Application of Hybrid Algorithms for Car Recommendation System

Abstract

Web based recommendation for any item is a hard and fast requirement in any E-commerce website. This research paper explores the system which is employed to recommend car to the users based on the requirement provided by the user. This methodology narrow the sphere right down to some criteria that car buyer consider like looks , cost, safety, functionality, performance , fuel economy, technology etc. A hybrid recommendation system is a special variety of recommendation system which might be considered as the combination of the content and therefore the collaborative filtering method.

Introduction

I. INTRODUCTION

Recommendation is such a thing that is not new to our society. This had always been a component of humankind where humans used to recommend their colleagues better way for decide on things .As globalization is a new normal, there are plentiful of products in every category. Recommendation techniques have always played a very crucial role in marketing activities. If we talk about vehicles, recommendation is a much needed thing for today’s buyers. People’s craze for wheels is not a new thing as in the last part of Stone Age; the invention of wheel was done. In fact wheels are the proof of intelligence of humans. But the wheel alone, without any further coinage, would not have done much for the mankind. As the time passes, technology grew and stands on what we see today. Owning a vehicle has become a mandatory requirement in the modern world. Automobile industries are investing a lot on producing different car models to cater the needs of their customers with different social and economic backgrounds. The foremost objective of this work is to recommend a car according the user need.

A recommendation system is a subclass of Information filtering Systems that looks for to predict the rating or the preference a user might give to an item. "Information Filtering" is a field of study designed for creating a systematic and scientific approach for extracting information that a particular person finds important from a larger stream of knowledge and  information. Recommendation system is active information filtering system that seeks to give the user information items the user is specifically interested in.

By its design , recommendation system is personalized system for user. Relevant items are shown using the content of the previously searched items for the users. This is content based filtering. Recommending items to users based on their interest and preference of other similar users is basically collaborative-based filtering. For eg:- When we shop on Amazon it recommends new products saying “Customer who brought this also brought”.

Recommendations are of two type: personalized and Non personalized. While personalized recommendation system suggests products to a user according to the user profile and their previous purchase history, a non-personalized recommender system displays products that are popular among the people in general during the time period. Hence, we can see, the more complete and precise each user profile leads more successful is the recommendation process.

A. Why Recommendation System

1. Customer Satisfaction: Many a time , customer gets product recommendation from their past browsing. This way they are guided to compare their thoughts and knowledge to get best of the thing they are searching.
2. Personalization: Seeking help from friend and family and then purchasing  a product is many a time influenced by their thoughts rather than customers own requirement and budget. Recommendation system help reducing the distractions.
3. More Sales for the Company: With the help of recommendation system companies are benefited with loyal customers . Also this helps them to identify what customers really need.

B Contribution to the System

1. Users: In order to achieve personalization, different parameters like ratings of the user, demographic attributes like age, gender, profession, income,place etc,
2. Behaviour: attributes like browsing pattern, click stream data, filtering pattern etc of the users are involved in the design of the system.
3. Models: In car recommend system client also check other features like car name,brand ,model,current model,fuel type and price

The dataset collected may not always be in a format that is suitables for machine learning algorithms to operates on RS.

Finally, values of the feature fuel type were mapped to four broader values according to their needs. They are cng and oil, pentrol, hybrid, and lpg and Oil.example: Lpg,octane, Petrol, diesel has been mapped to a broader categorys oil. car that run with both cng and pentrol are mapped to category cng and pentrol.. Other categories are like mapped in a similar manners

1. Content-based: In this system items are recommended that are similar to   items that the user have  searched or bought in past . The possible similarity of product  is calculated based on the feature which is very near to the demands of the user in the new one.
2. Collaborative Filtering System: It is called as “client-to-client correlation.” This is probably the most popular method of recommendation. This method is based on kind of partnership method where user and items are closely related to have recommendation.
3. Demographic: This works on the basis of the profile of the user . Profile of the user means what occupation he has, what is his age and also the geography around the user.
4. Knowledge-Based: Knowledge based recommendation system is most powerful tool if used properly. In this method different constraint is fed to the machine like which qualities of the item is best suited to the user. This actually works on data. Hence have a major drawback, that if data is not fed properly or not implemented as it is needed it may even produce false results,
5. Constraint-Based Systems: This system is very much similar to the knowledge base recommendation system. Here the recommendation works on some specific set of rules like what are the demands of user about car type, fuel type, cost of the car, space, and much more.
6. Community-Based: In this system the preference to the community of buyer is given.
7. Hybrid Recommendation System: This system is mixture of above all recommendation systems in appropriate amounts.

