Artificial Intelligence Computer- Human ITS = Human learning as supervised Interaction knowledge acquisition Artificial Intelligence Knowledge-based systems Planning Web Natural Language Processing technologies Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 5 Biarritz 6 1
- 2. Computer-Human Interaction Web technologies User (learner) modeling Distributed computing Personalization (Re)use web-based resources Intelligent interfaces Client-server, web services Cognitive psychology Huge amount of information available Cognitive ergonomics on the web Permanent evolution of the information on the web Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 7 Biarritz 8 Knowledge-based generation of web pages for tutoring Enhancing ITS with the advantages offered by the possibility of browsing the web : Intelligent reuse web resources Web resources for learning Integrate new information from the web Web rhetoric Personalized web pages Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 9 Biarritz 10 Learning on the web Resources on the web Web is a very good place for learning Databases New information must be coherently integrated in the body of knowledge in Knowledge bases (ontologies) order to keep a holistic character of the Dictionaries, glossaries, and thesauri body of knowledge Hypertexts and hypermedia Specific web rhetoric Computer programs (e.g. applets) Texts and corpora (annotated or not) Images, films, sound Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 11 Biarritz 12 2
- 3. Structure of resources on the Text perspectives web Unstructured (e.g. TEXT, images) - Signs (Peirce, de Saussure): syntax, hidden structure - Natural Language semantics, pragmatics - Semiotics Processing Linguistics Semi-structured (e.g. HYPERTEXT) - Metaphors HTML, XML Philosophy of language Structured (e.g. databases) Rhetoric Psycholinguistics Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 13 Biarritz 14 Text organization Hypertext Linear organization - essay, story Text with extra dimensions Hierarchical organization - treaty, Personalized reading manual Easy browsable with computer-human Network organization - hypertext, interfaces hypermedia Offers the possibility of mapping to a conceptual structure Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 15 Biarritz 16 Hypertext - facilitator of Hypertext - facilitator of human understanding: human understanding: Theodor Nelson, who coined the term Hypertext was introduced by Douglas "hypertext", defined it as the Engelbart, in the early sixties, as a : hyperspace of concepts from a given text or : "Conceptual framework for augmenting "A system for massively parallel creative human intellect" (Engelbart, 1995) work and study ... to the betterment of human understanding" (Nelson, 1995) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 17 Biarritz 18 3
- 4. World Wide Web Hypertext(media) + Internet + User Friendly Interfaces Text (+images ...) + Knowledge communication, distribution, agents + interfacing, cognitive ergonomics (HCI, CHI, HCD) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 19 Biarritz 20 Knowledge Knowledge-Based Systems Learning is a knowledge centered activity: Explicit representation, in a so-called “Knowledge Base”, of the knowledge needed One of the main goals of a learning by the program process is the articulation in the The knowledge base may easy evolve - the learner’s mind of a body of knowledge representation used must facilitate: for the considered domain. knowledge acquisition The skeleton of this body is usually a learning semantic network of the main concepts The same knowledge base used in several involved in that domain. processing regimes Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 21 Biarritz 22 Ontologies Ontologies "An ontology is a specification of a Knowledge base = Ontology + … (rules) conceptualization....That is, an ontology is a description (like a formal specification of Concepts + Attributes + Relations (+ Axioms) a program) of the concepts and relationships that can exist for an agent Multiple ontologies - Ontology alignment ! or a community of agents" (Gruber) Needed for agents inter-communication (share of same concepts) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 23 Biarritz 24 4
- 5. PROGRAMMING_CONCEPT PROGRAMMING_ABSTRACTION DATA_ABSTRACTION Ontologies - Concepts MAPPING ARRAY CONTAINER TABLE HASHTABLE The central part of the domain ontology is a INDEXTABLE ARRAY taxonomically organized knowledge base of SYMBOLTABLE COLLECTION concepts: IMPLICITCOL EXPLICITCOL SET SYMBOLTABLE Security BAG Bond DISPENSER STACK Share QUEUE HEAP OrdinaryShare CURSORSTR PreferenceShare LINKEDLIST CURSORTREE Stock CONTROL_ABSTRACTION Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 25 Biarritz 26 Ontologies - Relations Ontologies - Attributes Each concept has attributes. For example, Each concept may be related with other a share has the following attributes: concepts. Related terms with share are: the shareholder, earnings per share share capital, share premium account dividend. gain issue Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 27 Biarritz 28 Ontologies - Languages Ontologies on the web Description logics : LOOM, CLASSIC, General lexical ontologies : Fact WordNet XML-Based : DAML+OIL, OML EuroWordNet BalkanNet MikroKosmos FrameNet Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 29 Biarritz 30 5
