Living-consumption platform has become a very important platform for customers to extract information of businesses

and view or submit comments on the quality of services or products.It is common that fake reviews

as a commercial activity

are used to exaggerate or damage the reputation of a target business

which is extremely harmful.This paper chose an O2O (Online To Offline) platform

from which reviews are derived

to study fake reviews.With an in-depth study on features of fake reviews

it raised the user-credibility and shop-credibility evaluation model respectively from the credibility perspective.Based on features of reviewers

businesses

and review texts

it established a fake review identification model

and through testing this model showed excellent performance in identification.