Information access

The main focus for SICS research on information access is to help humans make sense out of large bodies of text. A first step is often information extraction, which is machine processing of text in order to identify keywords, phrases or other characteristics of bodies of text.

The output from information extraction can be used to look at the frequency with which keywords or phrases occur, in order to deduce something about the text. One such example is attitude analysis where a computer program tries to understand if a text expresses a positive or negative attitude towards a given concept or phrase, e.g. “How many of these 1000 articles mention SICS in a positive way?”. Another use of the output from the information extraction process is to understand something about the relation between different bodies of text, e.g. “List all articles which state that the function of protein X has been experimentally verified”.

Information access also deals with methods that assist humans in searching and assessing large amounts of information. Several such methods that were first studied by the academic community have recently gained broad acceptance on the Internet, e.g. social networks and user-based recommendations. Current research topics related to search and assessment of information include how to determine and present quality and reliability aspects on information.