Project evaluation method using non-formatted text information based on multi-granular linguistic labels

• Published in: Information fusion 2015-07, Vol.24, p.93-107
• Main Authors: Zhu, Jian-jun, Wang, He-hua, Ye, Chen, Lang, Qin
• Format: Article
• Language: English
• Subjects: Chinese text evaluation information; Decision analysis; Multi-granular linguistic label; Multiple-criteria decision analysis; Project evaluation
• ISSN: 1566-2535; 1872-6305
• DOI: 10.1016/j.inffus.2014.09.006

•We develop an analysis method for the non-formatted Chinese text evaluation information.•We propose a new decision making model derived from the text evaluation information.•We establish an estimation model for incomplete evaluation items with DS theory. We propose a novel project evaluation method for non-formatted Chinese text evaluation information. First, the non-formatted Chinese text evaluation information is determined and expressed using extensible markup language and a hypertext preprocessor. Then, the evaluation problem is transformed into a multiple-criteria decision-analysis problem based on multi-granular linguistic labels, including a comprehensive evaluation score for alternatives and an evaluation criteria point score for incomplete items. Next, we propose a weighting model for the criteria based on the minimal difference between the comprehensive evaluation score and the evaluation criterion point score of decision-makers. We establish an estimation model for incomplete evaluation items with the minimal evidence distance of Dempster–Shafer theory using maximal group consistency. In addition, we calculate a weighting for the decision-makers using the similarity of the group. Finally, we present a score modification method for alternatives based on weights of the criteria and the decision-maker. We use a soft science project evaluation and selection to illustrate the application process and feasibility of the proposed method.