Twitter Context and Predictive Analysis
The project aims to use context analysis and predictive analysis to solve problems using twitter data.
This team is looking for
In our city today,an important event may be happening on the other side of town and no know it because,a lot of data is being generated and meaning cannot be deduced by simply looking at hashtags and random text. Whatsmore,the bilingual nature of a majority of the populous makes text and sentiment analysis hard and translation of tweets nearly impossible. This is because,most people in the city use a mix of two languages or more and thus,a non city native cannot join a conversation between two people easily. After looking at various solutions to analyze text such as this http://aylien.com/text-api i have created a solution suited for our city that is suited for a multi language environment such as in our city. My solution involves giving keywords context and also pre-empting events courtesy of a predictive analysis component. A user is asked to suggest a keyword and also give some context. For instance,a resident near a known blackspot will create a keyword 'accident' and give a context,i.e 'accidents happen here a lot. A chain of night clubs near here don't make matters easier' . The user can also give the location,nearest hospital,nearest chemist,nearest garage etc We take all this data and store it and keep our bots on alert for the occurrence of the keyword 'accident'. Once a keyword is found,the system will establish the location from the information coming in and matches the various contexts with the current event with accuracy. The system goes ahead,gives advice based on the contexts available. This is the youtube link https://www.youtube.com/watch?v=A4a2CgKnzJY&feature=youtu.be
No comments yet