International Journal of Research in Advanced Technology - IJORAT

Vol. 1, Issue 5, MAY 2016

MARKET ANALYSIS AND SELLER TRUST

MINING BASED ON MDTE MODEL
S.Nirmala1, P. Suganthi2, S.Kavitha3
PG Scholar 1, Assistant Professor 2,3
Department of Computer Science and Engineering
1,2
K.L.N College of Information Technology, Madurai,
3
Velammal College of Engineering and Technology, Madurai
Abstract: The primary goal of CommTrust model to analyze the market and seller feedback, based on the
observation that buyers often express opinions openly in free text feedback comments. This survey paper deals with
discussion of various algorithm, technique and methods related to trust model, for computing comprehensive trust
proles for sellers in e-commerce applications. To improve the trust worthy of the sellers based on specific product
to the customers who all are wants the product, by computing dimension trust scores and dimension weights
automatically via extracting dimension ratings from feedback comments. The main contributions include: 1)
propose a multidimensional trust model for computing reputation scores from user feedback comments; and 2)
propose an algorithm for mining feedback comments for dimension ratings and weights, combining techniques of
natural language processing, opinion mining, and topic modeling. The new research is the first piece of work on
trust evaluation by mining feedback comments and user recommended system also implement in this application,
observe the user preferences and advertise the service to users.
Key Terms: multi-dimensional trust evaluation, mining feedback, natural language processing, opinion mining, and
topic modeling.
I.INTRODUCTION
Comment-based Multi-dimensional trust (CommTrust), a fine-grained multi-dimensional trust evaluation
model by mining e-commerce feedback comments.
Comprehensive trust profiles are computed for sellers,
including dimension reputation scores and weights, as well
as overall trust scores by aggregating dimension reputation
scores. Propose an approach that combines dependency
relation analysis, a tool recently developed in natural
language processing (NLP) and lexical based opinion mining
techniques to extract aspect opinion expressions from
feedback comments and identify their opinion orientations.
Further propose an algorithm based on dependency relation
analysis and Latent Dirichlet Allocation (LDA) topic
modelling technique to cluster aspect expressions into
dimensions and compute aggregated dimension ratings and
weights. As a result we make use of the structures on aspect
and opinion terms, as well as negation defined by
dependency relations to achieve more effective clustering.

II.

SYSTEM ARCHITECTURE

System design is the process of defining the

architecture, components, modules, and data for a system to
satisfy specified requirements. One could see it as the
application of systems theory to product development. There
is some overlap with the disciplines of systems analysis,
systems architecture and systems engineering. If the broader
topic of product development blends the perspective of
marketing, design, and manufacturing into a single approach
to product development, then design is the act of taking the
marketing information and creating the design of the product
to be manufactured. In figure 3.1 System design is therefore
the process of defining and developing systems to satisfy
specified requirements of the user.

Advantages

Compute
Multi-Dimensional
E-Commerce
Feedback taken
Efficiently identify the Seller rating
Create awareness of seller and product to Customer
Figure. 1 System Architecture

All Rights Reserved 2016 IJORAT

24

International Journal of Research in Advanced Technology - IJORAT

Vol. 1, Issue 5, MAY 2016

III.

EXISTING SYSTEM

In the olden days the merchant of the product may

show the fake feedback of the product and the customer may
receive the infrequent feedback and failed to get the product
from the same. Because the existing must have the product
feedback from the previous users and the drawback of the
existing is to lead the sellers business low and the improper
product purchased by the buyer. They wont know about the
quality of the sellers. The problem always lead the seller and
buyer to process the fault one. To overcome these they wont
follow any thing and this system become a failure one.
IV. OPINION EXPRESSIONS
Although buyers leave positive feedback ratings,
they express some disappointment and negativeness in free
text feedback comments, often towards specific aspects of
transactions. For example, a comment like The products
were as I expected. expresses positive opinion towards the
product aspect, whereas the comment Delivery was a little
slow but otherwise, great service. Recommend highly.
Expresses negative opinion towards the delivery aspect but a
positive opinion to the transaction in general. By analyzing
the wealth of information in feedback comments can
uncover buyers detailed embedded opinions towards
different aspects of transactions, and compute
comprehensive reputation profiles for sellers.
If a comment expresses opinion towards
dimensions then the dimension words and the opinion words
should form some dependency relations. It has been reported
that phrases formed by adjectives and nouns, and verbs and
adverbs express subjectivity
V.USER FEEDBACK COMMENTS
Our analysis of feedback comments on eBay and
Amazon reveals that even if a buyer gives a positive rating
for a transaction, s/he still leaves comments of mixed
opinions regarding different aspects of transactions in
feedback comments.
For example for comment c2, a buyer gave a
positive feedback rating for a transaction, but left the
following comment: bad communication, will not buy
from again. Super slow ship (ping). Item as described.
Obviously the buyer has negative opinion towards the
communication and delivery aspects of the transaction,
despite an overall positive feedback rating towards the
transaction.
VI. CLUSTERING DIMENSION EXPRESSIONS

