Research On The Portrait Of Scholars Based On Multi-source Information Fusion And Its Applications

Since the beginning of this century,advances in science and technology has greatly improved the quality of human life.In an era of rapid technological innovation,information technology,intelligent manufacturing,and biological science have been formed promote the development of productivity forces and social transformation.The rapid development of science and technology cannot leave hundreds of thousands of research scholars and cultivation.Therefore,to research the key group of research scholars can exert profound impact on the development of scientific and technological innovation.With the continuous iteration of Internet technology,the user portrait technology based on personal data model has been rapidly promoted.Using the multi-dimensional tagging form,user research and application support can be more intuitive and convenient.Therefore,the user portrait technology will be combined to construct and research the portrait of scientific researchers in this thesis.Firstly,the basic database is established by using the academic data resources collected on the Internet,which serves as the information basis for the construction and application of portrait.In the traditional academic research on the scholars,different data sources and structures have a direct impact on the results.Therefore,information fusion of multi-source dataset is necessary to ensure the quantity and quality of data.Secondly,after deeply understanding the principle of user portrait,we try to design the portrait model of scientific researchers reasonably by means of statistical analysis and data mining.The main contents of the portrait are divided into the basic dimension and the advanced dimension: the basic one includes the demographic dimension and the professional academic dimension,including the personal attribute information and the academic information of scholars,and using the information retrieval algorithm to expand the dimension.The advanced one mainly represents the individual characteristics of scholars,including their research interests and interest evolution.Finally,the portrait application research from multiple aspects.In the paper recommendation method,by considering the importance of the paper and scholars' research interest,we recommend high-quality papers for scholars.To mine the potential cooperative relationship among scholars,we use the link prediction method to control the weight of network features and interest features,so as to select a more accuratescheme and find out the potential cooperation among scholars.In addition,we use portrait tag shows the part of the academic index of the distribution characteristics,and analyzes the academic cooperation,specific include cooperation within the same area,colleges cooperation and global cooperation.By those means of data analysis,the differences in scientific research level and the development trend of science and technology are explained objectively and reasonably.