2.1 Product analysis based on
The paper deals about the fake reviews and neural
reviews given by the customer. The author wanted to identify those fake and neural reviews and he proposed
a system analyses that deals with user’s sentiments4. This system will
generate analysis for both manufactures and as well as customers. A report of
products will be created where the user can view the product wise ratings based
on customer reviews4, also the user can know about the features of the
product through feature based product analysis, due to this customer can get to know
about particular feature and as well as manufacturers can get to know the exact
reason of the failure of their product. Since opinions will change along with
time to come over this time-line analysis is introduced. The user can view the
performance of the product over time and can compare it with similar products
available in the market. Since the analysis is totally review based, a
product’s popularity will definitely show impact on the product’s quality.
The system also provides product recommendation to users if the
product he viewed has negative opinion overall. Also by tracking the kind of
products generally user reviews, author aim to build a recommendation engine
that suggests similar products to the user based upon his liking for products.
Also due the ever increasing competition between various manufacturers, and the
wide variety of options available to the customers, the churn rate of customers
is high. So the manufacturers can use the system for customer churn management.
2.2 Importance of customer product reviews from
customer prospective :
The problem with this paper is that the comparison of user reviews
is time consuming, because user reviews are in unstructured data. The solution
with this problem is Tag expression approach, it combines tags and ratings
allows user to get affect to this tag. For every tag user can state if they
like, dislike, or neutral about it. The drawback of this method is that user
can review only in 3 possible ways whether positive, negative, or neutral.
Online Costumer reviews for product Feature-based Ranking :
It is becoming increasingly difficult for customers to make
purchasing decisions based on only pictures and short product
Descriptions. Author provided the solution through feature
based product ranking
technique that mines thousand of
customer reviews. First thing is need to identify the specific
product features and analyze
their frequencies and relative
usage. For each individual feature they identified the
subjective and comparative sentences in reviews and then
by applying sentiment
orientations to the sentences.
By this result the relationship among products
were represented in a directed graph. At the
this graph will result in relative quality of product.