| As an increasing number of people choose to shop online and refer to online reviews of previous consumers,online consumer reviews(OCRs)have become important information resources for sellers,potential customers and researchers.Besides,what makes a review helpful to a potential customer in the process of making purchase decision has been a significant research topic.Although researchers have conducted plenty of studies on the helpfulness of OCRs from different perspectives,studies applying linguistic theories to achieve that goal are rare.This thesis aims to explore the relationship between the distribution of appraisal resources and helpfulness of OCRs by selecting 300 pieces of positive and negative reviews on books and phones in Amazon.com and annotating them according to specifications of Appraisal Theory(AT).Research questions of the thesis are presented as follows.(1)What are the similarities and differences of appraisal resources distribution between helpful positive reviews and negative ones?(2)Does the distribution of appraisal resources have impact on the review sentiment?(3)How can features of appraisal resources distribution explain the helpfulness of OCRs? The analysis of data in this thesis involves comparative approach and statistical method.Main findings of the thesis are as follows.Firstly,in terms of the distribution and frequency of appraisal resources in selected OCRs,generally engagement resources occupy the biggest percentage,followed by attitude resources,while graduation resources rank last.Specifically speaking,in the system of engagement obvious differences exist in terms of the percentage of proclaim and counter between positive and negative OCRs and no big gaps with regard to other engagement resources like entertain and attribute.The category of deny occupies a comparatively large percentage in either positive or negative OCRs.When it comes to the system of attitude,appreciation resources account for the largest percentage in both positive and negative OCRs,followed by affect resources,and last is the proportion of judgement resources.Besides,proportion of these attitudinal resources in positive OCRs is nearly same with that in negative ones.As far as the system of graduation is concerned,the percentage force resources occupy far exceeds that focus resources account for both in positive OCRs and negative ones.Specifically,resources of intensification occur most frequently in selected OCRs followed by resources of quantification while resources of soften occupy the smallest proportion in both kinds of OCRs.Secondly,according to the result of the Chi-square test,the distribution of appraisal resources is relevant with review sentiment,which means that the distribution of appraisal resources may decide whether an online consumer review is positive or negative to some extent.Furthermore,resources of attitude and graduation occur more frequently in positive OCRs while resources of engagement are more likely to occur in negative OCRs.Thirdly,the distribution of appraisal resources can partly explain the helpfulness of OCRs from four aspects,that is,abundant appraisal resources in OCRs,wide use of counter in positive OCRs,appropriate use of intensification and quantification in OCRs and more frequent use of attitude and graduation in positive OCRs and more resources of engagement in negative ones.These features provide some inspirations for indentifying helpful reviews.Findings of this thesis indicate that the distribution of appraisal resources does have some relevance with the helpfulness of OCRs and can make some explanations for the helpfulness of OCRs.AT,as the theoretical foundation of the study,has been proven to be an efficient theory to conduct such research. |
PDF Full Text Request