针对传统网络舆情评估难以定量化,评定因素片面、不够精确的特点,提出了以网络舆情分级评估指标体系为基础,采用模糊综合评判模型的网络舆情指数计算方法,兼顾了时间、广度、数量、语义等因素,全面地评估了网络舆情态势。同时,利用层次分析法对判断矩阵进行一致性修正,采用筛选判断矩阵调整相容性,合理地确定了各级舆情评估指标的权重,使舆情指数计算值更为精确,为网络舆情预警打下了坚实的基础。
As the traditional internet opinion evaluation method hard in quantification and of unilateral evaluating factors, worse precision, and difficulties in calculation, this paper, based on hierarchical evaluating index system with uses fuzzy comprehension evaluation model, proposes an algorithm for evaluating the internet public opinion index. This algorithm takes time, range, sum, semantic into account and comprehensively evaluates the status of internet public opinion. Meanwhile, the algorithm also implements consistency correctness of the judgment matrix with AHP, adjusts the compatibility by screening the judgement matrix and thus defines the weight of all evaluating factors reasonably and makes the evaluation more accurate. This could lay a solid foundation for early warning.