Skip to main navigation Skip to search Skip to main content

Generating hypothesis: Using global and local features in graph to discover new knowledge from medical literature

  • Vishrawas Gopalakrishnan
  • , Kishlay Jha
  • , Aidong Zhang
  • , Wei Jin
  • SUNY Buffalo
  • North Dakota State University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

13 Scopus citations

Abstract

Literature-based discovery, a method by which one estimates a plausible relationship between hitherto unconnected or unrelated terms is an important component of biomedical text mining. In this paper, we propose a new method which creates a graph-based knowledge base and formulates the problem as one of discovering and ranking paths in a graph. The proposed approach is contextually sensitive, while at the same time mindful of global characteristics. The proposed method, unlike many existing techniques, is not limited to immediate neighbor search, nor does it require any user intervention. At the same time it has the ability to discover meaningful and specific "bridge" terms with high precision. In some cases, it has also identified relationships well in advance when compared to popular publications investigating the corresponding discoveries.

Original languageEnglish
Title of host publicationProceedings of the 8th International Conference on Bioinformatics and Computational Biology, BICOB 2016
EditorsNurit Haspel, Thomas Ioerger
PublisherThe International Society for Computers and Their Applications (ISCA)
Pages23-30
Number of pages8
ISBN (Electronic)9781943436033
StatePublished - 2016
Event8th International Conference on Bioinformatics and Computational Biology, BICOB 2016 - Las Vegas, United States
Duration: Apr 4 2016Apr 6 2016

Publication series

NameProceedings of the 8th International Conference on Bioinformatics and Computational Biology, BICOB 2016

Conference

Conference8th International Conference on Bioinformatics and Computational Biology, BICOB 2016
Country/TerritoryUnited States
CityLas Vegas
Period04/4/1604/6/16

Fingerprint

Dive into the research topics of 'Generating hypothesis: Using global and local features in graph to discover new knowledge from medical literature'. Together they form a unique fingerprint.

Cite this