The main problems we found out in this research project is people find hard to buy products from pettah. They have lack of knowledge about the pettah market where sometimes they get scammed by the vendors buying low quality products for high prices. And, customers don’t know the exact streets in pettah to buy particular products they need. As we know in pettah we can find anything and purchase it for best price less than the market price outside of pettah. Customers also face problems like the cost of transportation, the lack of parking facilities and the frequent traffic congestion, which discourage customers to visit the pettah market to purchase the products they need. We also found out there is language barriers between vendor and customer to understand the name of the product they need. So, we have decided to give an overall solution for this kind of problem we have come through.
The solution proposed is to build a mobile application for the customers to search products using images with the help of image classification machine learning model to find out the similar products they need from a vendor in pettah market. They also can take a picture manually or upload an image in the mobile application.
Also, they can search the product using the voice input of three languages such as Sinhala, English and Tamil. With the help of natural language processing machine learning model, the system will generate the results according to the search products results they needed.
To make it easier for the customer we will also map the store location once the customer wants to buy the product physically from the store, so customer knows the exact location of the store or else they can even purchase it online through the application.
Novelty of this project is currently none of the shopping platforms in Sri Lanka provide this kind of solutions to the customers. To let people, search products by using images and their own voice input. Currently they use only text-based search. Also, some of the famous worldwide shopping platforms use image search and voice search with the help of google plugins. But in this research, we will be providing a solution using a custom algorithm to make our product unique.
Main objective is to build a mobile application to the customers to make the shopping in pettah easy and find out the products they need for the best price. With the help of Image classification, they can find products using an image search and with the help of natural language processing customer can search the product using their own voice through any languages they are well known (Sinhala, English and Tamil)
Search product through Image Inside the platform (Image classification)
a. User can upload picture
b. User can open the camera and take a picture of the needed product
Search product with any language using the voice input (English, Tamil, Sinhala) with the help of NLP (Natural language processing)
a. If a person searches the products using the voice with a Sinhala or Tamil language and the product is in English language, automatically the system should translate the name and search for the product. example if I search the word kiri (කිරි) using the voice input the system should know it means milk and listed milk products should appear.