- 6. Exchange of ontologies on the Ontologies on the web web Domain specific Particular ontologies are now sharable Supper Upper Ontology on the web with XML-based languages like DAML+OIL. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 31 Biarritz 32 Ontologies used in ITSs Ontologies in ITSs used for : Domain Learner modelling - overlay, buggy Tutoring Text processing Test generation and selection Human-computer interfacing Learner diagnosys Lexical Authoring Upper Level Knowledge acquisition Course planning Web page generation Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 33 Biarritz 34 Computer-Human Interaction (CHI) Among others, it studies: Cognitive ergonomics Computer-Human Interaction Immersive interfaces Learner (user) modeling Personalization Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 35 Biarritz 36 6
- 7. Important issues in cognitive Cognitive ergonomics ergonomics of web pages: Studies the ways in which human-computer Cognitive load interfaces can be tailored to users' cognitive characteristics. Lack of orientation It is very important to design cognitive Web rhetoric ergonomic web pages. Facilitate understanding If you design web pages that are not cognitive ergonomic, few people will stay browsing them (when they have the possibility of surfing a tremendous number of other pages). Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 37 Biarritz 38 Cognitive load Lack of orientation Mental (cognitive) effort needed to You could spend even whole days surfing in browse the web pages cyberspace, forgetting the starting point, the path you followed, or the starting goals (all One solution is to assure a holistic these might be one of the causes of its character for the body of knowledge attractiveness, but it may become something induced in the learner’s mind. The like drug-addiction). learning process must induce the sense Therefore, a well designed structure of the of the whole. New concepts must fit in links topology, easy to understand for the whole. anybody is very important. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 39 Biarritz 40 Web rhetoric Web rhetoric Similarly to a lawyer that uses rhetoric " In the course of designing a hyper document, an author is generally confronted with three sub to convince the jury, you must use problems which correspond to the classical fields of rhetoric in your web pages in order to rhetoric, i.e. inventio, dispositio and elocutio. He must: obtain the best results with generate and select relevant information (inventio), communication in your web pages structure resp. order the selected information (dispositio), and present the ordered information in an adequate way (elocutio).“ (Thuering, M., Hannemann, J., Haake, J.M., 1991) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 41 Biarritz 42 7
- 8. Understanding Empathy Explanation vs. Understanding "empathy is a phenomenon in which Understanding implies an emphatic one person can experience states, relation, which involves the immersion thoughts and actions of another person, of the learner in a context. (vonWright) by psychological transposition of the Different interpreters may have self in an objective human behavior different understandings of the same model, allowing the understanding of sign. the way the other interprets the world “ Understanding requires experiencing (…………..) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 43 Biarritz 44 Very important in immersion Immersion are the space and time perception or imagination in "The state of being overwhelmed or images (perceived or imagined) in deeply absorbed; deep engagedness". which objects are identified; (Webster Dictionary, 1999) the possibility and experience of real, "If you immerse yourself in something, simulated or mental walkthrough in the context of immersion; you become completely involved in it." (Collins Dictionary, 1999) the experience of actions (real of imagined) done by the immersed person. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 45 Biarritz 46 Immersion done by Flow state Flow state (Alan Cooper, “About Face”), e.g. Physically entering in a context of the domain driving a car or skiing - induced by a perfect (for example, learning to drive a car by immersion: entering the care, starting it and driving), Simulations through, for example, computer sense of control graphics facilities (starting from simple navigation interactive computer graphic till virtual reality); loose of the sense of time Mentally, as a result of mental imagery, as a consequence of reading a text or browsing web pages. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 47 Biarritz 48 8
