Yellow Pages proposed a challenged to build a Twitter bot that can answer queries to find products and goods.
What it does
The bot replies to tweets asking questions about products and goods that mention #askyp. It can also analyse pictures and attempt to find a business related to its content.
How we built it
The bot was built with node.js. We use the Twitter Streaming API to listen to the hashtags and the REST API to reply to users. The Watson Alchemy Language API is used to analyse the text in a tweet and the Clarifai API is used to analyse pictures to help users find what they are looking for.
Challenges we ran into
Learning and using the Watson API to recognize the text in the tweet. Connecting all the dots together, from the initial tweet to the final reply.
Accomplishments that we're proud of
Being able to experiment and work with new technologies. We're proud of our commitment to deliver a finished product.
What we learned
We learned during this hackathon to create a node.js application and to use various APIs such as Yellow Pages, Watson, Clarifai and Twitter. We were able to explore into query optimization.
What's next for AskYellowPages
Provide more continuous conversations and be able to reply to a wider array of queries.