Research On The Application Of The Association Rules In Novelty Search

Novelty Search is a form of advisory services ,that used the manual refers and the computer search , integrated use of the methods of analysis and com-parison, the legislation that for the evaluation of scientific research, results, pat-ents, inventions and other innovative, advanced and practical literature provide-d information based on, is an intelligence services for the management and scientific research provide policy-making basis .Data mining is tools for help people found information and knowledge in mass data.In recent years,data mining techniques has become the core in busine-ss intelligence technology ,has been widely applied to various fields. the paper mainly use the Association Rules in Novelty Search, analyse the Correlation o-f the investigation gathered by the new electronic documents ,tapping users ge-nuinely interested rules, and ultimately draw more real, scientific results.Thesis has understanded of the specific processes of the Novelty Search in-dept,Mastered the specific methods that comprehensive analyse the literature;Focus on the association rules mining algorithm,and using the Apriori algorithm in new search analysis can be realized.A major work in the following areas:①In many network database,CNKI network database has be selected as the experimental database objects.Look on "Data mining" as keywords to collect some related literature, Collecting some of the relevant literature and and preserve the part of indexing of the literature data as Excel.②The collected data through simple processing documents saved to Excel table,Further into the SQL database,has the data cleansing,integration,and transformation,Replace the text data for character data,after pretreatment,the data change into the form can be carried out.③Design and Construction the overall structure of the new system ,The inquiry focuses on the various subsystems of functional modules.Apriori algorithm will be applied to the data centralized association mining,Construction overall structure of the new system.According to a given minimum support (min_sup),finded all the frequent itemsetsin question in the database,according to the smallest setting Confidence (min_conf),produced strong association rules by frequent itemsets .Finally, the paper will be linked through the algorithm used for the analysis, proved to Novelty Search algorithm in the field of data analysis work has certain practicality.