tweetsent

Tweetsent is a simple, configurable Twitter bot designed to provide sentiment analysis for tweets. It uses a simple mashup of three modules:

Twit to query Twitter's REST service
Alchemy to perform sentiment analysis
Mongoose to store the results in a Mongo Database

An important note about the Alchemy API: Free access is limited to 1,000 requests/day.

Features

Highest impact: In order to provide the maximum benefit of its usage, tweetsent does some simple calculations (based on the number of followers and retweets) to include only tweets with the highest impact for analysis. If, for example, the bot retrieves 100 tweets during one interval, it will only select the twitter_max_tweets number for analysis.
No repetition: Each interval, the bot tracks where it left off, and will only request new tweets for analysis.

Dependencies

You'll need:

Twitter API credentials
Alchemy API key

Installation

$ git clone git@github.com:danielkennedy/tweetsent.git $ cd tweetsent $ npm install

Configuration

Remember that free usage of the Alchemy API is limited to 1,000 requests/day. That means you'll need to strike a balance between up-to-date data (frequent intervals) and tweet volume (max tweets). I like to set twitter_max_tweets to 2 and twitter_request_interval to 3.

$ cp config.js.sample config.js $ vim config.js

Operation

$ node bot