Vol 11, Issue 4 , April/ 2020

ISSN NO: 0377-9254

FLIPKART PRODUCT RECOMMENDATION SYSTEM

T. Keerthana1,T. Bhavani2, N. Suma Priya3 , V. Sai Prathyusha4, K.Santhi Sri5

IV B.Tech, Department of Information Technology, Vignan’s Nirula Institute of Technology &

1,2,3,4

Science for Women, Peda Palakaluru, Guntur-522009, Andhra Pradesh, India.

5
Assoc.Professor, Vignan's Foundation for Science, Technology & Research, Vadlamudi, Guntur-522213,
Andhra Pradesh, India.
[email protected]

Abstract recommended according to the maximum number of

With the increase in the demand for the e-commerce ratings given by the user in a genre.
websites, large amounts of information is available
due to which the users face difficulty in finding the In the Collaborative Based Filtering[5][6],
relevant information matching their preferences. recommendation for a user is governed by other
Recommender systems solve this problem by users’ profiles. An item is recommended based on the
searching through large volume of dynamically ratings of other users who have similar interests as
generated information to provide users with the user under consideration[7][8]. In another
personalized content and services. There are many approach, the content and collaborative based
machine learning techniques which can be used to filtering are combined to form the Hybrid Content-
realize the recommendation system. Among all these Collaborative Based Filtering. It includes the
techniques we are dealing with Content Based advantages of both the methods [19][20]and
Filtering, Collaborative Based Filtering, Hybrid outperforms both of them. In another approach, the
Content-Collaborative Based Filtering. Thus, we content and collaborative based filtering are
represent a system which will recommend products to combined to form the Hybrid Content-Collaborative
the user based on the customer reviews. Based Filtering[9][10]. It includes the advantages of
both the methods and outperforms both of them.
Keywords: Recommender system, Content based, Cold-start problem is one of the most commonly
Collaborative based, Machine Learning encountered challenges of the recommendation
system. It is also known as the new user problem as it
1. INTRODUCTION creates problem of generating recommendations for
the new user.
Recommendation system is an application which is
used for prediction in various domains throughout the When investigating recommendation mechanisms of
internet. A large amount of data flows through the the mobile version of an online shopping website for
internet and it gives away a lot of information coupon deals, it is found that, in real business case,
regarding the user searching activity[1][2]. The the information about customers is limited[11][12].
information extracted from the pattern of previously In most cases, information about customer
searched data can be molded into the prediction of profile,[17][18] product preference or product
relevant data for the user. The implementation of the ratings, is not available. The most relevant
system can be performed by various techniques. information about a particular customer is her
Different techniques like Content Based Filtering, purchase history and current location.
Collaborative Filtering based are used. The Content
Based Filtering approach takes into account a user’s A straightforward approach is to recommend deals
profile which is constructed based on his previous that are in the vicinity[13][14] Such recommendation
ratings. His ratings determine his inclination and will suffer from a higher probability of inaccuracy.
interests, forming the basis for recommending a new With regard to this problem, we propose a novel
item. A higher rating denotes a higher likelihood of approach to analyze the purchase history of
the user to visit similar items[3][4]. So, a new item is customers to find their purchase patterns, and apply
them to location based recommendations. The

www.jespublication.com Page No:515

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

objective is to achieve higher accuracy for the

recommendation system[15][16].