call dimensions. Different from the conventional topic

modelling approach, which takes the document by term
matrix as input, Lexical-LDA makes use of shallow lexical
knowledge of dependency relations for topic modelling to
achieve more effective clustering. To make use of two
types of lexical knowledge to supervise clustering
dimension expressions into dimensions so as to produce
meaningful clusters.
Comments are short and therefore co-occurrence of
head terms in comments is not very
informative. We
instead use the co-occurrence of dimension expressions with
respect to a same modifier across comments, which
potentially can provide more meaningful contexts for
dimension expressions.
Observe that it is very rare that the same aspect of
e-commerce transactions is commented more than once in
the same feedback comment. In other words, it is very
unlikely that the dimension expressions extracted from the
same comment are about the same topic.
6.1 Aggregate Dimension Evaluation
The weight for a dimension is proportional to the
total number of positive and negative ratings on the
dimension. Specifically we compute the total number of
(modifier, head) dimension expressions for the dimension.
Indeed only frequent dimension expressions with head terms
appearing in at least 0.1% of comments are included. The
total number of dimension expressions for dimensions are
normalized to produce the dimension weights.
6.2 User Observation
It is Important that a product is returned in the
highest rank position for as many user preferences as
possible. User history is being observed, and Estimate the
impact of a product compared to their competition product
then product is advertised to the potential customers.
Data mining also popularly known as
Knowledge Discovery in Databases refers to the nontrivial
extraction of implicit, previously unknown and potentially
useful information from data in databases. While data
mining and knowledge discovery in databases are frequently
treated as synonyms.
In principle, data mining is not specific to one type
data. Data mining should be applicable to any kind of
information repository. Data mining is being put into use for
relational databases, data warehouses, transactional
databases, World Wide Web, spatial databases, multimedia
databases, time-series databases and textual databases.

The Lexical-LDA algorithm cluster aspect

expressions into semantically coherent categories, which
All Rights Reserved 2016 IJORAT

25

International Journal of Research in Advanced Technology - IJORAT

Vol. 1, Issue 5, MAY 2016

QUERY PROCESSING
All database systems must be able to respond
to requests for information from the useri.e. process
queries. Obtaining the desired information from a database
system in a predictable and reliable fashion is the scientific
art of Query Processing. Getting these results back in a
timely manner deals with the technique of Query
Optimization.
E-COMMERCE
Electronic commerce, comm. Only known
as E-commerce, is trading in products or services using
computer networks, such as the Internet. Electronic
commerce draws on technologies such as mobile
commerce, electronic funds transfer, supply chain
management, Internet
marketing, online
transaction
processing, ElectronicDataInterchange (EDI), inventory
management
systems,
and
automated data
collection systems. Modern electronic commerce typically
uses the World Wide Web for at least one part of the
transaction's life cycle, although it may also use other
technologies such as e-mail.
E-commerce businesses usually employ some or all of the
following practices:

