Knarig Arabshian

"Many of life's failures are people who did not realize how close they were to success when they gave up.
~ Thomas A. Edison

A Framework for Personalized Context-aware Search of Ontology-based Tagged Data

The second project uses the general concept of the Personal Internet developed within the Personal Content Manager (PeCMan) to build a context-aware personalized search environment for a user. The World Wide Web is advancing towards greater personalization. Users have personal accounts on different web services and can access their private content or share it with others. Web services such as Flickr or Youtube allow users to store and search for data via keyword tags. However, as more web services are offered on the Web, it is becoming increasingly difficult for users to manage their content online. Currently, tagged data consists of flat keyword searches. The limitation to this type of search is that it does not use the information of the user or data that is being searched. I propose using a tag ontology to leverage information from the metadata of the user and resource, in addition to the tagged keyword, in order to present more meaningful search results to the user. Metadata about the user can include a user.s profile, demographic, or context; whereas metadata about the resource may include the type, size, or context of the document being searched. Additionally, with the use of ontologies, tags can now be shared and reused by different web services. A public repository can hold these tag ontologies and applications can download and make use of them. I.ve designed the following: 1) a tag ontology that includes context information of the user and data being tagged 2) a framework that performs context-aware searching of ontology-based tagged data by using this tag ontology. I.m in the process of implementing a prototype of this and hope to show that the use of tag ontologies provide more meaningful results within a user.s context compared to flat tag searches.