Fig.1 Recommendation System Architecture

2. LITERATURE SURVEY are usually derived from database of previous
There are generally four types of recommendation transactions[33][34].
systems: content-based filtering, collaborative Anveshini Dumala[31] et al. discovered localized
filtering, rule-based and hybrid approaches[21][22]. association rules among items that are purchased
Content-based filtering system performs together, which are helpful for target marketing. R S
recommendation in steps. First, an item profile M Lakshmi [27]et al. used data mining techniques
consisting of a set of features is extracted for each such as web usage mining, decision tree induction,
item. Second, user profiles are generated based on association rule mining to build a rule-based
features of items that are purchased by each user. recommendation system& proposed a sequential
Then similarity scores between user profiles and item pattern based system that predicts customer's time-
profiles are calculated to finally recommend items invariant purchase behavior for food items in a super
with top similarity scores[23][24]. This technique market [41] [42]. Hybrid approach combines multiple
was used to recommend documents such as news, techniques to overcome the limitations of individual
web pages, movies and books . The features used in systems. For example, a linear combination of rule-
building profiles are often a set of keywords. based and collaborative filtering recommendations
[25][26]The limitations of these systems are that considers customers purchase sequences over time as
recommendations are always similar to items that a well as their purchase data for latest period .When
user has already purchased, and it is difficult to applying these recommendation techniques to mobile
recommend items for new users. environment, we are facing some new challenges as
well as new opportunities[35][36].
Collaborative filtering systems are built on the
assumption that a good way to find interesting The mobile screens are smaller and thus few
content is to find other people who have similar products are browsed. As a result, it is difficult to use
interests and then recommend items rated highly by users' browsing behavior in methods like
those similar users[27][28] . There are two general collaborative filtering. However, more information
classes of collaborative filtering algorithms. Memory can be obtained from users' mobile devices, such as
based algorithms operate over the whole user location, dwell time and direction [43] [44]. Location
database to make recommendations[29][30]. has been widely used in mobile environment to
provide value-add services. For recommendation
Model based collaborative filtering, in contrast, systems related to location based business such as
learns a model from the user database, which is then restaurants or tourist attractions, the information of
used for recommendations[31][32]. Collaborative location can be helpful[37][38].
filtering recommendation systems, however, also
have limitations that it is difficult to recommend For example, Lakshman Narayana [12]et al. aimed at
items for new users, to recommend items which have recommending restaurants. Their system use a
not been rated before, and to recommend when rating Bayesian network that models the probabilistic
information is insufficient. Rule-based approach is a influences of the user's personal profile and
simple but popular way of recommendations. Rules contextual information on the restaurant attribute
values. K.Santhi Sri[19]et al. used location history of

www.jespublication.com Page No:516

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

a user to derive personal preferences and social attributes of items in order to generate predictions.
opinions mined from local experts to facilitate When documents such as web pages, publications
personalized location recommendations [45] [46]. In and news are to be recommended. Content-based
this paper, we are targeting at a problem with limited filtering technique is the most successful. In content-
amount of user information[39]. User profiles, based filtering technique, recommendation is made
preferences, location histories, item ratings are not based on the user profiles using features extracted
available. Furthermore, in the mobile shopping from the content of the items the user has evaluated
website for coupon deals, the life span of deals is in the past. Collaborative filtering is a domain-
only several days. All these limitations and independent prediction technique for content that
constraints make it difficult to apply existing cannot easily and adequately be described by
techniques to build recommendation systems [47] metadata such as movies and music. Collaborative
[48]. Therefore, we are proposing a novel approach filtering technique works by building a database of
to analyze the purchase history of the user, find their preferences for items by users. It then matches users
purchase patterns related to the user behavior and with relevant interest and preferences by calculating
then predict information of the next possible purchase similarities between their profiles to make
in a particular location[40]. recommendations.
Collaborative filtering techniques can be
3. PROPOSED METHOD  Model-based filtering technique
The proposed method uses the content based,  Memory-based filtering technique
collaborative based and hybrid filtering techniques  User-based technique
to achieve accurate recommendations. Content-  Item-based technique
based technique is a domain-dependent algorithm
and it emphasizes more on the analysis of the

Fig.2.ProposedFramework
4. RESULTS strict setting is used in the experiment, such that the
To make recommendations to a particular user in the second part only contains 1 transaction. Assume that
test data set, the user’s transaction history is divided a user is located near a set of available coupon deals.
into two parts. The first part is considered as The experiment aims to predict the correct category
purchase history and it is used by the conditional of the user’s next purchase, so as to recommend the
probability model to make recommendations. The deals that are the most likely to be accepted.
second part is the next purchase after first part, and it
is considered as recommendation ground truth. A

