Searching Documents with Semantically Related Keyphrases

In this paper, we present a tool, called SemKPSearch, for searching documents by a query keyphrase and keyphrases that are semantically related with that query keyphrase. By relating keyphrases semantically, we aim to provide users an extended search and browsing capability over a document collection and to increase the number of related results returned for a keyphrase query. Keyphrases provide a brief summary of the content of documents, and they can be either author assigned or automatically extracted from the documents. SemKPSearch uses a set of keyphrase indexes called SemKPIndex, and they are generated from the keyphrases of documents. In addition to a keyphrase-todocument index, SemKPIndex also contains a keyphrase-tokeyphrase index which stores semantic relation scores between the keyphrases in a document collection. The semantic relation score between keyphrases is calculated using a metric which considers the similarity score between words of the keyphrases, and the semantic similarity score between two words is determined with the help of two word-to-word semantic similarity metrics based on WordNet. SemKPSearch is evaluated by human evaluators, and the evaluation results showed that the evaluators found the documents retrieved with SemKPSearch more related to query terms than the documents retrieved with a search engine.