Knowledge acquisition for the SEASALT apprentice agent using twitter feeds

Seneviratne, Chathuri Nilushika, Sauer, Christian and Roth-Berghofer, Thomas (2013) Knowledge acquisition for the SEASALT apprentice agent using twitter feeds. In: The 18th UKCBR 2013 Workshop, 10 December 2013, Cambridge, UK.

Full text not available from this repository.

Abstract

Web 2.0 content, including blogs, forum posts and tweets, is mostly expressed in an unsystematic manner. Due to this reason, retrieving and reusing this content has become challenging. As a solution, Reichle et al. presented a novel architecture named SEASALT. A main feature of the SEASALT architecture is the use of topic agents implemented as topic specific Case-based reasoning (CBR) systems. Another key component of the SEASALT architecture is the Apprentice Agent which supports a knowledge engineer in the SEASALT Architecture by automatically extracting vocabulary items and taxonomic similarity measures for the CBR-based topic agents from a virtual community of experts, comprising the knowledge input for the SEASALT Architecture. A first implementation of such an apprentice agent was presented by Bach et al. with the Knowledge Extraction workbench (KEWo) which extracted vocabulary items and similarity measures form an online community of travel medicine experts. The work presented in this paper extent the KEWo to use Twitter feeds as a knowledge source. A Multi Agent System is developed to acquire Twitter feeds which are then transferred for further knowledge extraction to the Apprentice agent component KEWo. Further Twitter is analysed as a knowledge source in terms of the amount of data it can provide on a specific topic and how this provided amount of tweets has an impact on the performance and quality of knowledge extracted from them. Furthermore, the paper analyses how well the hash tag feature provided in Twitter can be employed as a source of structuring information. As the ultimate output, this paper contributes to the extension of a virtual community within the SEASALT architecture by including the content from Twitter users.

Item Type: Conference or Workshop Item (Paper)
Subjects: Computing
Depositing User: Vani Aul
Date Deposited: 10 Jan 2014 09:57
Last Modified: 21 Oct 2015 13:39
URI: http://repository.uwl.ac.uk/id/eprint/604

Actions (login required)

View Item View Item

Menu