www.jespublication.com Page No:517

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

Fig.3.Sample Code

Fig.4. Main Category Classification

Fig.5.Second Category Classification

www.jespublication.com Page No:518

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

Fig.6. Tertiary Classification

5. CONCLUSION of Packet Drop using 2ACK scheme in
MANETs”, Smart Intelligent Computing and
Due to fast moving lifestyle, online shopping has Applications, Vo1.1, pp.649-658. DOI:
been growing drastically through out the world. The 10.1007/978-981-13-1921-1_63 DOI:
increase of online shopping has become a trendy way 10.1007/978-981-13-1921-1_63
for consumers to shop over internet. Machine [5]. Lakshman Narayana Vejendla and Bharathi C
learning techniques provide recommendation R, (2018), “Effective multi-mode routing
mechanisms to identify customer interests. Using mechanism with master-slave technique and
recommendation techniques we can find the frequent reduction of packet droppings using 2-ACK
products purchased by the customers and suggest scheme in MANETS”, Modelling,
products to customers or users. Measurement and Control A, Vol.91, Issue.2,
pp.73-76. DOI: 10.18280/mmc_a.910207
References [6]. Lakshman Narayana Vejendla , A Peda Gopi
and N.Ashok Kumar,(2018),“ Different
[1]. Lakshman Narayana Vejendla and A Peda techniques for hiding the text information
Gopi, (2019),” Avoiding Interoperability and using text steganography techniques: A
Delay in Healthcare Monitoring System Using survey”, Ingénierie des Systèmes
Block Chain Technology”, Revue d'Information, Vol.23, Issue.6,pp.115-
d'Intelligence Artificielle , Vol. 33, No. 1, 125.DOI: 10.3166/ISI.23.6.115-125
2019,pp.45-48. [7]. A Peda Gopi and Lakshman Narayana
[2]. Gopi, A.P., Jyothi, R.N.S., Narayana, V.L. et Vejendla (2018), “Dynamic load balancing for
al. (2020), “Classification of tweets data based client server assignment in distributed system
on polarity using improved RBF kernel of using genetic algorithm”, Ingénierie des
SVM” . Int. j. inf. tecnol. (2020). Systèmes d'Information, Vol.23, Issue.6, pp.
https://doi.org/10.1007/s41870-019-00409-4. 87-98. DOI: 10.3166/ISI.23.6.87-98
[3]. A Peda Gopi and Lakshman Narayana [8]. Lakshman Narayana Vejendla and Bharathi C
Vejendla, (2019),” Certified Node Frequency R,(2017),“Using customized Active Resource
in Social Network Using Parallel Diffusion Routing and Tenable Association using
Methods”, Ingénierie des Systèmes d' Licentious Method Algorithm for secured
Information, Vol. 24, No. 1, 2019,pp.113-117.. mobile ad hoc network Management”,
DOI: 10.18280/isi.240117 Advances in Modeling and Analysis B,
[4]. Lakshman Narayana Vejendla and Bharathi C Vol.60, Issue.1, pp.270-282. DOI:
R ,(2018),“Multi-mode Routing Algorithm 10.18280/ama_b.600117
with Cryptographic Techniques and Reduction