It is also estimated that the e-commerce market is about 57%

from small towns and the balance from the largest
metros. The most popular use of e-commerce is on travel
websites, which is done by about 70% of all e-commerce
consumers in India.
There are challenges to e-commerce companies operating in
India - including some reluctance of consumers to pay for
goods or services online. Some companies are trying to get
around this problem by providing cash-on-delivery as a
service to their consumers.
Sentiment analysis
Sentiment analysis may involve analysis of movie
reviews for estimating how favorable a review is for a
movie. Such an analysis may need a labeled data set or
labeling of the affectivity of words. Resources for affectivity
of words and concepts have been made for Word
Net and Concept Net, respectively.
Text has been used to detect emotions in the related
area of affective computing. Text based approaches to
affective computing have been used on multiple corpora
such as students evaluations, children stories and news
stories.
Implications

Provide E-tail or virtual storefront on websites with online

catalogs, sometimes gathered into a "virtual mall" Buy or
sell on websites or online marketplaces. Gather and use
demographic data through web contacts and social media.
Use electronic data interchange, the business-to-business
exchange of data. Reach prospective and established
customers by e-mail or fax (for example, with newsletters).
Use business-to-business buying and selling. Provide secure
business transactions
Advantages of E-commerce

It can help increase profits; it can increase sales and

decrease costs.
It can help organizations do business 7 days a week
and 24 hours a day.
It can help organizations have customers all around
the globe and not be limited to a specific region.
It helps organizations bring higher return on
advertisements, if managed properly.
It helps organizations identify new suppliers,
partners and customers.

The e-commerce market in India is estimated to grow from

$10 billion every year to between $ 70 - 260 billion every
year by 2025.

All Rights Reserved 2016 IJORAT

Until recently, websites most often used text-based

searches, which only found documents containing specific
user-defined words or phrases. Now, through use of
a semantic web, text mining can find content based on
meaning and context (rather than just by a specific word).
Additionally, text mining software can be used to
build large dossiers of information about specific people and
events. For example, large datasets based on data extracted
from news reports can be built to facilitate social networks
analysis or counter-intelligence. In effect, the text mining
software may act in a capacity similar to an intelligence
analyst or research librarian, albeit with a more limited scope
of analysis.
Trust Representation
According to authors YonghongWang and Munindar P.
Singh [3] considers a distributed system of software agents
who cooperate in helping their users to find services,
provided by different agents. The agents need to ensure that
the service providers they select are trustworthy. Because the
agents are autonomous and there is no central trusted
authority, the agents help each other determine the
trustworthiness of the service providers they are interested
in. This help is rendered via a series of referrals to other
agents, culminating in zero or more trustworthy service

26

International Journal of Research in Advanced Technology - IJORAT

Vol. 1, Issue 5, MAY 2016

providers being identified. Munindar P. Singh [3] describes

trust network is a multiagent system where each agent
potentially rates the trustworthiness of another agent. This
paper develops a formal treatment of trust networks. At the
base is a recently proposed representation of trust via a
probability certainty distribution. The main contribution of
this paper is the definition of two operators, concatenation
and aggregation, using which trust ratings can be combined
in a trust network. This paper motivates and establishes
some important properties regarding these operators, thereby
ensuring that trust can be combined correctly. Further, it
shows that effects of malicious agents, who give incorrect
information, are limited.
The mechanisms that support finding trust estimations are
called reputation systems. There are two kinds of reputation
systems: centralized and distributed. Centralized reputation
systems include collaborative filtering (social information
filtering) systems (Breese, Heckerman, &Kadie 1998;
Resnicket al. 1994; Dellarocas 2004), and online reputation
systems (Dellarocas 2004). Distributed reputation systems
include peer-to-peer (P2P) systems (Aberer&Despotovic
2001; Xing& Liu 2004) and referral systems (Yu & Singh
2002; 2003).This paper uses a probabilistic theory of
evidence to represent the trust between agents in terms of
referrals and the quality of service obtained. To support an
agent computing trust in a service provider with which it has
no direct interactions, this paper defines operators to
determine trust by concatenating and aggregating paths from
an agent to the given service provider. The underlying
architecture of a distributed system is common to several
approaches, and not novel to this paper. This paper
formalizes some key properties of the concatenation and
aggregation operators. Lastly, this paper shows how our
operators mitigate the effects of deception by some agents.
A reputation system for an open environment should support
the following features. First, agents may join or leave an
open environment arbitrarily. Second, the agents need not be
cooperative. An agent may give biased information on its
own experience. It may also provide many witnesses whose
ratings of the intended target are biased.
VII. CONCLUSION
At one fine moment the decried problem can
overcome with the new techniques and the analyze on the
user and the merchants, the feed backs from the existing
user can collected and identify the best one to produce the
best seller for the important one of the site as new user.
This is the end of the new developed site and they must
survey with the new feedback collection and identify the
best seller with the best product and the range as well as
rate of the sellers business margin. The best seller can be
show to the new users of the site.

