Recognizing People, Places, Conferences and Events in the Web 2.0
Thesis type |
|
---|---|
Status | Finished |
Submitted in | 2010 |
Presentation on | 27. Apr 2010 16:30 |
Presentation room | Seminarraum I5 |
Add presentation to calendar |
![]() ![]() |
Supervisor(s) | |
Advisor(s) |
|
Web2.0 consists of numerous user-friendly feautures that aim
to enhance information sharing and collaboration among users. These concepts are
realized by social network sites, folksonomies, wikis, etc. In spite of the
Web2.0 developing and usage transperency, there are neither standars nor
structure that could be used by web applications. The information systems
need the cooperation between human and machine intellegence to reach a new step
of Web evolution.
We apply machine intellegence concept to extract semantical data of handled
objects, in our case publicly available information in Web 2.0 like wikis,
blogs, podcasts, mailing lists etc. One of the services that can be used to
achieve our aim is
Open Calais, an attempt of Reuters Press Agency to create an appropriate
Semantic Web API for Web2.0. The application of the API solves a lot of Web 2.0
flaws like automatic tagging based on semantics, identity management,
information compatibility. It gives the World Wide Web community a great utility
hopefully leveraging the interoperability of information systems. The API is
successfully used for WordPress blog,
in Gnosis Firefox plug-in and
many others.
The master/diploma thesis will focus on applying of Semantic Web technologies in the analysis of publicly available media. This analysis should be combined with existing multimedia analysis means already implemented using standards like MPEG-7 and MPEG-21. The task is to create an information system where you have apply the API on existing sources like the PROLEARN MediaBase or the GRAECULUS MediaBase to add meaningful semantics to largely unstructured collections of text and multimedia. In a second step, the enriched information should be connected to existing social network analysis data gained from a parallel analysis of the sources.
The applicant should be open for interdisciplinary work. She/he should be experienced in databases, XML and Java or the other object-oriented programming language and has a good mathematic background. The applicant should also be a good team player to cooperate with colleagues. Informatik 5 offers an open-minded, congenial research atmosphere within different national and international research projects like the excellence cluster Ultra High-Speed Information and Communication (UMIC), the collaborative research centre "Media and cultural communication" and the Cairo University E-Learning Centre (CUELC).