www.jespublication.com Page No:519

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

[9]. Lakshman Narayana Vejendla and Bharathi C [17]. R S M Lakshmi Patibandla, Santhi Sri Kurra
R,(2017),“Identity Based Cryptography for and N.Veeranjaneyulu, (2015), “A Study on
Mobile ad hoc Networks”, Journal of Real-Time Business Intelligence and Big
Theoretical and Applied Information Data”,Information Engineering, Vol.4,pp.1-6.
Technology, Vol.95, Issue.5, pp.1173-1181. [18]. K. Santhisri and P.R.S.M. Lakshmi,(2015), “
EID: 2-s2.0-85015373447 Comparative Study on Various Security
[10]. Lakshman Narayana Vejendla and A Peda Algorithms in Cloud Computing”, Recent
Gopi, (2017),” Visual cryptography for gray Trends in Programming Languages
scale images with enhanced security ,Vol.2,No.1,pp.1-6.
mechanisms”, Traitement du Signal,Vol.35, [19]. K.Santhi Sri and PRSM Lakshmi,(2017),
No.3-4,pp.197-208. DOI: 10.3166/ts.34.197- “DDoS Attacks, Detection Parameters and
208 Mitigation in Cloud Environment”,
[11]. A Peda Gopi and Lakshman Narayana IJMTST,Vol.3,No.1,pp.79-82.
Vejendla, (2017),” Protected strength approach [20]. P.R.S.M.Lakshmi,K.Santhi Sri and Dr.N.
for image steganography”, Traitement du Veeranjaneyulu,(2017), “A Study on
Signal, Vol.35, No.3-4,pp.175-181. DOI: Deployment of Web Applications Require
10.3166/TS.34.175-181 Strong Consistency using Multiple Clouds”,
[12]. Lakshman Narayana Vejendla and A Peda IJMTST,Vol.3,No.1,pp.14-17.
Gopi, (2020),” Design and Analysis of CMOS [21]. P.R.S.M.Lakshmi,K.Santhi Sri and
LNA with Extended Bandwidth For RF M.V.Bhujanga Ra0,(2017), “Workload
Applications”, Journal of Xi'an University of Management through Load Balancing
Architecture & Technology, Vol. 12, Issue. Algorithm in Scalable Cloud”,
3,pp.3759-3765. IJASTEMS,Vol.3,No.1,pp.239-242.
https://doi.org/10.37896/JXAT12.03/319. [22]. K.Santhi Sri, P.R.S.M.Lakshmi, and
[13]. Chaitanya, K., and S. M.V.Bhujanga Ra0,(2017), “A Study of
Venkateswarlu,(2016),"DETECTION OF Security and Privacy Attacks in Cloud
BLACKHOLE & GREYHOLE ATTACKS IN Computing Environment”,
MANETs BASED ON IJASTEMS,Vol.3,No.1,pp. 235-238.
ACKNOWLEDGEMENT BASED [23]. R S M Lakshmi Patibandla and N.
APPROACH." Journal of Theoretical and Veeranjaneyulu, (2018), “Explanatory &
Applied Information Technology 89.1: 228. Complex Analysis of Structured Data to
[14]. Patibandla R.S.M.L., Kurra S.S., Mundukur Enrich Data in Analytical Appliance”,
N.B. (2012), “A Study on Scalability of International Journal for Modern Trends in
Services and Privacy Issues in Cloud Science and Technology, Vol. 04, Special
Computing”. In: Ramanujam R., Issue 01, pp. 147-151.
Ramaswamy S. (eds) Distributed Computing [24]. R S M Lakshmi Patibandla, Santhi Sri Kurra,
and Internet Technology. ICDCIT 2012. Ande Prasad and N.Veeranjaneyulu, (2015),
Lecture Notes in Computer Science, vol “Unstructured Data: Qualitative Analysis”, J.
7154. Springer, Berlin, Heidelberg of Computation In Biosciences And
[15]. Patibandla R.S.M.L., Veeranjaneyulu N. Engineering, Vol. 2,No.3,pp.1-4.
(2018), “Survey on Clustering Algorithms [25]. R S M Lakshmi Patibandla, Santhi Sri Kurra
for Unstructured Data”. In: Bhateja V., and H.-J. Kim,(2014), “Electronic resource
Coello Coello C., Satapathy S., Pattnaik P. management using cloud computing for
(eds) Intelligent Engineering Informatics. libraries”, International Journal of Applied
Advances in Intelligent Systems and Engineering Research, Vol.9,pp. 18141-18147.
Computing, vol 695. Springer, Singapore [26]. Ms.R.S.M.Lakshmi Patibandla Dr.Ande
[16]. Patibandla, R.S.M.L., Veeranjaneyulu, N. Prasad and Mr.Y.R.P.Shankar,(2013),
(2018), “Performance Analysis of Partition “SECURE ZONE IN CLOUD”, International
and Evolutionary Clustering Methods on Journal of Advances in Computer Networks
Various Cluster Validation Criteria”, Arab J and its Security, Vol.3,No.2,pp.153-157.
Sci Eng ,Vol.43, pp.4379–4390.

www.jespublication.com Page No:520

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

[27]. Patibandla, R. S. M. Lakshmi et al., (2016), OPNET Modeler”, International Journal of

