Deriving dynamic knowledge from academic social tagging data: A novel research direction

Dong, Hang; Wang, Wei; Coenen Frans

Permalink

https://hdl.handle.net/2142/96693 Copy

Description

Title

Deriving dynamic knowledge from academic social tagging data: A novel research direction

Author(s)

• Dong, Hang
• Wang, Wei
• Coenen Frans

Issue Date

2017

Keyword(s)

• Academic social tagging
• Data cleaning
• Ontology learning
• Concept extraction
• Knowledge evolution

Abstract

Academic social tagging is an important activity in the age of Web 2.0 (and Science 2.0) whereby researchers collaborate online to organize academic resources. Compared to general social tagging, academic social tagging has a more complex nature in terms of semantics and sparsity of the data. It is worth exploring the knowledge structure hidden in the academic social tags and to see how they reflect the evolution of scientific knowledge. This poster presents a research direction comprised of four phases: (i) data cleaning, (ii) concept extraction, (iii) relation learning and (iv) knowledge evolution of academic social tags. For the data cleaning phase, a workflow is presented and evaluated using the Bibsonomy dataset. Future studies will focus on cluster-based outlier detection and topic modeling to extract concepts and derive relations from academic social tags; chronological analysis will be conducted to discover the dynamics of knowledge structure reflected in academic social tags.

Publisher

iSchools

Series/Report Name or Number

iConference 2017 Proceedings

Type of Resource

text

Language

en

Permalink

http://hdl.handle.net/2142/96693

Copyright and License Information

Copyright 2017 Hang Dong, Wei Wang, and Coenen Frans

Owning Collections