Inspiration
Nowadays, we can not reject that chat is one big part in our daily life. We use chat for work, for education, for appointment or event for shopping.
Let's focus on shopping, when we desire to buy something, most of the time we always ask a simple question like shown below:


From the example we found that it take a long response time. Therefore the merchant might lost their customer.
So we introduce a chatbot that response to the customer as soon as possible with useful information.
What it does
Our chatbot will ask for the basic information from customer for example, product category, brand, model or customer's budget. After gathering enough information, chatbot will recommend product best product based on customer's intput.
How we built it
We build it with 2 services with are
Chat engine
- For smart dialog ####Recommendation engine
- For recommend product

Challenges we ran into
- Multi channel for chat
- Find appropriate data set
- Time limit
Accomplishments that we're proud of
- Engine is work!
- Our chatbot can recommend products to user.
What we learned
- We learned how to do content-based recommendation system
- We learned how to cleansing data(NaN)
What's next for SID TECH TALENT CHAT
- Looking for Thai dataset
- Add more features for product classification
- Tuning accuracy
Built With
- node.js
- platform
- python
- scikit-learn
Log in or sign up for Devpost to join the conversation.