| As a popular way for consumers to obtain product information and user experience,online reviews play an important supporting role in helping potential consumers reduce online shopping uncertainty and making purchasing decisions.Prior research on online reviews is usually conducted from the perspective of information diagnosis,and found that useful reviews should provide specific and sufficient information.While surveys show that reviews written by consumers are becoming more and more brief and short,little attention has been paid to abstract product online reviews.Generally,abstract reviews are treated in research as a contrast for specific comments,and the definitions and measurement methods of review specificity and abstraction in the existing literature are not consistent.Given the huge number of such online reviews exist,it is not sufficient to only focus on specific reviews.Rather,it is necessary to further explore the value of abstract reviews.Therefore,this study focuses on these "useless" but widespread abstract reviews and explores what utility values these abstract reviews have.First of all,based on different product types,this study introduces three objective measurement variables of review abstraction,namely review length,review information volume,review language characteristics,adopts the consumers’ subjective judgements to determine whether a review is abstract or not,and uses machine learning to select the most effective objective measurement method for review abstraction.Further,this research extracts a large amount of reviews from online shopping platforms,and based on the optimal measurement method obtained from previous machine learning results,the specificity and abstraction of the reviews are measured.In addition,the result indicates that abstract reviews are indeed perceived less useful than specific reviews,which is consistent with prior research.In response to the question of what practical values abstract reviews bring to consumers,this study conducts a laboratory experiment to explore how abstract reviews affect customers’attitudes towards the product.The results of the study indicate that,no matter what product type is,search goods or experience goods,the amount of information is the best measure for review abstraction.Further,although abstract reviews are of low usefulness,they have practical values for consumers’ decision-making.That is,abstract reviews can significantly change consumers’attitudes toward products.Through further analysis,this thesis finds that the overall attitude change towards the product generated by a set of reviews is stronger than the average attitude change of a single review in the review set.In other words,abstract reviews have a greater effect on consumers’attitude change when presented as a review set.This study focuses on abstract reviews,an important but often overlooked type of online review,and deepens the understanding of abstract reviews.By comparing the performance of different measurement methods,the optimal measurement method and its boundary range of comment abstraction are obtained,and the performance differences of the abstraction measurement methods are clarified.Secondly,this thesis explores the role of abstract reviews in shaping consumer decision-making,and discusses it from the perspective of influencing consumers’ attitudes towards the product.Therefore,it not only makes theoretical contributions to current research on online reviews,but also has practical implications for online e-commerce practitioners. |
PDF Full Text Request