There are a lot people around us using cigarettes. It’s harmful for them, the people around them and also the environment. Most young people are not aware the danger and some of them want to quit but are not motivated enough. This also acts as a anonymous data collection platform to help people quit smoking.

What it does

It’s a chatbot that helps people to quit smoking. It excels to inform people of the dangers of smoking in a conversational manner, we help them to know better about the cigaratte and its influence. For people who have already aware the danger and want to quit, we provide them daily inspiration once with an option to subscribe. We also provide ways to get detoxed, and link to research and self help videos. It even helps you calculate how much you will save if you stop smoking now. One of the best features is the join chatroom, where users of the bot can chat with each other and talk about their progress and essential help each other. We have built and educative, conversational bot. We connect users with experts on anti smoking on call. Along with which the bot takes feedback from the user and stores it in the mongodb for future analysis. Currently we do sentiment analysis on this feedback to gather more data.

How we built it

Chatfuel + python + mongodb + sentiment analysis

Challenges we ran into

Conversational UI is a tricky subject, getting it right was a challenge. How to utilize feedback data, it is a paragraph. We ended up doing sentiment analysis for a start.

Accomplishments that we're proud of

Created from scratch and deployed. Working on a social cause feels great. We can feel we are making a difference.

Sentiment analysis on feedback, utilizing machine learning is always fun.

What we learned

Using mongodb Using chatfuel Creating intuitive conversational interfaces Collecting data for analysis

What's next for StopSmoking

Use feedback from users about the bot, about their quitting status, about their calls with experts and do data analysis on that. Such data analysis will help us help our users better.

Share this project: