Please use this identifier to cite or link to this item: https://doi.org/10.1007/978-3-642-29038-1_31
Title: Assessing web article quality by harnessing collective intelligence
Authors: Han, J.
Chen, X.
Chen, K.
Jiang, D. 
Issue Date: 2012
Source: Han, J.,Chen, X.,Chen, K.,Jiang, D. (2012). Assessing web article quality by harnessing collective intelligence. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 7238 LNCS (PART 1) : 428-439. ScholarBank@NUS Repository. https://doi.org/10.1007/978-3-642-29038-1_31
Abstract: Existing approaches assess web article's quality mainly based on syntax, but seldom work is given on how to quantify its quality based on semantics. In this paper we propose a novel Semantic Quality Assessment(SQA) approach to automatically determine data quality in terms of two most important quality dimensions, namely accuracy and completeness. First, alternative context with respect to source article is built by collecting alternative web articles. Second, each alternative article is transformed and represented by semantic corpus and dimension baselines are synthetically generated from these semantic corpora. Finally, quality dimension of source article is determined by comparing its semantic corpus with dimension baseline. Our approach is promising way to assess web article quality by exploiting available collective knowledge. Experiments show that our approach performs well. © 2012 Springer-Verlag.
Source Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
URI: http://scholarbank.nus.edu.sg/handle/10635/41254
ISBN: 9783642290374
ISSN: 03029743
DOI: 10.1007/978-3-642-29038-1_31
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

Page view(s)

58
checked on Dec 16, 2017

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.