WikiSeeAlso: Suggesting tangentially related concepts (see also links) for wikipedia articles

Wikipedia is the pervasive knowledge source for widely utilized applications like Google’s Knowledge Graph, IBM’s Watson and Apple’s Siri system. Wikipedia articles contain internal links and See also section links. According to Wikipedia, one of the purposes of See also links is to enable readers to explore tangentially related topics. Currently, Wikipedia relies on human judgments for adding See also links. We attempt to automate the process of See also recommendation by utilizing the aspects of Wikipedia articles like category knowledge, Backlink and the ESA concept vector similarity and external knowledge retrieved by web search engine. Our proposed ensemble based approach combines similarities obtained from these aspects to give a final prediction score. We evaluate our approach on datasets of Wikipedia articles and present our empirical comparison and case studies results with the state-of-the art approaches. We envisage that this work will aid Wikipedia editors and readers to facilitate information search.