Title: Cross-domain citation recommendation based on hybrid topic model and co-citation selection

Authors: Supaporn Tantanasiriwong; Sumanta Guha; Paul Janecek; Choochart Haruechaiyasak; Leif Azzopardi

Addresses: Department of Computer Science and Information Management, School of Engineering and Technology, Asian Institute of Technology (AIT), P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand ' Department of Computer Science and Information Management, School of Engineering and Technology, Asian Institute of Technology (AIT), P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand ' Department of Computer Science and Information Management, School of Engineering and Technology, Asian Institute of Technology (AIT), P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand ' Speech and Audio Technology Laboratory (SPT), National Electronics and Computer Technology Center, National Science and Technology Development Agency, Thailand Science Park, Klong Luang, Pathumthani 12120, Thailand ' Department of Computer and Information Sciences, University of Strathclyde, 26 Richmond Street, Glasgow, G1 1XH, UK

Abstract: Cross-domain recommendations are of growing importance in the research community. An application of particular interest is to recommend a set of relevant research papers as citations for a given patent. This paper proposes an approach for cross-domain citation recommendation based on the hybrid topic model and co-citation selection. Using the topic model, relevant terms from documents could be clustered into the same topics. In addition, the co-citation selection technique will help select citations based on a set of highly similar patents. To evaluate the performance, we compared our proposed approach with the traditional baseline approaches using a corpus of patents collected for different technological fields of biotechnology, environmental technology, medical technology and nanotechnology. Experimental results show our cross domain citation recommendation yields a higher performance in predicting relevant publication citations than all baseline approaches.

Keywords: cross domain recommender system; citation recommendation; cross domain citation recommendation; CDCR; topic model; co-citation selection; CCS; information retrieval; keyphrase extraction tool; similarity measures; evaluation; analysis of variance; ANOVA.

DOI: 10.1504/IJDMMM.2017.086566

International Journal of Data Mining, Modelling and Management, 2017 Vol.9 No.3, pp.220 - 236

Received: 19 Feb 2016
Accepted: 14 Oct 2016

Published online: 08 Sep 2017 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article