- 9. Immersion on web sites The World Wide Web has been proved as a very attractive and, meanwhile, very useful space to wander for almost anyone, including students. Therefore, it may be considered it as a very suitable medium to provide immersive learning CHI - Personalization The immersion illusion can be supported both by a structure of web pages Web browsing may generate a flow state Flow state may be useful for learning Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 49 Biarritz 50 Personalized web pages Personalized web pages From an ideal perspective, everybody has Are adapted to each users': to find WWW structured according to knowledge - ITS student model his needs, goals and cognitive learning style particularities. psychological profile goals (e.g. lists of concepts to be learned) level (novice, expert) preferences (e.g. style of web pages) context of interaction Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 51 Biarritz 52 Student model Learning style Keeps track of the concepts known, unknown Exploratory vs. interactional or wrongly known by the student (………) David Kolb’s learning styles : Inferred from results at tests or from Accomodator interaction (visited web pages, topics searched etc.) Diverger Is usually defined in relation with the domain Converger ontology (concept net, Bayesian net) Assimilator Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 53 Biarritz 54 9
- 10. Psychological profile Psychological profile Inferred from results at psychological Self-confidence tests or from interaction (time of Motivation visiting different types of web pages) Concentration Personality types Social interaction Intelligence Emotion profile Context dependence Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 55 Biarritz 56 Preferences Context of interaction Explicitly chosen by the learner Avoid monotony, fatigue or cognitive Inferred from behavior overload Inferred from the psychological style Rhetoric schemata Speech acts Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 57 Biarritz 58 Web page generation Content Structuring Web page generation Styling Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 59 Biarritz 60 10
- 11. Web rhetoric " In the course of designing a hyper document, Web page generation … generate and select relevant information (inventio), Content structure resp. order the selected information (dispositio), and present the ordered information in an adequate way (elocutio).“ (Thuering, M., Hannemann, J., Haake, J.M., 1991) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 61 Biarritz 62 Content types Content types - text Text Descriptions Questions and tests Justifications Explanations Links Questions Images and sounds Glossary Programs (e.g. applets) Index Links Help Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 63 Biarritz 64 Content types Content semantics Textual Conceptual structure Visual Semantic density Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 65 Biarritz 66 11
- 12. Content pragmatics for learning purposes Source of content Created (edited) by the professor - authoring Context tools Reused - Information retrieval - search Prerequisites for a content module agents Relations to other content modules text html Speech act role of content xml jpeg, mpeg etc. Automatically generated (text, tests) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 67 Biarritz 68 Dimensions of texts on the web Text structuring 1. Raw text 2. Text shown by the browser Bracketing 3. Annotated text (HTML, XML) Knowledge extraction and semantic 4. Style of presentation (CSS, XSL) 5. Hyperlinks relations 6. Structure of web pages Text segmentation 7. Knowledge in texts Rhetoric schema identification 8. Goals of the writer 9. The history of browsing web pages Automatic link generation 10. Effect on the reader Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 69 Biarritz 70 Text annotation Text segmentation Syntactic Identification of structures (e.g. lexical chains Part of speech - G. Hirst) of semantically related words “Bracketing” Uses WordNet or other lexical ontologies, which provides semantic relations among Semantic words Pragmatic synonims Rhetoric hypernims, hiponims meronyms, holonims Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 71 Biarritz 72 12
- 13. Natural Language Processing Natural Language Processing (NLP) approaches Parsing Annotation Grammar-based Knowledge extraction Statistical Document categorization Search for relevant documents Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 73 Biarritz 74 XML XML “eXtensible Markup Language” Universal markup language <Student> <ID>7321</I <FName>Steven</FName> Extends HTML facilities <Name>Collins</Name> <Year>4</Year> Simplified SGML </Student> Keeps 80% from SGML Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 75 Biarritz 76 XML additional features XML similarities with HTML comparatively to HTML Easy to use on Internet Extensibility - new types of annotations XML documents are easy to create and may be introduced process Universal representation language XML documents may be read with an ordinary text editor Separation of content, structure and visualization SGML compatible Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 77 Biarritz 78 13
