Experiences in Information Mining from a Legacy CMS
1 like207 views
The document discusses experiences in information mining from a legacy content management system (CMS) using D2RQ mapping based on business ontology. It outlines the motivation to improve user experience and data visibility, the challenges posed by the CMS's structure, and the strategic actions taken for ontology engineering and mapping adaptation. The ongoing development aims to extend capabilities and enable access to additional systems while supporting the transferability of the ontology.
Experiences in Information Mining from a Legacy CMS
- 1. Experiences in Information Mining from a Legacy CMS via D2RQ Mapping Based on Business Ontology Presentation at VMBO 2014 Workshop Malte Hahnenwald, Vera G. Meister March 2014
- 2. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Motivation What do we want to achieve? Improve User experience Usability of data Visibility of Information Mine data from different sources Show feasibility Develop transferability 05.06.2014 2 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 3. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Initial Situation Poor onsite search • Specific search plug-in • No global search SixCMS • Not normalized RDB • Encrypted structure • Adjourned re-launch No consistent knowledge organization system • Each department acts self-dependent 05.06.2014 3 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 4. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Course of Action Ontology engineering Deep structure research Mapping adaptation Technology decisions Application implementation Concept adjustment Enabling transferability 05.06.2014 4 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 5. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Ontology Engineering 05.06.2014 5 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister University Faculty Person Role Module Compulsory Module Optional Module StudyCourse ModuleGroup Bachelor Course Master Course hasSubUnit Room PracticeRoom LectureHall Office Lab Event Researcher Full_Professor Staff Contract_Lecturer Student Admin_Staff Tutorial Exam Seminar Lecture hasRole memberOf organizes performs contains hasPart hasRoom contains offers isLocatedIn role enrolledIn
- 6. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Reuse of Ontology Patterns 05.06.2014 6 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister Description Standard Ontology Concepts / Relations University as an organization org: <http://www.w3.org/ns/org#> Role / role Staff contact information foaf: <http://xmlns.com/foaf/0.1/> Person / familyName, givenName, title, phone, mbox, homepage, interest Actors, events and outcome of scientific work iswc: <http://annotation.semantic web.org/iswc/iswc.daml#> Event (Conference, Tutorial, Workshop), Full_Professor, Associate_Professor, etc.
- 7. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page D2RQ Mapping Adaptation SixCMS structure is sparse and encrypted • not normalized • 70 tables without foreign keys • source data in one table • relations in proprietary ID’s No automated mapping possible Specific mapping features needed 05.06.2014 7 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 8. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page D2RQ Mapping Excerpt ########## Class -> Module ########## map:Module a d2rq:ClassMap; d2rq:dataStorage map:sixcms; d2rq:condition "sixcms_article.area_id = 1105"; d2rq:class :Module; d2rq:uriPattern "Module/@@sixcms_article.title|urlify@@". map:moduleID a d2rq:PropertyBridge; d2rq:belongsToClassMap map:Module; d2rq:property :moduleID; d2rq:column "sixcms_article_data.value"; d2rq:condition "sixcms_article.area_id = 1105"; d2rq:condition "sixcms_article_data.fieldname = 'lv_nr' "; d2rq:join "sixcms_article.id <= sixcms_article_data.article_id"; d2rq:join "sixcms_article_data.area_id => sixcms_article.area_id"; d2rq:propertyDefinitionLabel "Module Number". 05.06.2014 8 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 9. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Application Architecture 05.06.2014 9 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister RDB D2RQ Engine - Mapping D2R Server SPARQL Ontology JenaRDF Faceted browsing
- 10. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page User Interface 05.06.2014 10 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister
- 11. University of Applied Sciences Brandenburg a. d. H. · Department of Business and Management Page Ongoing Development Skim more from Six CMS Extend to other systems Enable search engine access Streamline the business ontology Support transferability 05.06.2014 11 Experiences in Information Mining from a Legacy CMS ∙ Hahnenwald / Meister