“Significance of Embedded Systems to IoT.”, Computer Science Trends and Technology
International Journal of Computer Science and (IJCST) – Vol. 4 No. 5, pp. 38-42.
Business Informatics, Vol.16,No.2,pp.15-23. [35]. VellalacheruvuPavani and I. Ramesh Babu
[28]. AnveshiniDumala and S. PallamSetty. (2019) ,”Three Level Cloud Storage Scheme
(2020),“LANMAR routing protocol to support for Providing Privacy Preserving using Edge
real-time communications in MANETs using Computing”,International Journal of Advanced
Soft computing technique”, 3rd International Science and Technology Vol. 28, No. 16, pp.
Conference on Data Engineering and 1929 – 1940.
Communication Technology (ICDECT-2019), [36]. VellalacheruvuPavani and I. Ramesh Babu,”A
Springer, Vol. 1079, pp. 231-243. Novel Method to Optimize the Computation
Overhead in Cloud Computing by Using
[29]. AnveshiniDumala and S. PallamSetty. Linear Programming” ,International Journal of
(2019),“Investigating the Impact of Network Research and Analytical Reviews May 2019,
Size on LANMAR Routing Protocol in a Volume 6, Issue 2,PP.820-830..
Multi-Hop Ad hoc Network”, i-manager’s [37]. Anusha Papasani and Nagaraju
Journal on Wireless Communication Networks Devarakonda,(2016),"Improvement of Aomdv
(JWCN), Volume 7, No. 4, pp.19-26. Routing Protocol in Manet and Performance
[30]. AnveshiniDumala and S. PallamSetty. Analysis of Security Attacks", International
(2019),“Performance analysis of LANMAR Journal Of Research in Computer Science &
routing protocol in SANET and MANET”, Engineering ,Vol.6,No.5, pp.4674-4685.
International Journal of Computer Science and [38]. Sk.Reshmi Khadherbhi,K.Suresh Babu , Big
Engineering (IJCSE) – Vol. 7,No. 5, pp.1237- Data Search Space Reduction Based On User
1242. Perspective Using Map Reduce ,International
[31]. AnveshiniDumala and S. PallamSetty. (2018), Journal of Advanced Technology and
“A Comparative Study of Various Mobility Innovative Research Volume.07, IssueNo.18,
Speeds of Nodes on the Performance of December-2015, Pages: 3642-3647
LANMAR in Mobile Ad hoc Network”, [39]. B.V.Suresh kumar,Sk.Reshmi Khadherbhi
International Journal of Computer Science and ,BIG-IOT Framework Applications and
Engineering (IJCSE) – Vol. 6, No. 9, pp. 192- Challenges: A Survey Volume 7, Issue VII,
198. JULY/2018 pg.no 1257-1264
[32]. AnveshiniDumala and S. PallamSetty. [40]. P.Sandhya Krishna,Sk.Reshmi
(2018),“Investigating the Impact of IEEE Khadherbhi,V.Pavani, Unsupervised or
802.11 Power Saving Mode on the Supervised Feature Finding For Study of
Performance of LANMAR Routing Protocol Products Sentiment ,International Journal of
in MANETs”, International Journal of Advanced Science and Technology, Vol 28
Scientific Research in Computer Science and No 16 (2019).
Management Studies (IJSRCSMS) – Vol.7, [41]. K.Santhi Sri, Dr.Ande Prasad (2013), “A
No. 4. Review of Cloud Computing and Security
[33]. AnveshiniDumala and S. PallamSetty. Issues at Different Levels in Cloud
(2016),“Analyzing the steady state behavior of Computing” , International Journal on
RIP and OSPF routing protocols in the context Advanced Computer Theory and Engineering
of link failure and link recovery in Wide Area Vol. 2,pp 67-73.
Network”, International Journal of Computer [42]. K.Santhi Sri, N.Veeranjaneyulu(2018), “A
Science Organization Trends (IJCOT) – Vol. Novel Key Management Using Elliptic and
34 No 2, pp.19-22. Diffie-Hellman for Managing users in Cloud
[34]. AnveshiniDumala and S. PallamSetty. Environment”, Advances in Modelling and
(2016),“Investigating the Impact of Simulation Analysis B,Vol.61,No.2,pp 106-112.
Time on Convergence Activity & Duration of [43]. K.Santhi Sri, N.Veeranjaneyulu(2019),
EIGRP, OSPF Routing Protocols under Link “Decentralized Key Management Using
Failure and Link Recovery in WAN Using Alternating Multilinear Forms for Cloud Data

www.jespublication.com Page No:521

 Vol 11, Issue 4 , April/ 2020
ISSN NO: 0377-9254

Sharing with Dynamic Multiprivileged

Groups”, Mathematical Modelling of
Engineering Problems,Vol.6,No.4,pp511-518.
[44]. S.Sasikala, P.Sudhakar, “interpolation of CFA
color Images with Hybrid image denoising”,
2014 Sixth International Conference on
Computational Intelligence and
Communication Networks, DOI 10.1109/.53
193 DOI 10.1109/CICN.2014.53, pp. 193-197.
[45]. Me. Jakeera Begum and M.Venkata Rao,
(2015), “Collaborative Tagging Using
CAPTCHA” International Journal of
Innovative Technology And Research,
Volume No.3, Issue No.5,pp,2436 – 2439.
[46]. L.Jagajeevan Rao, M. Venkata Rao, T.Vijaya
Saradhi (2016), “How The Smartcard Makes
the Certification Verification Easy” Journal of
Theoretical and Applied Information
Technology, Vol.83. No.2, pp. 180-186.
[47]. Venkata Rao Maddumala, R. Arunkumar, and
S. Arivalagan (2018)“An Empirical Review on
Data Feature Selection and Big Data
Clustering” Asian Journal of Computer
Science and Technology Vol.7 No.S1, pp. 96-
100.
[48]. Singamaneni Kranthi Kumar, Pallela Dileep
Kumar Reddy, Gajula Ramesh, Venkata Rao
Maddumala, (2019), “Image Transformation
Technique Using Steganography Methods
Using LWT Technique” ,Traitement du
Signalvol 36, No 3, pp. 233-237.

www.jespublication.com Page No:522