- 14. XML additional features comparatively to HTML XML encourages semantics HTML XML <table> <?xml version="1.0"?> Facilities for semantic encoding <tr> <StudentsList> <td>7612</td> <Student> Allows different (personalized) <td>John</td> <td>Freeman</td> <ID>7612</ID> <FName>John</FName> presentations of the same document <td>3</td> </tr> <Name>Freeman</Name> <Year>3</Year> (by means of XSLT transformations) <tr> <td>7321</td> </Student> <Student> <td>Steven</td> <ID>7321</ID> <td>Collins</td> <FName>Steven</FName> <td>4</td> <Name>Collins</Name> </tr> <Year>4</Year> </table> </Student> </StudentsList> Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 79 Biarritz 80 XML Perspectives XML Perspectives Allows the definition of a grammar for a Universal markup of documents (simplified markup language: SGML) Explicitly, with a DTD or a schema Universal document structuring - allows a (“valid XML document”) linear representation of any structure Implicitly, even in the absence of a DTD Universal modality of exchange of information or schema, starting from the annotation on Internet structure (“well formed document”) Language for federated databases Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 81 Biarritz 82 XML languages XSLT XSLT Transformation of XML files into other XPointer XML, HTML or text files Tree (source) to tree (destination) XLink transformation rules DAML+OIL Example-based programming LOM XSLT programs are XML files User defined Uses XPath language for addressing inside XML documents Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 83 Biarritz 84 14
- 15. XML annotation for learning XSLT purposes <xsl:stylesheet xmlns:xsl="http://www.w3.org/TR/WD-xsl"> <xsl:template match="/"> Universal way of content structuring <html> <body> <h2>List of students</h2> <xsl:apply-templates/> and annotation </body> </html> Reuse of learning modules through the </xsl:template> <xsl:template match="StudentsList"> web <xsl:for-each select="Student"> ID= <xsl:value-of select="ID"/> First name:<xsl:value-of select="FName"/> Name:<xsl:value-of select="Name"/> Year:<xsl:value-of select="Year"/> </xsl:for-each> </xsl:template> </xsl:stylesheet> Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 85 Biarritz 86 Semantic editing E-learning standards IEEE-LTSC - IEEE Learning Technology Standards Committee (LTSC) ARIADNE - Alliance of Remote Instructional Authoring and Distribution Networks for Europe IMS - Global Learning Consortium, Inc. SCORM - Sharable Content Object Reference Model - ADL - Advanced Distributed Learning AICC - Aviation Industry CBT (Computer-Based Training) Committee DC - Dublin Core Metadata Initiative Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 87 Biarritz 88 XML based annotation in Learner Object Metadata E-learning standards XML-based Metadata - LOM (“Learning <?xml version="1.0"?> <lom Object Metadata”) - elementary xmlns="http://www.imsglobal.org/xsd/imsmd_rootv1 p2p1” ...> learning module <general> ... </general> <lifecycle> ... </lifecycle> <metametadata> ... </metametadata> IMS packages of learning modules <technical> ... </technical> <educational> ... </educational> <relation> ... </relation> <annotation> ... </annotation> <classification> ... </classification> </lom> Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 89 Biarritz 90 15
- 16. Learner Object Metadata Learner Object Metadata <educational> <technical> <interactivitytype> <format>text/html</format> <langstring>Expositive</langstring> <location type="URI"> </interactivitytype> http://www.racai.ro/foo/c.html <learningcontext> </location> <langstring>Higher Education</langstring> </technical> </learningcontext> <description> <langstring>Online CoursePack</langstring> </description> </educational> Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 91 Biarritz 92 Learner Object Metadata <relation> <kind> Web page generation <langstring>Requires</langstring> </kind> <resource> <description> Structuring <langstring>Description of resource</langstring> </description> </resource> </relation> Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 93 Biarritz 94 Web page generation Structuring Content Linear Structuring Hierarchy Styling Network Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 95 Biarritz 96 16
- 17. Structuring Generate web pages Usually, learning systems on the web Adaptable – with usual browsers generate a linear, “tutorial” order, e.g. Adaptive – (Brusilovsky-AH) ELM-ART DCG, APHID, ELM-ART, ID Generated for a group, with adaptable features Simple hierarchical links -lessons, (reorder links, show/hide links, map adaptation) sections, subsections, and terminal Customization vs. optimization pages ELM-ART II Personalized (individualized) – DCG, APHID, Very simple network links – index, Larflast glossary, references Generated for a single person Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 97 Biarritz 98 Scope of generation Generation horizon Generate an entire site Local – satisfy “requires” links Generate page by page Holistic - Larflast Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 99 Biarritz 100 Goal of generation Generation procedure Convert printed to electronic textbooks, Personalized generation is achieved by e.g. ELM-ART filtering the conceptual structure Sequencing of modules – starting from (semantic network, domain ontology) a student model and relations among according to the learner model (known learning modules, e.g. DCG or unknown concepts) or to the Glossary, index, and references links abstraction level (e.g. ID) Hypertext links – using NLP techniques Planning – AND/OR graph (DCG), Bayes Believe Net – APHID Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 101 Biarritz 102 17
