Research And Implementation Of Scientific Research Team Matching Recommendation For Scientific And Technological Requirements Of Enterprises

In China,there are a large number of small and medium-sized enterprises,involving many aspects of technology industry,which are a cornerstone of China's economic development.At present,there are some common questions in small and medium-sized enterprises,such as a shortage of high quality scientific and technological talents,a lack of innovative capacity and so on.And there are not many cases in which excellent research results of many scientific research teams in colleges and universities have been put into practical application,which does not have a significant role in promoting social progress.There exists grievous information asymmetry between small and medium-sized enterprises and scientific research teams in colleges and universities,and the docking progress of innovation chain and industrial chain is not ideal.In view of this phenomenon,this paper constructs the industry classification model of enterprise technical requirements and the recommendation model with scientific research team,so as to realize the purpose of scientific research team recommendation for enterprise technical requirements.Firstly,based on the cyclic convolution network model combined with attention mechanism,this paper completes the industry classification of enterprise technical requirements,so as to narrow the technical scope of text matching between enterprise technical requirements and scientific research team;secondly,under the condition of fusing multiple text features,this paper constructs the text similarity matching model between enterprise technical requirements and scientific research team,extracts the TOPK of similarity results,and complete the recommendation of the research team.Finally,we design and complete a research team recommendation system for enterprise technical requirements,recommend suitable research teams for enterprises,and overcoming the technical barriers faced by enterprises in publishing requirements.This paper's main work is as follows.(1)Text classification of enterprise technical requirements based on RCNN?ATT model.In the era of big data,the text information of SME's technical requirements is complex,and it is more and more hard to distinguish,mine and manage the technical problems.After analyzing the traditional machine learning method to construct text classifier for enterprise technical requirements,the deep learning method is considered to classify the enterprise technical requirement text,attention mechanism is added to cyclic convolution neural network and this paper puts forward a text classification method based on RCNN?ATT model,which makes the text of technical requirements automatically classified according to its industry.The method is tested on the text data sets of four online enterprise demand publishing platforms in China,the results show that the classification performance of this model is better compared with the existing classical neural network model,can narrow the matching range of scientific and technological texts and improve the efficiency of matching calculation.(2)Research on scientific research team recommendation based on multi feature fusion and text matching.In order to solve the problem of information redundancy and feature sparsity in enterprise requirement texts,this paper constructs a text matching model based on multi-level feature fusion,which uses enterprise technology requirement texts to represent technical problems,and takes research direction texts of scientific research teams as technical resource to calculate the similarity between enterprise technology requirement texts and research direction texts of scientific research teams.The content-based recommendation method is used to get the TOPK of scientific research team,automatically recommending research teams for small and medium-sized enterprises to meet their personalized needs.Experimental results show that,compared with single feature text matching model,this model has better recommendation performance.(3)Design and implementation of research team recommendation system for enterprise technical requirements.The platform adopts B/S architecture based on PHP,analyzes the overall framework and module composition of the recommendation system,and designs the processing flow of scientific research team recommendation in the system.Main functional modules are composed of enterprise management,scientific research team management,project management and user communication in system.The platform can effectively analysis and deal with the enterprise technical requirement texts and the research direction texts of scientific research teams published by users,and finally recommending the scientific research teams that meet the requirements of the enterprises.