[Purpose/Significance] Regarding to the phenomenon that news reports and public opinions of the same event were interdependent and also deviated from each other,the paperexplore the differences and similarities of public opinions triggered by news reports in subject content and time phasing by topic detection and subject association analysis.This research planed to study the resonance and deviation between public events reflected in public opinions and social reality reflected in news reports,which would provide a reference for exploring the communication rule of social public opinions and offer a basis for government to make scientific decisions.[Process/Method] Based on Lasswell's 5W communication model,LDA model and Python,the paper designed research ideas and process,crawled the corpus of news reports and their corresponding comments from Tencent News and Quaro and deeply analyzed the processed corpus.[Results/Conclusion] The research results indicated that more derivative latent subjects would be found in public opinions events which partly diverged from the subjects of social reality;the developing trends of public opinions events conformed with that of social reality;in addition,subjects of public opinions events reflected in social media were more than news media,while the subject content of social reality seemed similar on these two platforms.

Considering the similarities between microblog and traditional literature,this paper aimed to explore the application of h-index,g-index and p-index in the evaluation of microblogs' impact and compared the evaluation results of the three to explore the rationality and characteristics of the three indexes on the evaluation of microblogs' impact.First of all,based on the definitions of h-index,g-index and p-index in the field of information measurement,this study put forward nine indexes for the evaluation of microblogs' impact for forwarding,commenting and liking.Then,it took the campus microblog as the research object,through the overall comparison,correlation analysis and other methods,to compare the evaluation results of microblogs' impact based on h-index,g-index and p-index.The results showed that the three indices had their rationality in the evaluation of microblogs' impact.In terms of discrimination,both the g-index and the p-index were better than the h-index.Among them,the p-index had the most advantages and could find high-impact inactive users.In terms of sensitivity,the h-index was less susceptible to extreme values,which was more suitable for the evaluation of microblogs' impact.

[Purpose/Significance] Social online reviews have more prominence propagation spread and influence than traditional professional reviews.Emotional factors in text comments cannot be fully reflected by mere quantification scores.The measurement and analysis of the emotional factors in text comments can help full angle identify and reveal the online commentary,and reflect the values of online reviews more objectively and accurately.[Process/Method] By extracting the user's text comment data,an algorithm of supervised machine learning was used to calculate the emotion classification score,emotion tendency score and comprehensive emotion score of text comment respectively.The emotion score and total score were cross-comparison analyzed from multiple dimensions of genre,area,number of participants.[Results/Conclusions] The results showed that the emotional factors contained in the text comments had a partial influence on the total score.User's cognitive preferences,social cultural background and the proportion of commentators would affect the usefulness of emotional factors.

This paper discussed the "Internet plus" change of Information Science arising from the challenges,opportunities and power in three dimensions.Firstly this paper introduced the connotation of "Internet plus",followed from the basic theory,research areas and research methods summed up the challenge of three levels of the "Internet plus" brings.Then from the distribution of information,intelligence analysis techniques and intelligence service model introduced Information Science's opportunities which "Internet plus" brings.Finally,it summarized the power of information technology,information needs and information infrastructure for the development of Information Science brought under the "Internet plus" background.

In order to understand the evolution and diffusion of knowledge in the field of mobile library,and guide the related researches in China,the paper scanned the citation network in the mobile library,and understood the citation relationship and citation network path.Using HistCite to detect the citation subgroups of the citation networks of different nodes,it determined the main path of the citation network.With the 3 main path algorithms of Pajek,the difference of main path components of different nodes was analyzed,the core literature set and key knowledge nodes were found.Through the analysis of the core literature and its relationship in the field of mobile library,the main body of knowledge evolution in the field of mobile library was revealed.On the basis of this,the main path of the research on the innovation of mobile library was planned in combination with the domestic research results.

Compare CiteScore Percentile,CiteScore Percentile Quartile which are the new evaluation indexes in the Scopus database with the Journal Impact Factor Percentile,Journal Impact Factor Quartile in Web of Science database.The paper selected four subjects empirical comparison of the above indicators in order to facilitate the in-depth understanding of CiteScore bibliometrics,and for the further application of CiteScore and its extension indexes to provide a reference.