III. TEST CASES AND RESULTS

Clicking on “option” in the MENU FRAME brings you on this frame. This window looks as follows:

1) This window provides suggestion of car(s). User needs to inputs .

Conclusion

The hybrid recommendation algorithm is efficient in suggest recommendations. This system mainly helps users to purchase car without more knowledge about car, also suggests many option available based on the customer requirement. It can correctly suggest the user car model based on their requirement. To provides accurate and current information to the client, links are provided for each car model for user can know review, detail, image of model. The user can click on the link and there by update information on the internet. This is useful as the users can view the images, view of the car, reviews of other users, ratings, details, etc. Data Visualizations help to see the analysis of the available car models at a glances. If user is low on budget, there is an option to view Car Loans of banks and direct link to get. This system can surely help to choose the appropriate car, meeting the users requirement.

References

[1] A.J. Patel, J.S. Patel, Ensemble systems and incremental learning, in: 2013 International Conference on Intelligent Systems and Signal Processing (ISSP), 2013, pp. 365–368. [2] A. Prioletti, A. Mogelmose, P. Grisleri, M.M. Trivedi, A. Broggi, T.B. Moeslund, Part-based pedestrian detection and feature-based tracking for driver assistance: real-time, robust algorithms, and evaluation, IEEE Trans. Intell. Transport. Syst. 14 (3) (2013) 1346–1359. [3] P. Viola, M. Jones, Rapid object detection using a boosted cascade of simple features, in: IEEE Conference on Computer Vision and Pattern Recognition, January 2001, pp. 511–518. [4] L. Shao, X. Zhen, D. Tao, X. Li, Spatio-temporal laplacian pyramid coding for action recognition, IEEE Trans. Cybernet. 44 (6) (2014) 817–827. [5] Ambeth Kumar, V. D., Malathi, S., Venkatesan, R., Ramalakshmi, K., Vengatesan, K., Ding, W., & Kumar, A. (2019). Exploration of an innovative geometric parameter based on performance enhancement for foot print recognition. Journal of Intelligent & Fuzzy Systems, 1–16. [6] S. Sivaraman, M.M. Trivedi, Looking at vehicles on the road: a survey of vision-based vehicle detection, tracking, and behavior analysis, IEEE Trans. Intell. Transport. Syst. 14 (4) (2013) 1773–1795. [7] Kesavan, S., Kumar, E. S., Kumar, A., & Vengatesan, K. (2019). An investigation on adaptive HTTP media streaming Quality-of-Experience (QoE) and agility using cloud media services. International Journal of Computers and Applications. [8] Vengatesan, K., Kumar, A., Naik, R., & Verma, D. K. (2019). Anomaly based novel intrusion detection system for network traffic reduction. Proceedings of the International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud), I-SMAC 2018. [9] M. Feng et al., \"Big Data Analytics and Mining for Effective Visualization and Trends Forecasting of Crime Data,\" in IEEE Access, vol. 7, pp. 106111-106123, 2019. [10] M. Li, H. Wang and J. Li, \"Mining conditional functional dependency rules on big data,\" in Big Data Mining and Analytics, vol. 3, no. 1, pp. 68-84, March 2020. [11] E. Lee et al., \"Game Data Mining Competition on Churn Prediction and Survival Analysis Using Commercial Game Log Data,\" in IEEE Transactions on Games, vol. 11, no. 3, pp. 215-226, Sept. 2019. [12] S. G. Teo, J. Cao and V. C. S. Lee, \"DAG: A General Model for Privacy-Preserving Data Mining,\" in IEEE Transactions on Knowledge and Data Engineering, vol. 32, no. 1, pp. 40-53, 1 Jan. 2020. [13] A. M. Sainju, D. Aghajarian, Z. Jiang and S. Prasad, \"Parallel Grid-Based Colocation Mining Algorithms on GPUs for Big Spatial Event Data,\" in IEEE Transactions on Big Data, vol. 6, no. 1, pp. 107-118, 1 March 2020. [14] Z. Feng, S. Zhu, J. Wu and H. Guo, \"Theory and Method of Time-varying Computational Experiments for the Fully Mechanized Mining Process in an Artificial System Environment,\" in IEEE Access, vol. 7, pp. 168162-168174, 2019. [15] K. Vrotsou and A. Nordman, \"Exploratory Visual Sequence Mining Based on Pattern-Growth,\" in IEEE Transactions on Visualization and Computer Graphics, vol. 25, no. 8, pp. 2597-2610, 1 Aug. 2019.

Copyright

Copyright © 2022 Supriya Pandey , Ujjawal Raj, Sandeep Kumar , Shalini Kanwar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.