Proceedings of the 2nd International conference on Electronics, Communication and Aerospace Technology (ICECA 2018)

IEEE Conference Record # 42487; IEEE Xplore ISBN:978-1-5386-0965-1

Smart Local Shopping System

Rohan Padaya Sumeet Suvarna
Department of Information Technology Engineering, Department of Information Technology Engineering,
Vidyalankar Institute of Technology Mumbai, India Vidyalankar Institute of Technology Mumbai, India,
[email protected] [email protected]

Ankit Channe Chintan Shah

Department of Information Technology Engineering, Assistant Professor, Dept. of Information Technology
Vidyalankar Institute of Technology Mumbai, India, Vidyalankar Institute of Technology
[email protected] [email protected]

Abstract— With ever increasing need for buying are lagging behind in the online scenario. Due to attractive
commodities, it has become very essential that the delivery speed incentives and wide varieties provided by e-stores, there is an
is up to the mark. More and more products are sold via e- adverse effect of online shopping on the fixed shop retailers.
commerce because the ease of access and efficiency of Internet This has provoked the fear of uncertainty and helplessness into
has made trading very popular. Youth are trying to order most of the mind of shopkeepers [5]. Local shops lack exposure to the
their daily necessities online, but surely till now, this hasn’t
current online trade which in turn affect the economy of local
affected the sales of day to day commodities. These shopkeepers,
who are willing to enter the online arena are overwhelmed by all markets and create a prejudice among the local shops against
the technological ideologies and technical nomenclatures. They online websites. On the other hand, online system cannot
want to market themselves online with very little technical provide instant delivery of daily necessity products to their
knowledge involved in the process. The proposed solution is a customers.
platform for shopkeepers to market their commodities and Competition in the market is beneficial for the average
services to an average consumer. This solution is a middleware consumers as this tends to increase the quality of products and
between shopkeepers and consumers which acts as a chance of decrease in the cost of the product. If the local
recommendation system. We are trying to bridge this gap where vendors are given an opportunity to enter the online shopping
consumers are more technology-oriented and shopkeepers lie at
system, then it would be a great opportunity for them to
another end of the spectrum. In turn, the shopkeeper can also
benefit by getting a data feed about most searched commodity in increase their sales and widen their business in the midst of
their vicinity. The consumers, on the other hand, get to choose such healthy competition.
which shops to visit or get their product delivered from. This In online shopping systems, users come across millions of
would reduce the monopoly caused by the e-commerce giants like products to choose from. So it becomes difficult for them to
Amazon, Flipkart in the market. Also, the consumer enjoys shop from the vast array of products. This consumes a lot of
instant knowledge about the availability of the product. time and efforts of consumers in searching for the desired
product at an optimal cost [9]. This lead to the idea of the
Keywords— E-commerce, Recommendation system recommender system. A recommender system acts as a
personal assistant that helps you in choosing from a
I. INTRODUCTION comprehensible list of items, thus saving time and energy.
Recommender system simplifies the job of users to fetch the
A significant amount of demographic is choosing online desired item by recommending the related item which the user
shopping – which is comfortable – over legacy schemes like may like [6], [11]. This recommendation is based on user’s
window shopping. This has made people aware of the benefits previous search history and previously purchased items. So
that e-commerce holds over regular commerce. This this system uses recommender to prioritize the item based on
generation, who has made online shopping as their default user preferences.
way of buying lifestyle commodities is yet to harness a similar
way of shopping for day-to-day commodities. A. Usefulness of web vendors

E-commerce giants like Amazon, Flipkart have cemented a ● Save business projects from collapsing and
market of their own as they provide smallest of the household increasing profits.
item to huge luxurious items at the fingertips of the customers ● Curtail expenditures spent on manpower
[4]. Variety of items are available online on these e-commerce ● Opportunity to serve both local and foreign
sites. This has created instability in the market as local shops customers

978-1-5386-0965-1/18/$31.00 ©2018 IEEE 868

Authorized licensed use limited to: Victoria University. Downloaded on January 18,2025 at 01:13:04 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2nd International conference on Electronics, Communication and Aerospace Technology (ICECA 2018)
IEEE Conference Record # 42487; IEEE Xplore ISBN:978-1-5386-0965-1