[Purpose/Meaning] With the strategy execution of "Internet+ government Service",the e-government service which takes the Internet as its carrier has been paid much attention to by the society.This is an urgent need for large sample,multi-channel empirical research.[Method/Process] Therefore,based on the government website,WeChat,micro-blog,APP,by using the existing mature evaluation index system,this paper evaluated the city above prefecture-level and provincial e-government services.Then,it selected the administrative level,economic level,regional balance as the differentiation attribute.It was found that both administrative level and economic level were positively correlated with government service ability in all the channels.It showed weakening its adverse effect helps to improve the ability of e-government service.The distribution of channel capacity varied from administrative and economic attribute,city and provincial government level,but it was not restricted by level and region.It provided direction and strength management to balance and promote channel service ability[Results/Conclusion] This paper summed up the revelation of the government in the intensity and direction based on administrative and economic attribute,city and provincial government level,and put forward some suggestions to eliminate the differentiation,learn reference and promotion.

In Main developed countries public digital culture construction system characteristics can be embodied in three aspects.The first aspect was macro guidance system level——regulations,policies and strategy supporting each other,namely relatively perfect cultural laws and regulations system,cultural policy system to ensure strong and the strategic planning of digital culture more comprehensive.The second aspect was the basic guarantee system level——management,fund and personnel collaborative guarantee,that is,the management mechanism was used to complement each other,the cultural capital input source was diverse,the personnel team was relatively standardized and orderly.The third aspect was the construction of digital resources——supply,utilization and evaluation orderly interaction,namely,resources supply subjects were of diversification,the resource utilization mode emphasizes innovation,and the resource service paid attention to performance evaluation.It is helpful to promote the construction of public digital culture in China that foreign experience were discussed and summarized.

Under the big data environment,the sparsity problem of recommendation system project becomes more and more serious.In addition,the traditional single interest representation method is difficult to accurately described,resulting in the reduced accuracy of recommendation result.In view of this,this paper put forward with a kind of multiple interest representation based on recommendation model for library big data knowledge service,by dividing the level of user interest more situations,using ant colony level mining advantage to some target user's nearest neighbor cluster.According to the cluster within the same user rating to forecast the goal of the project not scored,this paper finally implemented parallel processing method for collaborative filtering with the MapReduce data.The experimental results showed that the modeling method generates new multiple item clustering interest tree by hierarchical partition mechanism,enhanced the mining depth of situational interest,and the MapReduced ant colony clustering method also greatly reduced the overall computation time,significantly improved the efficiency of the recommendation system.

The arrival of the era of big data has promoted the rapid development of computer science and technology,and also brought opportunities for the development of scientific and technological information analysis methods.Effective data analysis methods are the basis for acquiring valuable information.Data analysis of science and technology policy refers that it use processing technology of computer to automatically extract concise and representative statements from the scientific and technological policy data,identify the core content of the data.Based on the characteristics of science and technology policy,this paper proposed a data analysis method suitable for the content of science and technology policy,designed and constructed the experimental system about data analysis of science and technology policy,and verified the effectiveness of the method proposed in this paper.It provided a new way for exploring the deep analysis method of scientific and technical information.

Principal component analysis and factor analysis were widely used in science and technology evaluation,however,the selection of evaluation methods was lack of inspection.In this paper,the test framework and test system for the applicability of principal component analysis and factor analysis were established.It had three perspectives:pre evaluation,test in evaluation,and post evaluation test.The pre test includes KMO test,Bartlett test and normal distribution test;Test in evaluation included information loss test of evaluation index;The post test includes the explanatory power test,the representativeness test,the index monotonicity test and the weight rationality test.This paper suggested that the method suitability test should be carried out by principal component analysis and factor analysis;Factor analysis was not generally applicable to science and technology evaluation;Principal component analysis was not suitable for the evaluation with more objects.

[Purpose/Significance] Through dynamic monitoring and abnormal perception of network public opinion data,it made early warning of abnormal network public opinion and provided theoretical models and practicable methods for grasping the preemptive advantage of network public opinion decision.[Method/Process] It analysed the abnormal phenomena of network public opinion data such as skyrocketing data and trend fluctuation under big data environment,made clear the monitoring mechanism of abnormal data,such as trend prediction,dynamic perception of abnormal data,etc.On this basis,the first step was using Gompertz model to predict the trend of network public opinion theoretical interval.Secondly,the deviation degree was defined for data anomaly rating,and the early warning level was determined to realize the timely capture and rapid early warning of abnormal data.[Result/Conclusion] The feasibility of the model was proved by an example,which could provide the measurement basis for the guidance degree of network public opinion,and also provided the algorithm support for compiling intelligent monitoring software of network public opinion.