All Rights Reserved 2016 IJORAT

7.1 FUTURE ENHANCEMENT

The new researcher can improve the rating method with the
upcoming technique and improve the rating method for the
merchant and the quality of the best seller can be improved
too better.
REFERENCES
[1]

B. Yu and M. P. Singh, Distributed reputation

management for electronic commerce, Comput.
Intell., vol. 18, no. 4, Nov. 2002, pp. 535549.

[2]

G. Qiu, B. Liu, J. Bu, and C. Chen, Opinion word

expansion and target extraction through double
propagation, Comput. Linguist., vol. 37, no. 1,
2011, pp. 927.

[3]

H. Zhang, Y. Wang, and X. Zhang, Efficient

contextual trans- action trust computation in ecommerce environments, in Proc. 11th IEEE
TrustCom, Liverpool, U.K., 2012.

[4]

H. Zhang, Y. Wang, and X. Zhang, A trust vector

approach to transaction context-aware trust
evaluation in e-commerce and e- service
environments, in Proc. 5th IEEE Int. Conf. SOCA,
Taipei, Taiwan, 2012.

[5]

H. Wang, Y. Lu, and C. Zhai, Latent aspect rating

analysis with- out aspect keyword supervision, in
Proc. 17th ACM SIGKDD Int. Conf. KDD, San
Diego, CA, USA, 2011, pp. 618626.

[6]

S. Brody and N. Elhadad, An unsupervised aspectsentiment model for online reviews, in Proc. HLT,
Los Angeles, CA, USA, 2010, pp. 804812.

[7]

Liu, Sentiment Analysis and Opinion Mining. San

Rafael, CA, USA: Morgan & Claypool Publishers,
2012.

[8]

D. M. Blei, A. Y. Ng, and M. I. Jordan, Latent

Dirichlet alloca- tion, J. Mach. Learn. Res., vol. 3,
pp. 9931022, Jan. 2003.

[9]

Hofmann, Probabilistic latent semantic indexing,

in Proc. 22nd ACM SIGIR, New York, NY, USA,
1999, pp. 5057.

[10]

Y. Lu, C. Zhai, and N. Sundaresan, Rated aspect

summarization of short comments, in Proc. 18th
Int. Conf. WWW, New York, NY, USA, 2009.

27

International Journal of Research in Advanced Technology - IJORAT

Vol. 1, Issue 5, MAY 2016

[11]

H. Wang, Y. Lu, and C. Zhai, Latent aspect rating

analysis with- out aspect keyword supervision, in
Proc. 17th ACM SIGKDD Int. Conf. KDD, San
Diego, CA, USA, 2011, pp. 618626.

[12]

H. Wang, Y. Lu, and C. Zhai, Latent aspect rating

analysis on review text data: A rating regression
approach, in Proc. 16th ACM SIGKDD Int. Conf.
KDD, New York, NY, USA, 2010, pp. 783792.

[13]

S. Ramchurn, D. Huynh, and N. Jennings, Trust in

multi-agent systems, Knowl. Eng. Rev., Vol. 19,
no. 1, pp. 125, 2004. .

[14]

B. Yu and M. P. Singh, Distributed reputation

management for electronic commerce, Comput.
Intell., vol. 18, no. 4, pp. 535549, Nov. 2002.

[15]

M. Schillo, P. Funk, and M. Rovatsos, Using trust

for detecting deceptive agents in artificial
societies, Appl. Artif. Intell., vol. 14, no. 8, pp.
825848, 2000.

[16]

J. Sabater and C. Sierra, Regret: Reputation in

gregarious soci- eties, in Proc. 5th Int. Conf.
AGENTS, New York, NY, USA, 2001, pp. 194
195.

All Rights Reserved 2016 IJORAT

28