基于情感分析和关系网络的影视产品评论数据文本挖掘研究

Text Mining of Comment Dada on Video Electronic Product Based on Sentiment Analysis and Relational Network

[目的/意义]挖掘观众对影视产品的主观感受,建立合适的模型评估影视产品的质量。[方法/过程]对Web影视产品评论数据建立评估模型,通过LDA和关系网络进行分析研究,并将结果与影视产品的网站评分做比较,从而得到较为准确的分析结果。[结果/结论]对影视产品的评论数据进行文本挖掘分析,能够得到观众的主观情感评分,更加客观地反映影视产品的质量高低。模型对于主观性强的评论数据具有较为可信的评估效果,研究方法也可应用于电商评论分析、舆情数据情感分析等方面。

[Purpose/significance]The paper is to dig the audience’s subjective sentiment on video electronic product,and establish a suitable model to evaluate the quality.[Method/process]The paper establishes an evaluation model of comment data on Web video electronic product,analyses it by LDA and relational network,compares the results with the scores from film website,and then obtains more accurate results.[Result/conclusion]Text mining of comment data on video electronic product gets the audience’s subjective sentiment scores,and reflects the quality more objectively.The model gives a believable result on strong-subjective review data,so it can be applied to the e-commerce comments and sentiment analysis of public opinion data.