High-Dimensional Multi-Attributeed Data Visual Analysis System for Negative Comment Guidance
With the maturity technology of Internet platform and multi-user social network,the reference value of group user consumption experience is expanding.In the massive comment data,negative comments are particularly important for enterprises and consumers.Therefore,effective visual analysis of negative comments is necessary.According to the multidimensional and multivariate characteristics of comment data,the paper takes negative comments as the starting point and defines their scopes.Furthermore,the paper proposes an interactive visual analysis system,which provides enterprises and consumers with a new per-spective of comment analysis.Firstly,the paper uses the emotion analysis and opinion mining methods to process user data,and proposes a quantitative strategy for the difference of the consumer individual influ-ences.Secondly,the paper designs a series of interactive visual representation methods,such as the thematic sentiment ripple graph and the comment comparison view.Then,a multi-dimensional correlation view is constructed to explore the factors that affect the generation and causes of negative comments,and their indi-vidual differences with dynamic interactive methods.Finally,three case studies are used to verify the effec-tiveness and practicality of the system,which can also be extended to comment data of other visual analysis fields.
NETL
NSTL
万方数据
