Started off with exploring the Power of Wolfram Alpha. Super powerful, got everything. Then we stumbled upon their Facebook Page, and just sent a message to see how smart their ChatBot was. Boom! No reply, not even automated.That's where we realized - Wolfram Alpha is way too powerful, and the world does not know enough. Sure, Wolfram Alpha is big, but there is still scope for it to get bigger. One way for the world to know that is to put them out there, something fancy for the world to see and for the world to realize this power. Why Facebook? Well, Facebook messenger has had a boom in the number of user per hour, and Facebook on the whole has a 50+minute/day average daily users. So you're on facebook, you want to know the weather, or something else for which you have to open google, you wont have to do that anymore. You just switch to Wolfram Alpha chatBot. We believe the data wolfram would collect be richer than the data google collects regarding what humans search, because its through facebook, so basic data can always be shared with the page, to develop good demographics, although the data volume will not be very high compared to google in the initial years. But more than the data, it's something new and it can be used to make Wolfram Alpha even more recognized than it is. I believe it would go viral if launched. People love to see new stuff like this on the internet, and on top of that, this is on Facebook. People pay facebook to have their ads, which most of us ignore, but this would attract people. I see a lot of potential, but it must be acted upon soon. The impact might not be as high if other companies come up with smart bots like this one.
What it does
Technically speaking, It takes input of queries on the Facebook Messenger that are passed to a Natural Language Processing API of Wolfram Alpha, which then looks into its database and retrieves the top results on the Messenger itself. But the answer to what it does,is bigger than that. It shows the world what Wolfram Alpha is capable of. A very small audience knows about the cool API's that wolfram has.
How we built it
We understood the API's of Wolfram alpha and Facebook Developer API. Learnt about Heroku servers and used node.JS to integrate them all together.
Challenges we ran into
Integrating Wolfram alpha and Facebook Dev together was a little tricky.
Accomplishments that we're proud of
Realizing that not every big impact idea has a complex solution was one of the great achievements, apart from everything else I mention below that I'm proud to learn.
What we learned
1)We learned about the Facebook Chat Bot 2)We were surprised to see how strong Wolfram Alpha API is but is not as recognized as it should be. Coursera/Lynda has a lot of courses on Machine Learning, Image processing etc. If they use Wolfram API to teach these subjects to people who do not want to reinvent the wheel, but instead use the what's out there, they should use the Wolfram Alpha API. 3)Learnt a lot of new technologies that were on our "Things-To-Learn" list, but we had not until yesterday.
What's next for Wolfram_Beta ChatBot
1)Improve the way Facebook chat displays and figure out how to display graphs and images on the chat. 2)Have Wolfram Alpha use it on their Facebook page and see it go viral. 3)Analyze the data searched on the ChatBot and relate it to demographics of people using it, which can be used for data analytics. 4)Watch the penetration of Wolfram Alpha rise and see their well documented API's used more in the tech world.