Micropublication,a representative semantic publishing mode,was introduced and analyzed in this paper.It first introduced micropublication's concepts and micropublication ontology,then it reviewed micropublication's current applications.Finally,it applied the model to represent a psychology document and analyzed the characteristics of micropublication.The research showed that as an argument-based semantic publishing mode,micropublication semantically represented the processes of argumentation that took one document's conclusions as argument,takes statements,data,methods,etc.as evidences,but it could not semantically represent concrete chunks,so it was needed to combine other concept models to realize semantic representation for scientific documents in different degrees.

Nowadays,the introduction of talents in universities has become an important measure to enhance the competitiveness of universities,and the discovery and evaluation of high-level talents is of great significance.In view of the three practical problems,discovery of high level scholars,the analysis and orientation of academic influence of scholars and the comprehensive judgment of academic level of scholars,this paper put forward the methods of screening,analysis and evaluation for high-level overseas scholars,which could be used for reference for the introduction of talents in schools.

This paper aimed at the evaluation of the construction of the library's foreign language electronic resources,took the citation analysis of SCI literature as the breakthrough point,based on the database technology,SQL language,and the C# development of the.NET framework,constructing a directional automatic mining analysis system based on WOS citation analysis.And test ran.The Results were accurate,the time of analysis and the intensity of work were greater than that of DDA or Citespace,to a certain extent.This was a good solution to the problems of professional analysts and lack of funds.It also enabled non professionals to quickly obtain basic analytical data.

[Purpose/Meaning] In the "Internet+" era,cross-border services for the urban library are overwhelming in the development of the library industry,so how to accelerate and take advantage of them to promote China's development of library cause will become an important issue.[Mthod/Process] This paper,from the perspective of cross-border services and industrial convergence,making an analysis on the characteristics of cross-border services for the city library,advanced a structural model for multi-dimensional joint cross-border services system dominated by the city library and designed its operating model.[Consequence/Conclusion] It was found that not only interaction between all the elements within the cross-border services system but also their normal operating of growth mechanism,sharing mechanism,competition mechanism and coordination mechanism could improve multi-dimensional joint cross-border services capabilities for the city library.

Resources discovery system is an important construction object of college library,which is obviously demanded by users,and there is hidden difference in these demands,the relevant analysis can provide reference for the construction,utilization and future development of resources discovery system.The researcher investigated the study situation by Sitespace.The researcher analyzed the resources security acquisition,related knowledge and sharing,objective knowledge analysis and other user demands based on the Maslow theory,selected typical users as subjects,obtained test score data by the user experience of task operation and Liket type questionnaire scale,promoted the experience quality by using EEG and SC emotion metering at the same time,gained the conclusion by SPSS data analysis.The research found the demand level and group difference of college users.Accordingly,some construction suggestions were given.

Information resources are the cornerstone of the survival and development of university think tanks.Through the investigation of the information resources construction actuality of 24 typical university think tanks in our country,the paper found that the information resources construction of university think tanks had much room for improvement.The information resources construction of university think tanks guided by university libraries not only conformed to the policy guidance of think tanks,but also helped to better realize the functions of university libraries and promote the sustainable development of university think tanks.University libraries could carry out the information resources construction of university think tanks by formulating scientific planning,establishing a user-driven think tank resource construction model,revealing existing think tank resources,using the main construction mode combined with purchasing and building database,collecting network think tank results and paying attention to the sharing of think tank resources.

In this article,the author knew that research topics and academic papers had the close relationship between the support each other.In order to explore and utilize the existing analysis tools to carry out the financing innovation services,reasonable mining hottest interdisciplinary fields or the most advantage of research topics,in the future practice to build a feasible quantitative prediction model,it used the essay analysis tools to explore the research topics,discarding the commonly used inherent patterns of the ESI and Incites databases on the ESI discipline and the potential disciplinary analysis,using Incites database and ESI platform for the coupling relationship.It declared the National Natural Science Foundation of selected topics as an example,the statistics of the past 3 years the fund issued in the quantitative analysis,to study the quantitative analysis of a high-level SCIE/SSCI papers as a starting point,combining with the deeper "dark" data visual information processing method,suitable for scientific research subject topics provided ideas and reference,and books combining with a deeper "dark data" visualization information processing methods,to provide ideas and reference for the appropriate research topics,and for intelligence in the construction of high-level university for "dark data" visualization integrated intelligence analysis and the complementary service ability to ascend to explore practical research.

