Online Hotel Selection Based On Review Mining And Pythagorean Fuzzy Sets

With the rapid development of e-commerce,many online shopping platforms and online shops have emerged,but the quality of goods varies.This makes it difficult for consumers to make quick decisions when purchasing goods.Usually,before placing an order,people first must review the alternative products,compare the pros and cons of the products' features let they care about,and then buy the product with the highest score in their minds.Considering the limited ability of manpower to collect and analyze massive reviews,we propose to use the theory and methods of review mining and the ability of Pythagorean fuzzy sets to describe and integrate the information to help people make decisions when shopping.However,the current research and application based on comment mining and Pythagorean fuzzy sets still have many shortcomings: 1)In comment mining,the process of identifying and extracting feature words and their corresponding evaluation information are more complicated.Research often ignores the sentiment analysis of neutral attitudes;2)When using the existing Pythagorean fuzzy sets algorithm to calculate information,the calculation of the membership degree and the non-membership degree is independent of each other.This will lead to loss of interactive information and insufficient results;3)The stability and applicability of the existing Pythagorean fuzzy aggregation operators are poor.They are easily affected by extreme values or 0-1 special values and will produce wrong decision results.This article is dedicated to solving the above problems and has carried out the following research:(1)We introduce T-norm function and T-conorm function into the existing Pythagorean fuzzy sets algorithm,and propose a more complete Einstein's interactive algorithm.This enhances the interaction between the membership degree and the non-membership degree in the calculation,and solves the problem that 0-1 special value is not applicable.Combined with the power average operator,this paper also constructs a series of Pythagorean fuzzy aggregation operators based on the interactive operation rules,and proposes multiple attribute group decision making method based on Pythagorean fuzzy Einstein interactive power weighted aggregation operator and Pythagorean fuzzy Einstein interactive weighted aggregation operator.This weakens the influence of extreme values on the results during information integration.By using MATLAB to simulate the decision-making process,and comparing with the results of some representative methods,we test the effectiveness,stability and applicability of the operators and decision-making method proposed in this article.(2)This article takes four indistinguishable hotels as the research object.We use Python to collect reviews on Ctrip,Tongcheng,and Qunar platforms,and use ROST-CM for text preprocessing such as deduplication,sentence segmentation,and word segmentation.After extracting feature-opinion phrases and performing sentiment analysis,we convert the text information into Pythagorean fuzzy sets according to relationships such as sentiment ratios.Using the multiple attribute group decision making method based on Pythagorean fuzzy Einstein interactive power weighted aggregation operator and Pythagorean fuzzy Einstein interactive weighted aggregation operator,we rank and compare the four hotels,and verify the reliability of the results.In this article,by expanding the Pythagorean fuzzy aggregation operator and combining the theory and method of comment mining,we propose a novel decision-making method.This solves the problem that consumers are difficult to choose when shopping,and provides effective technical supports and scientific purchase advices for the consumers,and greatly reduces the workload of decision-making.