基于多决策模型的百科词条质量评价方法研究--以百度百科为例

Research on Evaluation of Article Quality in Internet Encyclopedia Based on Multi-decision Model:Case Study of Baidu Encyclopedia

提出一种基于多决策模型的百科词条质量评价方法,将质量评价转化为多属性决策问题,利用多分类器集成学习的模式对百科词条进行评价,结合属性权重得出质量评分与分类。在实验上,以百度百科作为数据源,进行了模型的统计分布验证与评分可信度验证,并与传统分类模型进行对比,获得了较高的准确度。实验表明,多决策模型在百科词条评价上有较高的可行性、合理性和有效性,为百科词条质量评价提出了新的解决方案。

This paper proposes a method for evaluating the quality of internet encyclopedia articles based on a multi-decision model.The method transforms the quality evaluation into a Multi Attribute Decision Making problem,uses multi-classifier in ensemble learning mode to evaluate encyclopedia articles,and combines attribute weights to obtain quality classification and scoring.In the experiment,this article uses Baidu Encyclopedia as the data source,carrying out the statistical distribution verification and rating credibility verification of the model.Comparing with traditional classification models,this model obtains high accuracy result.Experiments show that the multi-decision model has high feasibility,rationality and effectiveness in the evaluation of encyclopedia articles,and a new solution is proposed for the evaluation of article quality in internet encyclopedia.