Research On Intelligent Evaluation Of Academic Papers' Key Factors

Scientific research achievements based on academic papers are an important part of national knowledge innovation,and the scientific evaluation of academic papers is the premise of performance appraise of knowledge innovation.At present,the evaluation of scientific research achievements is mainly carried out anonymously by experts,which is limited by the experts' own academic knowledge level and the scope of the subject field.Besides,the limitations of the evaluation results may cause the missing or delaying of some high-quality scientific research achievements to be published,and some other achievements with relatively low academic value to be published in important academic journals.This will bring serious consequences to the growing of scientific research and negative effects on the development of national innovation capacity.Nowadays,the academia research fraud and other misconduct phenomenon has appeared in academic circles frequently,which has given birth to the urgent need in the whole society for the evaluation of scientific research achievements,especially academic papers.Therefore,based on the theory and technology such as artificial intelligence,our study combined the academic paper evaluation theory and practice,building intelligent evaluation key indicators of the academic papers,in order to make the evaluation changing from mainly relying on artificial evaluation to intelligent evaluation,overcoming the shortage of purely relying on anonymous evaluation of experts.On the basis of combing and analyzing the current research related to academic paper evaluation,our paper will conduct in-depth research from the following five aspects.The first step is that we constructed the key index system of academic paper evaluation.On the basis of depth analysis of academic papers,our paper uses grounded theory method to code and analyze the expert evaluation opinions.By doing the research and comparative analyze of the five main categories,we systematically combs the relationship between the main categories and the evaluation factors of academic papers,and constructs the key indicators of academic paper evaluation into clear structure levels.Moreover,the main component analysis method is used to determine the index weight,which highly improved the key index system of academic paper evaluation.The second is the establishing of an intelligent evaluation framework for academic papers.Based on the traditional evaluation process of academic papers,big data,artificial intelligence and other information technologies are integrated into the process of intelligent evaluation of academic papers.Also the intelligent framework is divided into three main modules: one is the automatic normative inspection module;the other is the expert evaluation system module;and the third one is the processing module of the editorial department.The intelligent evaluation framework is expected to form a practical application process for the academic community that based on the theory and method of intelligent evaluation.While implementing scientific evaluation,it will promote the standard of writing an academic paper,and also improve the writing quality of these papers.Finally,it will promote the better dissemination and exchange of scientific research results.The third one is the proposing of an intelligent evaluation method for the novelty of academic paper topics.The process of identifying and judging the novelty of academic paper topics is constructed in this paper.And then,based on this process,the intelligent evaluation process of topic selection novelty is proposed.The intelligent evaluation of topic selection novelty index is realized by constructing research hotspot map,feature vector analysis and novelty judgment.Finally,the feasibility of this method is verified by empirical research.The fourth,an intelligent evaluation method for the content innovation of academic papers is proposed.Based on the analysis of the academic papers' content,we constructed the ontology model of knowledge element of academic papers,and the rules of knowledge element extraction are put forward.Word2 ve and naive Bayes are used to classify the innovation of theoretical methods of academic papers,and SVM model is used to build the rule base of knowledge element extraction.On the basis of the construction of knowledge Meta base of academic papers,this paper proposed the basic methods of intelligent evaluation of research questions,theories,methods and conclusions of academic papers.We also construct the process of intelligent evaluation of innovation of academic papers.Finally,the feasibility of the methods is verified by the researches and demonstration.Finally,we proposed a scientific and intelligent evaluation method for citation of academic papers.We analyze the citation level and construct the citation evaluation dimension and put forward the idea and steps of intelligent evaluation of academic papers based on citations,including the identification of highly relevant academic papers,the identification of citation types of target articles,etc.By verified the feasibility of the evaluation model and the feasibility of our method,through the optimizing of the research results,we finally verified the effectiveness of our method.Based on the related theories and methods of artificial intelligence,from the research perspective of the intelligent evaluation of the key indicators of academic papers,this paper extends the traditional academic paper which based on the limited peer evaluation mode.It also overcomes the subjective drawbacks of the peer evaluation as the main evaluation mode,and improves the efficiency and quality of the evaluation of academic papers at the same time.At the theoretical level,by establishing the key indicators and intelligent evaluation framework of academic papers,the theory,technology and method of artificial intelligence are integrated into the evaluation system of academic papers,which enriches and improves the theory and method system of intelligent evaluation of academic papers.At the practical level,the intelligent evaluation framework of academic papers is constructed.Through empirical research,we discussed the intelligent evaluation process of three main indicators: topic novelty,content innovation and citation scientific.This research has formed a paper intelligent evaluation process for the practical application of the academic community,promoted the standardized writing of academic papers,promoted the better dissemination and exchange of scientific research achievements,improved the quality and efficiency of scientific achievements evaluation,and promoted the development of national scientific and technological innovation achievements evaluation.In the future,an intelligent comprehensive evaluation system will be established to realize the judgment of innovation,scientific,value of academic papers and the writing of automatic evaluation opinions,in order to promote the in-depth application of big data and artificial intelligence technology in the field of academic evaluation...