- 18. GenWeb (Trausan-Matu, PEDAGOGICAL KNOWLEDGE Domain knowl. acquisition 1997) Test generation DOMAIN Centered around a domain knowledge base Student (ontology) Eval. KNOWLEDGE BASE Adapts lesson planning according to different Rev.eng. of predefined student personalities stud. programs Generates simple explanations in natural language Explanation Generates automatically multiple answers tests generation STUDENT MODEL (knowledge about the user) Evaluates students results for tests, and develop a student’s model RETHORICAL Understands (reverse engineering) student programs KNOWLEDGE Generates a highly structured collection of web pages HYPERTEXT LINGUISTIC GENERATION KNOWLEDGE FOR WWW Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 103 Biarritz 104 LARFLAST LARFLAST LeARning Foreign Language Scientific Terminology COPERNICUS EU project Browsing a holistic, understandable structure may induce a flow state • Leeds University – UK, • Manchester University - UK, Adaptation of the content of the generated • Montpellier University - France, web pages to the incoming information from • RACAI – Romania, • Sofia University - Bulgaria, the web. New information is extracted, • Sinferopol University - Ukraine annotated and coherently integrated in the body of knowledge in order to keep the Objective: To provide a set of tools, available on the web, for supporting the learning of foreign terminology in finance holistic character of the body of knowledge. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 105 Biarritz 106 Serendipitous information LARFLAST acquisition (Cerri & Maraschi) Dynamic generation of personalized web pages Runs from an Apache servlet Adapts to the learner’s model, transferred from another web site Parameterized, easy to configure for new patterns of web pages and structures Includes relevant metaphors and texts from a corpus Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 107 Biarritz 108 18
- 19. Semantic editing (Trausan) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 109 Biarritz 110 Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 111 Biarritz 112 Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 113 Biarritz 114 19
- 20. Web page generation Styling Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 115 Biarritz 116 Web page generation Styling Content Different presentation attributes (color, Structuring shape, highlighting, background etc.) Styling Correspond to user’s preferences Performed Declaratively – CSS, XSLT Procedural – JavaScript, Java Client vs. server (ASP, JSP, XSP, PHP) Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 117 Biarritz 118 References References P. De Bra, P. Brusilovsky, G. Housen, Adaptive Hypermedia: From Kettel, Thomson, Greer, Generating Individualized Hypermedoia Systems to Framework, ACM Computing Surveys 31(4) 1999. Apploications, Procs. Of the Int. Workshop on Adaptive and Intelligent Clibbon, K., Conceptually Adapted Hypertext For Learning, Proceedings Web-based Educational Systems, Montrel, Canada, 2000, pp. 37-49 of CHI’95, (APHID) http://www.acm.org/sigchi/chi95/Electronic/documnts/kc_bdy.html Sickmann and all, Adaptive Course Generation, Procs. Of the Int. Dimitrova, V., Self, J., Brna, P., 'Maintaining a Joinly Constrcted Workshop on Adaptive and Intelligent Web-based Educational Systems, Student Model', in S.A.Cerri (ed.), Artificial Intelligence, Methodology, Montrel, Canada, 2000 , pp. 73-84, (ID) Systems, Applications 2000, Springer-Verlag, ISBN 3-540-41044-9, Nelson, T.H., The Heart of Connection: Hypermedia Unified by pp.221-231. Transclusion, Communications of the ACM, vol.38, no. 8, pp. 31-33, Engelbart, D.C., Toward Augmenting the Human Intellect and Boosting aug. 1995. our Collective IQ, Communications of the ACM, vol.38, no. 8, pp. 30- Thuering, M., Hannemann, J., Haake, J.M., What’s Eliza doing in the 33,aug. 1995. Chinese Room? Incoherent Hyperdocuments - and how to avoid them, Gruber, T., What is an Ontology, Hypertext'91, San Antonio, 1991, pp. 161-177. http://www.ksl.stanford.edu/kst/what-is-an-ontology.html Thuering, M., Hannemann, J., Haake, J.M., Hypermedia and Cognition: Designing for Comprehension, Communications of the ACM, vol.38, no.8, pp. 57-66, aug. 1995. Stefan Trausan-Matu, ITS 2002, Stefan Trausan-Matu, ITS 2002, Biarritz 119 Biarritz 120 20
- 21. References Trausan-Matu, St. (1997) 'Knowledge-Based, Automatic Generation of Educational Web Pages', in Proceedings of Internet as a Vehicle for Teaching Workshop, Ilieni, June 1997, pp.141-148, See also http://rilw.emp.paed.uni-muenchen.de/99/papers/Trausan.html Trausan-Matu, St. (2000) 'Metaphor Processing for Learning Terminology on the Web', in S.A.Cerri (ed.), Artificial Intelligence, Methodology, Systems, Applications 2000, Springer-Verlag, ISBN 3- 540-41044-9, pp.232-241. Gerhard Weber and Marcus Specht, User Modeling and Adaptive Navigation Support, in WWW-based Tutoring Systems, http://www.psychologie.uni-trier.de:8000/projects/ELM/Papers/UM97- WEBER.html - (ELM-ART) J. Vassilieva, http://julita.usask.ca/homepage/AIED'97.ps - (DCG) Louis Weitzman, Kent Wittenburg, Grammar-Based Articulation for Multimedia Document Design, Multimedia Systems CACM (1996) 4, pp. 99-111 Stefan Trausan-Matu, ITS 2002, Biarritz 121 21