Grocery Recommender System / Naajihah Mustapa
Retail industry is the largest contribution to the country but with the immerse growing of the retail industry, an increasing number population and unstoppable growth of technology, the retail industry needs a solution to keep maintain their business. Instead of staying offline purchase, they have t...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/95082/1/95082.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Retail industry is the largest contribution to the country but with the immerse growing of the retail industry, an increasing number population and unstoppable growth of technology, the retail industry needs a solution to keep maintain their business. Instead of staying offline purchase, they have to venture into online business. In early 2020, a Covid-19 outbreak has changed everything. A new norm needs to be adopted by everyone. Since then, supermarkets, hypermarkets, and other store departments start to sell online due to restricted movement control orders by promoting on a social media platform and some of the big companies even create an application that allows the user to place an order and deliver on the same day and people tend to spend less time in the store. However, the implementation still has their limitation, the number of daily products only could reach a thousand with different brands lines. As a user, it would be helpful, if the website/application could help them to decide to choose the products. A recommender system is a well-known system that has been used for a long time, the advantages of the recommender system has given benefits on both side, retailers, and consumer. To understand how the recommender system work, a preliminary study helps to understand more and choose the best algorithm to be implemented into the application. A collaborative filtering algorithm is selected. The development of the application is successfully developed with the result of precision 94%, recall 88%, and f-measure 91%. |
|---|