Intelligent information extraction from scholarly document databases

Extracting knowledge from big document databases has long been a challenge.Most researchers do a literature review and manage their document databases with tools thatjust provide a bibliography and when retrieving information (a list of concepts and ideas), thereis a severe lack of functionality. Researchers do need to extract specific information from theirscholarly document databases depending on their predefined breakdown structure. Thosedatabases usually contain a few hundred documents, information requirements are distinct ineach research project, and technique algorithms are not always the answer. As most retrievingand information extraction algorithms require manual training, supervision, and tuning, itcould be shorter and more efficient to do it by hand and dedicate time and effort to perform aneffective semantic search list definition that is the key to obtain the desired results. A robustrelative importance index definition is the final step to obtain a ranked importance concept listthat will be helpful both to measure trends and to find a quick path to the most appropriatepaper in each case.