Inspiration

Since I started getting socially active online a few years ago, I have joined so many online communities that allowed me to communicate with people who have similar interests as mine. Although most of the people I got to know through online communities lived thousands of kilometers away, I have learned a lot from them and their experiences.

For example, about a year ago, I have joined a Facebook group where members share their work and experience working on machine learning projects. Consequently, in the following months, I got to learn a lot more about the field of machine learning and the community members were very helpful and willing to help and encourage people to learn about machine learning and how it works.

However, recently I have noticed that the group members weren't as active as they used to and it was frustrating to me since I truly enjoyed checking the group everyday but I wasn't able to contribute to the community by sharing my work because I don't have much experience in the field.

Therefore, I decided to create Moy. A chatbot built to help pages and groups to stay active by suggesting relevant content to share & discuss with their community members!

What it does

Moy is a Messenger chatbot and a web app that's built specifically to help communities to stay connected by suggesting relevant content to share & discuss with their community members.

The Messenger bot is an easy-to-use tool that requires minimal commands to take full advantage of Moy. When Moy receives its first message from a user, it responds with a welcome message and requests the user to add a topic of their interest. When Moy receives the user input, it runs a web crawler to find news and content that can be relevant to the topics the user is interested in.

After finding relevant content, Moy curates the content in a simple style and sends the user the latest content with a URL to the full list of content that may be relevant to share with their online community!

How we built it

Using the Facebook Graph API and Messenger API we were able to communicate with Facebook's servers to send and receive messages from Moy's subscribers.

Then we've developed an NLP class in addition to using the NLP information given by Messenger to perform text & context analysis of the users' input.

After done analyzing the text and parsing out the needed information we run a web crawler to find news and trending content that's relevant to the users' interests. We've used and open-sourced PHP class called crawler in order to find relevant content. Additionally, we used the NewsAPI to find trending news topics that's relevant to users' interests.

Moy, both the Messenger bot and the web application, are written in PHP 7 and used SQL to enhance the bot's user experience!

What's its impact on communities

Personally, as a student who have learned and benefited a lot from online communities and was able to find my passion because of the communities I have joined in the past few years. I believe there are many people around the world just like me, that one day a community they'd join will change everything in their lives to the best!

Hence, Moy will encourage community members to engage in the groups and pages they join and be up to date to the latest news and trends that's relevant to their interests. Which one day may have a great impact on someone's passion.

Challenges we ran into

My brother and I worked together on this project while we are thousands of kilometers away from each other. It was quite a challenge to plan how the product will work and assign tasks. However, thanks to Messenger and Trello, we were able to efficiently work together and get the first version of Moy ready and working!

Additionally, I got to use PHP and the Facebook API for the first time and it was a challenge to learn a new syntax and how Facebook's API works in a very short amount of time. Yet, I really enjoyed learning a whole new programming language and the challenge of using the language to build a new product from scratch :)

Accomplishments that we proud of

We are truly proud that our product may have a big impact on many people's lives and how my brother and I collaborated efficiently to finish the first version of Moy.

What we learned

While working on Moy, we both got to learn new concepts in machine learning and natural language processing to improve the bot's understanding of the user's inputs.

Additionally we used what we learned to create a PHP class to analyze page & group content to extract keywords and topics that would be relevant to community members' interests.

What's next for MoyBot

At this moment the published version of MoyBot only uses the user manual input of the topics they're interested in. However, because we've already developed a tool that would extract keywords and topics from a page's/group's content, in the next version we will include a feature that would allow users to simply Login with Facebook and select a group or page they'd like to share content on and get news & trend that are relevant to the community's interest.

Built With

Share this project:

Updates