Academic paper search sorting method based on document scoring model and relevancy

摘要: The invention discloses an academic paper search sorting method based a document scoring model and relevancy. comprises the steps that preprocessing is conducted on data in library, basic information of including year publication, quality published journals, number times citation, obtained, meantime being clicked viewed recorded used as feedback value; established for by three factors time, quality, feedback; overall score each search-term related collection calculated combining term, relevancy model; results according to sorted are returned. takes into account multiple aspects whole process, built papers, deficiency handling multi-term engine overcome, so accuracy experience improved greatly.