Sentiment Analysis and Opinion Mining via Twitter (Basic Model)
Introduction & Background
This application is similar to a service built at Stanford University called www.sentiment140.com.
- Data-mines the opinions using social media streams (Twitter in this case)
- Leverages pre-determined sentiment words list to refine analysis
- Can potentially adapt machine learning techniques to further expand its analytical capabilities
This application leverages a list of positive and negative opinion words or sentiment words for English (around 6800 words).
- This list was compiled over many years starting from the authors’ first paper (Hu and Liu, KDD-2004).
- Authors’ website: Sentiment Words Sources: Opinion Mining, Sentiment Analysis, and Opinion Spam Detection
- Credits: Bing Liu and Minqing Hu, http://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html
How Does the Application Work?
- Prerequisite: Copy the word lists “positive-words.txt” and “negative-words.txt” to the application directory.
- In step #5, specify a keyword/phrase (e.g. data mining, machine learning, etc.) and the 50 most recent tweets generated that match the visitor-specified keywords are retrieved.
- Using a pre-supplied list of sentiment words, the application assesses the tweets and identifies which ones are positive, which ones are neutral, and which ones are negative.
- In the end, a bar graph is generated to display the resulting data.
The Analysis Results
The HTML formatted report can be found here.