Sentiment Survey

This project implements a simple sentiment survey. Respondents are shown a series of tweets and asked to determine the amount of positive and negative sentiment.

Tweets are picked randomly from a CSV file (which is not made publically available). Each tweet evaluation is stored in another CSV file for further analysis. Following file paths are used:

• ./data/tweets.csv for input data
• ./data/result.csv for the sentiment evaluation results

This sentiment analysis survey is a part of my dissertation research which focuses on using social media data, sentiment analysis and topic modeling as decision support tools for municipal decision-making.

The survey is indended for respondents from Czechia so the interface is in Czech.

Starting the Application

Option 1: NodeJS

$ npm install
$ node app.js

The app now runs at http://localhost:3000

Option 2: Docker

$ docker build -t sentiment-survey .
$ docker run -p {HOST_PORT}:3000 -v {DATA_DIR}:/app/data sentiment-survey

The app now runs at http://locahost:{HOST_PORT}