Abstract: 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.

Key words: Internet encyclopedia, Quality evaluation, Machine learning, Multi-decision model, Online information evaluation