● Promote their products and create noticeable online Pooja Vashisth et al [10] provided insights on Interest Based
shops Recommendation agents which are used to generate
● Provide excellent customer experience by launching interesting recommendations. The agents are evolving in
fast functioning and accessible web stores nature which uses feedback from the users for improving
In the rest of the paper, we first review the related work of future recommendations for them. The feedback includes
the online shopping system and recommender systems in user’s behavior on the web, user preference, and user
the Section II. Then we state the Aim and Objective of profile.
this system in Section III. The Section IV describes the
current situation of market from customers as well as III. AIM AND OBJECTIVE
shopkeeper’s point of view. Then we describe the A. Aim:
proposed system in Section V. The next section, Section The important aim of this project is to bring the concept
VI talks about the scope of the system followed by the of online shopping into local market thereby giving an
detailed methodology used for the system in the Section opportunity to the local shops to widen their business.
VII. Finally, we draw conclusion and give references. The proposed system also aims at providing a
platform which allows users to get information about the
II. RELATED WORK availability of desired product in their respective local
shops at optimal cost and distance.
Previous work has been referred which proposes a lot of
challenges related to the online shopping system and B. Objective:
recommendation system. The following are objectives of Smart Local
In the year 2003, Junzhong Ji et al [1] proposed a paper on Shopping System:
an online recommendation based on customer shopping
model. In this paper, he introduced the idea of using the ● To minimize the time and human efforts of
Bayesian network for building a customer shopping model users required to find a particular product.
which helps in identifying the customer’s shopping ● To provide the statistics of the demand for
items in a particular locality to the shopkeeper
behavior.
thereby helping them to maintain the supply of
In 2016, Supriya Jaiswal et al [2] has written a paper where
commodities.
the author analyzes the user's search history based on
● To encourage healthy competition in the local
recommender system. Recommender systems are based on market and avoid monopoly.
information about a user’s past patterns and consumption ● To provide appropriate recommendations to
patterns in general and recommend new items to the user. users based on their search history.
Using genetic algorithm combined with K nearest neighbor,
Hong-Wei-Yang et al [6] developed a method that selects IV. PROBLEM STATEMENT
the appropriate products for customers based on their In today’s era of internet, a variety of products are
demands and personal preferences. available through online shopping to the customers at their
As per Gediminas Adomavicius et al [3], recommender doorstep which in turn makes them prefer online shopping
systems are classified into following categories based on over local shops. Hence, there is a significant dip in sales of
how recommendations are made the local shops.
● Content-Based Recommender System: The user will be Also, there is lack of the credible source in the shopping
recommended items similar to the ones the user preferred in domain which provides the statistics regarding the demand of
the past commodities to the shopkeeper in their locality. The
● Collaborative Recommender System: The user will be shopkeepers need to rely on the indirect sources of
recommended items that people with similar tastes and information such as the advertisement, sales of similar
preferences liked in the past [7]. products in neighboring shops and the profit margin of a
● Hybrid Recommender System: Combination of both particular product to understand the demand in their vicinity.
content-based and collaborative recommender systems. On the other hand, in online shopping, the customers have to
Kavinkumar.V et al [12] also put forth the same idea of wait for at least a day to get their product delivered.
hybrid recommendation with the additional feedback This delay in delivery of a product is not feasible for daily
analysis feature in which customers were provided with need commodities. To overcome these problems, we have
better suggestions depending upon user based and item proposed Smart local shopping system which provides
based collaborative filtering. information about user’s desired product at their fingertips.
Badrul Sarwar et al [8] has compared collaborative filtering
with other recommender system algorithms and concluded V. PROPOSED SYSTEM
that collaborating filtering is the most successful The Smart Local shopping system is an e-commerce
recommender system algorithm on the web. based recommender system that provides a platform

978-1-5386-0965-1/18/$31.00 ©2018 IEEE 869

Authorized licensed use limited to: Victoria University. Downloaded on January 18,2025 at 01:13:04 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2nd International conference on Electronics, Communication and Aerospace Technology (ICECA 2018)
IEEE Conference Record # 42487; IEEE Xplore ISBN:978-1-5386-0965-1

independent application which can run on both Android and server for authentication. During the authentication of shops,
IOS. the location of the shop will be taken. Once the customer is
The system will keep the track of the most searched items in a authenticated he can search for the product. Customer’s
locality. This statistic will be updated on weekly basis. runtime location will be taken by the server and then by
The application will show the statistical data to matching the location of the customers and shops, the server
shopkeeper in a visual format like pie charts. The system will retrieve the list of shops which are nearer to the customer.
includes an inbuilt barcode scanner which allows the Depending upon the most searched and purchased product in
shopkeeper to scan new products and to register it in the the market, Shopkeeper will be provided with the statistics
shopkeeper’s database. which will help them to know the demand in the market.
A. Customer module
VI. SCOPE
The user has to download a phone application which will
The Smart Local Shopping system has extensive scope in be used for all further interactions with the server. Every user
the trade market to help the shopkeepers to maintain the will be validated by making sure that they register with server
demand and supply of commodities. The system will process before interacting with the application. Once successfully
the historical data of user’s search history and items registered, the user is prompted to enable location services or
purchased. A hybrid algorithm will be developed which will he/she can enter their address accordingly. User is provided
notify users. The data will be analyzed and tuned to optimize with a search bar to enter desired product. He is then provided
accuracy and diversity. with search results based on his location. User can filter his
● Our Platform will reduce human effort and time searches according to prices, accessibility and credibility of
required to know the location of the local shops the shop.
where required item is available.
● Shopkeepers will be provided with the statistics B. Machine Learning Module
regarding the demand in the market. Based on user’s search histories and collective purchase
● Through our application, Customers will get results, a Machine Learning module will recommend products
customized notification on weekly basis based on which the users are most likely to consume. This module will
their preferences. also target potential audience for a new product.
● Shopkeepers can post advertisements about new
products introduced in the market according to his
C. Shopkeeper Module
leisure.
The shopkeeper has to provide their shop location while
VII. METHODOLOGY registering in the application. After signing up, the shopkeeper
The Smart Local Shopping System consist of three needs to register all his products using the barcode scanner
different modules: provided by the application. Based on search history in that
locality the shopkeeper is presented with statistics of most
searched items.

