Squiggle Med: Semantic Search for Medical Digital Library

I. INTRODUCTION Searching everything everywhere is becoming our habit when we need to find something. However, finding what we need is often a hard job. Current search engine technology is very good in finding complete Web pages, but it lacks the desired precision and recall when searching for multilingual resources such as within pan-european Medical Digital Libraries. For instance, searching " Diagnosi preventiva del cancro al seno " (which in Italian means " early diagnosis of breast cancer ") in a syntactic search engine results in a subset of the relevant document, those written in Italian. Squiggle [1] is an extensible semantic search framework for the development of semantic search engines. By adding a conceptual flavour to the crawling and the indexing of resources, Squiggle can exploit ontological elements to improve and enrich searching time, without undermining the user experience. These features, together with the employment of SKOS [2] model, make Squiggle a powerful and reusable framework to build engines with both syntactic and semantic functionalities. II. SEARCHING WITH SQUIGGLE MED The interaction of a final user with Squiggle is intuitive and very similar to the use of a traditional search engine; however, the results are better and more meaningful. In the following we provide examples of searches with Squiggle Med (a demonstrative semantic search engine for medical digital library) that uses the SKOS version of MeSH [3] 1 st step: Syntactic search and query analysis The user is presented with a simple search form in which he can insert some keywords. Squiggle, like syntactic search engines, can immediately retrieve all results containing those keywords. For example, a General Practitioner could search for " Diagnosi preventiva del cancro al seno " ; and the user not only obtains syntactically-matching results (documents containing the searched keywords), but his query is also analyzed in order to identify its meaning. This is possible because of Squiggle disambiguation capabilities: the search engine can access an ontology of the domain (i.e., the SKOS version of MeSH) and try to identify the concepts that could have some connections with the query (in the previous example: early diagnosis, breast neoplasm, primary prevention, breast diseases). 2 nd step: Semantic search The results of the previous disambiguation phase are displayed, together with the syntactic results, in a lateral box under the heading " Did you mean...? ". Therefore, the user can manually choose the meaning of his query among …