标题：Probabilistic linguistic TODIM method for selecting products through online product reviews
作者：Liu, Peide; Teng, Fei
作者机构：[Liu, Peide; Teng, Fei] Shandong Univ Finance & Econ, Sch Management Sci & Engn, Jinan 250014, Shandong, Peoples R China.
通讯作者：Liu, Peide;Liu, PD
通讯作者地址：[Liu, PD]Shandong Univ Finance & Econ, Sch Management Sci & Engn, Jinan 250014, Shandong, Peoples R China.
关键词：Big data; Online product reviews; Multiple attribute decision making;; TODIM method; Probabilistic linguistic term sets
摘要：Online product reviews (OPRs) provide abundant information for potential customers to make optimal purchase decisions, and they apply Big Data to better understand product performance. OPRs contain an enormous stockpile of information; therefore, it is difficult for potential customers to make a comprehensive evaluation of alternative products through qualitative reviews. To facilitate consumer purchase decisions, ranking the alternatives based on the OPRs posted on social media platforms is a worthwhile research topic, although relative study is comparatively rare. Therefore, this article provides an extended probabilistic linguistic TODIM (PL-TODIM) method for assisting potential customers to evaluate alternative products through consumer opinions regarding product performance. In other words, this study introduces a novel multiple attribute decision making (MADM) method to rank products based on OPRs. To realize this goal, some basic theories of probabilistic linguistic term sets (PLTSs) are reviewed. Moreover, a possibility formula is first proposed to compare the probabilistic linguistic term sets (PLTSs). Furthermore, a combined weighting method is developed to determine objective weights based on cross entropy and entropy measures. Thus, the specific steps of the extended PL-TODIM method are described. After that step, in order to testify to the effectiveness and practicality of the proposed method, a case study of OPRs for SUVs is designed. Last, comparisons with other existing methods are further performed to show its advantages. (C) 2019 Elsevier Inc. All rights reserved.