Research Of Academic Paper Recommendation System Based On Ugc Of ResearchGate

With the advent of the Internet and the era of mobile Internet,the producer of information has changed from the earliest expert to the whole Internet user.Human access to information also evolved from the earliest portal sites to the use of search engines.However,with the increasing of human informatization,people's need for information has changed from the earliest commonality to the present personalization.In view of the research and application of the recommender system,it has received a lot of attention in academic circles and industry.More mature recommendation systems have been applied to news content,video,music songs and e-commerce products.Correspondingly,relevant researches on recommendation system are increasing at home and abroad.At present,however,there is less research on the recommendation system of academic paper,especially on the basis of social network.As is known to all,the academic achievements of scientific research are growing rapidly with the development of scientific research.Researchers are getting interested in its scientific research achievements,mainly rely on in the different database or search engine to find,or the need to periodically browse relevant academic journals in the field of or related to your own website,in order for the current academic learning about the latest scientific research dynamic.But such browsing and finding work is time-consuming and cumbersome.Researchers sometimes get lost in the thousands of scientific achievements.Considering the increasing number of interdisciplinary studies in scientific research,researchers often have no prior knowledge in acquiring unfamiliar academic fields.Therefore,it is necessary to use the recommender system to provide information service for researchers.At present,there is not much research on the related recommendation system of academic achievements.Most of the published academic papers or research findings are based on the analysis of the content of the scientific research.For example,using LDA(document topic generation model)to generate document theme for each scientific research result,and reuse the similarity between themes to carry out the recommendation of academic achievements.In addition,some researches are based on the external attributes of scientific research results,such as the quality of scientific research achievements and the citation network of scientific research results.The related research is the main use of prior knowledge on the academic achievements has been formed,and did not consider the timeliness of scientific research as well as for academic researchers interested in converting considerations.In this study,proposed the use of the current popular research social platform ResearchGate(the door)of scientific research,scientific research personnel and the attention of researchers published information on academic topic and knowledge skills,thus to form a scientific research personnel academic interest model.At the same time,this study also studies the academic citation of researchers.By researchers published a paper in a longer time frame as well as relevant academic references behavior research,found the researchers cited in academic achievements,its cited frequency of logarithm of the author can be a good fit.The author's key words in his long-term research results also reflect his academic direction and field.Most of these authors can also be found in the content of their ResearchGate social platform.These are the prerequisites for building a recommendation system based on user-generated content on ResearchGate's social networking platform.At the end of this study,the recommendation system of the academic paper was tested.The system mainly realizes the acquisition,storage and cleaning of user-generated content on the relevant academic social platform and personalized recommendation based on scientific research personnel.Finally,this study based on the review of the recent recommendation system common indicators,according to generate content based on the scientific research personnel and its generation with its focus on user generated content fusion recommended by the two different information sources for scientific research performance related tests have been carried out.The test proves the effectiveness of the proposed system.The academic contribution of this research mainly lies in the following aspects:This research uses the content generated by the users in the research social platform to establish the academic interest of researchers.Similar to traditional social media platforms,research social platforms can provide more diverse and rich information for the recommendation system.Moreover,the cold start problem of recommender system can be solved well.This study uses the key words of the author to construct the lexical dictionary of scientific research.This has largely solved the problem that the content of the social platform of scientific research is biased,and the content of scientific research social platform and academic achievements are relatively low.This paper probes into the methods of generating content on social platform of scientific research.At present,ResearchGate social platform does not open its data interface,and data acquisition is difficult.Access to these data can provide a variety of services for the agency's researchers.