VIII. CONCLUSION

Our solution targets shopkeepers who are

technologically less advanced. This comes with an expectation
for an extra layer of simplicity. Currently, the solution comes
with simple but feature-rich UI for shopkeepers. Since the
solution broadens the local reach of the shopkeeper, it will
result in determining trending commodities. Accurate
predictions will further expand local businesses and will
enable newcomers with potential to have an equal share in the
economy. The customers are likely to experience increased
consumption of goods because of simplicity in their
Fig. 1. Block diagram of Smart Local Shopping System availability. With more people moving online for simple
purchases, the shop centers are likely to experience few rush
This System will contain a Centralized database which hours. Less time spent in shopping queues will result in fewer
involves the details of the product and the information of the frustrating shopping experience
shops in different tables. ML engine is used for customized
notifications for customers depending on the customer’s
previous search history and advertisements posted by
shopkeepers. Customer and shopkeeper communicate with

978-1-5386-0965-1/18/$31.00 ©2018 IEEE 870

Authorized licensed use limited to: Victoria University. Downloaded on January 18,2025 at 01:13:04 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2nd International conference on Electronics, Communication and Aerospace Technology (ICECA 2018)
IEEE Conference Record # 42487; IEEE Xplore ISBN:978-1-5386-0965-1

REFERENCES [7] Jian Shi, Junyun Chen, “An Application Study on Collaborative
Filtering in E-commerce”, Zhuolan Bao Research Center of
Corporate Governance of Nankai University Business School of
[1] Junzhong Ji, Zhiqiang Sha, Chunnian Liu College of Computer
Nankai University Tianjin, China.
Science and Technology, Beijing University of Technology
Beijing Municipal Key Laboratory of Multimedia and Intelligent
[8] Badrul Sarwar, George Karypis, Joseph Konstan, and
Software Technology Beijing 100022, China.
John Riedl, “Analysis of Recommendation Algorithms for E-
Commerce”, GroupLens Research Group / Army HPC Research
[2] Supriya Jaiswal, Aarohi Ajgaonkar, “Online shopping
Center Department of Computer Science and Engineering
recommendation optimization based on user’s previous search
University of Minnesota Minneapolis,MN 55455.
history”, Student, M.C.A. Department, Sardar Patel Institute of
Technology, Mumbai, India, Volume: 03 Issue: 04, April-2016.
[9] Andrea Horch, Holger Kett and Anette Weisbecker,” Mining
E-Commerce Data from E-shop Websites.”, Fraunhofer Institute
[3] Gediminas Adomavicius, Member, IEEE, and Alexander
for Industrial Engineering IAO Stuttgart, Germany.
Tuzhilin,Member, IEEE “Toward the Next Generation of
Recommender Systems: A Survey of the State-of the Art and
[10] Pooja Vashisth, Ounam Bedi, “Interest-Based Persosonalized
Possible Extensions”,VOL. 17, NO. 6, JUNE 2005.
Recommender System”, Department of Computer
Science,University of Delhi, INDIA.
[4] Grey Linden, Brent Smith and Jeremy York,” Amazon.com
Recommendations Item-to-Item Collaborative Filtering”,
Amazon.com, February-2003.
[11] J. Ben Schafer, Joseph A. Konstan, John Riedl, “E-commerce
Recommender Applications”, GroupLens Research Project
[5] Amit Shah, A Study on “The impact of online shopping upon
Department of Computer Science and Engineering University of
retail trade business”, Royal School of Commerce Royal Group
Minnesota Minneapolis, MN 55455.
of Institution (RGI) Guhawati, Assam.
[12] Kavinkumar.V, Rachamalla Rahul Reddy, Rohit
[6] P.Resnick and H.R.Varian Recommendations systems,
Balasubramanian, Sridhar.M, Sridharan.K Dr. D.Venkataraman,
Communications of the ACM, vol. 40, pp. 56-58, 1997.
”A Hybrid Approach for Recommendation system with Added
Feedback Component “, Department of Computer Science and
Engineering, Amrita School of Engineering, Amrita Vishwa
Vidyapeetham (University), India

978-1-5386-0965-1/18/$31.00 ©2018 IEEE 871

Authorized licensed use limited to: Victoria University. Downloaded on January 18,2025 at 01:13:04 UTC from IEEE Xplore. Restrictions apply.