View on GitHub

Pretrender17

Prediction of trend patterns.

Pretrender

What is it:

> A java based application to predict trend in entities on an upcoming date by leveraging entities trend in the past.
> User can either enter a hashtag or someone's twitter ID to analyze their tweets for its past sentiment trend.
> This trend is then used to plot a pattern which is extrapolated on the future date fed by the user.
> The slope on the future date is then compared for similar slope patterns in the past, which is given as output prediction.
> This is based on the concept that trend pattern for entities tend to repeat.
> UI can either be through java application or by installing apk in the repo.
> Can goto files folder inside src and tweets and their sentiment.

How to run:

> Import Server folder as eclipse General/Existing Projects into workspace
> Get all JAR files in JARs folder as external libraries
> Also get 'stanford-corenlp-3.5.2-models.jar' as file is too big to upload on repo
> Get various developer keys from ur twitter profile - 'https://apps.twitter.com/' and mention them in twitter4j.properties file in src
> Run client.java inorder to access platform through system
> Run server.java to interact with system using phone application
> Pretrender2 folder has phone application that can edited in Android Studio