A Semantic Approach to Enhance HITS Algorithm for Extracting Associated Concepts Using ConceptNet

Common sense knowledge base creation and usage is an active field of research, many researches in this field are trying to invent new methods to make machines more intelligent. ConceptNet is one of the most popular freely available common sense knowledge bases with millions of common concepts and assertions. In this paper, we propose a novel algorithm called SMHITS to extract associated concepts which have conceptual connections with a group of input concepts using ConceptNet knowledge base. We modify HITS algorithm to take into account the semantic meaning of concepts in the context by defining weights for concepts. We also eliminate the domination of the famous concepts that have huge number of predecessors and successors by using weights for hub and authority values. Evaluation shows the superiority of SMHITS over current state of the art methods, with about 10% improvement in performance comparing with AnalogySpace which is the current ConceptNet association method. Evaluation also confirms that this improvement is statistically significant. We use two versions of ConceptNet for evaluation, results show that SMHITS method performance will remarkably be increased with the coming on versions of ConceptNet.

[1]  David Baxter,et al.  On the Application of the Cyc Ontology to Word Sense Disambiguation , 2006, FLAIRS.

[2]  Hugo Liu,et al.  ConceptNet — A Practical Commonsense Reasoning Tool-Kit , 2004 .

[3]  Douglas B. Lenat,et al.  CYC: a large-scale investment in knowledge infrastructure , 1995, CACM.

[4]  Vijayan Sugumaran,et al.  Usability of upper level ontologies: The case of ResearchCyc , 2010, Data Knowl. Eng..

[5]  Vijayan Sugumaran,et al.  Using Semantic Knowledge to Improve Web Query Processing , 2006, NLDB.

[6]  Chihli Hung,et al.  A Personalized Word of Mouth Recommender Model , 2008, Webology.

[7]  Weibo Li,et al.  Optimization of WEB Data Collection Technology Based on HITS Algorithm , 2013, 2013 International Conference on Computer Sciences and Applications.

[8]  Henry Lieberman,et al.  Digital Intuition: Applying Common Sense Using Dimensionality Reduction , 2009, IEEE Intelligent Systems.

[9]  Ramanathan V. Guha,et al.  CYC: A Midterm Report , 1990, AI Mag..

[10]  Hong Yu,et al.  An Improved Weighted HITS Algorithm Based on Similarity andPopularity , 2007 .

[11]  Mehdi Keshavarz,et al.  Ontology matching by using ConceptNet , 2012 .

[12]  Henry Lieberman,et al.  AnalogySpace: Reducing the Dimensionality of Common Sense Knowledge , 2008, AAAI.

[13]  Junpeng Chen,et al.  Combining ConceptNet and WordNet for Word Sense Disambiguation , 2011, IJCNLP.

[14]  Catherine Havasi,et al.  ConceptNet 5: A Large Semantic Network for Relational Knowledge , 2013, The People's Web Meets NLP.

[15]  Franco Zambonelli,et al.  Bridging vision and commonsense for multimodal situation recognition in pervasive systems , 2012, 2012 IEEE International Conference on Pervasive Computing and Communications.