In order to understand the progress of Date Mining research,published articles on Date Mining during 2007-2016,which were recorded in the database of WOS,were chosen to be analyzed in this paper.By applying the method of bibliometric and Visualization,the paper analyzed the research status in this area according to the time,journals,disciplines,the research power,important literatures,research focus and frontiers through Cite Space.The study found that there are four clusters of international data mining research:network and computer technology,data mining method,data mining and management,data mining and application.Research front detected three development periods,each period had its own hot spots.Based on the knowledge map,the research status and development of international Date Mining were analyzed and summarized,which in order to provide reference for the researching about the Date Mining in our country.

[Objective] The low automation and dynamic processing capability of knowledge graph analysis systems and models are major obstacles in the advance of industrial innovation evaluation and prediction applications.[Context] In order to deal them,a novel industrial innovation evaluation and prediction model was proposed based on knowledge graph analysis.And its frameworks,functional modules,and working flows were given as following.By a novel grey sub graph algorithm,the model made industrial innovation project planes content definition and possibility assessment.And a knowledge graph growth and evolution algorithm was used to predict the future trend of industrial innovation.[Results] Actual measurement results showed that the model had better sub-graph convergence rate,graph coverage,and prediction efficiency than the traditional does.

CiteSpace Ⅲ was used to do the knowledge map analysis to all the innovation diffusion research literature collected in the Web of Science core set database,it was concluded that time and space distribution,key national network,co-cited reference and co-word knowledge maps,such as in the field of innovation diffusion is analyzed on the distribution,change trends and hop topics.The study results showed that:China and the USA were leading in the field of innovation diffusion research;earlier studies mainly focused on the macro level to forecast the new product sales by using the Bass model and its extended model,the current research had focused on from micro level to analyze the evolution of the innovation diffusion behavior by simulation method;the hot topics of innovation diffusion research mainly included the acceptance and adoption of innovative technology,the influencing factors of innovation diffusion,innovation diffusion power,innovation diffusion model and simulation study.

[Purpose/Significance] Based on the previous research on big data policy,the paper analyzed the focus and shortcomings of the big data policy in China.And the results provided reference for the subsequent policy innovation.[Method/Process] The paper used 9 National Big Data Comprehensive Experimental Zone issued policies in November 2014 to November 2017 as the research sample,with the grounded theory,word frequency analysis and method of policy strength measure to establish the weight of policy keywords.Then,using the experience of Ochiia coefficient matrix establish a keyword co-word matrix,and using visual analysis tools for common words network.By using the methods of degree centrality,cluster analysis,multidimensional scaling analysis and social network analysis to research the node's characteristics and network structure.[Result/Conclusion] Found that the policies issued by National Big Data Test Comprehensive Area were focus on financial fiscal support for the development of the big data,the big data industry cluster development,basic system construction and infrastructure construction,application of innovation and development,etc.There were deficiencies in the industry guidance fund,core enterprise cultivation,the deep application of big data and the introduction of external technology.

[Purpose/Significance] Open innovation is playing a more and more important role in the national science and technology innovation ecosystem.Carrying on the review and prospect of this field helps us to reflect deficiency and clarify the development direction so as to provide references for the related researches.[Method/Process] Based on the bibliometrics method,Bicomb and SPSS22.0 were used to measure and analyze the 572 academic papers with entitled "open innovation",including keywords co-word analysis,cluster analysis and multidimensional scale analysis.[Result/Conclusion] Knowledge map of open innovation researches in China was drawn and six research hotspots were revealed,accompanying which were four suggestions including focusing on the micro perspective of researches,strengthening the empirical researches,excavating mechanism of open innovation cooperation and paying attention to the open innovation platform construction.

Science and technology novelty search is an important part of sci-tech information service.It has significant meaning to scientific and technological innovation and transformation.The paper acquired the relative articles from CSSCI to find out the core authors,and seven research fields by biclustering algorithm.Furthermore,it analyzed the knowledge bases and research focuses of these fields systematically.