It's estimated that 34% of E-Commerce users don't eventually make a purchase, that's mainly attributed to the lack of competitive pricing and costs the user a lot of time and effort trying to find the best product with the best value all while being distracted by lots of irrelevant suggestions.

What it does

  • World Market digs through a variety of E-Commerce websites listing different prices for the same product and exposing e-commerce websites that aren't necessarily well known to most users.
  • World Market analyzes search output offering a minimal interface showing 3 options: Cheapest, Most Expensive and Top Rated product after analyzing the search output.
  • Semantically evaluate the search query and search for products in the corresponding segment (e.g. Only search for top rated phones if query is "high quality headphones").
  • Offer on-site chat bot for enhanced user experience.

How we built it

We have built the backend server with django to retrieve the product data and do the required analysis. The frontend is built with react framework. We have used Microsoft language understanding service; LUIS to extract semantics from the search queries and accordingly decide what quality level to map the search to. We also use LUIS to facilitate our chat bot logic.

Challenges we ran into

Communicating with django from react required a variety settings adjustments that we weren't previously familiar with. We have also found some challenges when it came to retrieving data from a variety of resources and then building a standard data structure to handle all of them.

Accomplishments that We're proud of

Being the first e-commerce service to semantically understand search queries. Building our application while considering business and commerce limitations.

What's next for World Market

Since we have a chat bot that facilitates direct human interaction, we aim to expose the chat bot to a call service to allow users that don't have internet access to reach e-commerce solutions.

Share this project: