We were inspired by an idea that sometimes in such a big city as Moscow people sometimes want to go out somewhere, but don't know in advance, where they want to go. So our dream is to build a service that could reflect interesting things that take place in the city using peoples' feedback in social nets such as twitter or instagram -- people often post tweets or make selfies while they're in public places. So, retrieving moods of the posts, we could build a look at the event so other people could understand if it is worth going there.
What it does
At this 12 hours hackathon we have build a scratch of an application that is able to show moods of people on the map, using tweets.
How we built it
As a background map we used Yandex Maps. To get users sentiment data we used Twitter API to get twits and then applied modified Russian word2vec to it to get mood of a tweet.
Challenges we ran into
Working with new API of Yandex Maps, working with Russian version of word2vec and generating an algorithm that could reflect tweets' sentiments.
What's next for Mood Map
- Start using other social networks for retrieving data, such s Instagram.
- Moving in a direction of making idea described in "Inspirtion" section come true: start recognizing events using posts and understanding their mood.