Opinion Mining
Project Title: Opinion Mining: Movie Review Analysis
Project Duration: October 2015 - December 2015 (3 months)
Project Type: Academic Project
​
Technology Used: Java, Weka (Tool for visualization)
Find Implementation of Algorithm Here (GitHub Code link): [code]
​
Colleagues:
-
Rohit Maurya
-
Arshpreetsingh Pabla
​
Abstract:
To analyse the movie reviews and classify the movie as positive or negative according to the reviews given to it on scale of 5.
Opinion Mining Steps:
-
Collection of data
-
Pre-process the data
-
Train the model
-
Testing the unknown data
-
Analysis of the movie
​
Basic Opinion Rules to be followed:
-
Neg -> Negative
-
Pos -> Positive
-
Negation Neg -> Positive
-
Negation Pos -> Negative
-
Desired value range -> Positive
-
Below or above the desired value range -> Negative
-
Decreased Neg -> Positive
-
Decreased Pos -> Negative
-
Increased Neg -> Negative
-
Increased Pos -> Positive
Removal of Stop words:
Stop words usually refer to the most common words in a language, there is no single universal list of stop words used by all processing of natural language tools, and indeed, not all tools even use such a list. Some tools specifically avoid removing these stop words to support phrase search.
​
Classifier: We have used J48 as our classifier to train the data. It uses C4.5 algorithm for building decision trees.
​
Filter: We have used StringToWordVector().
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
Applications:
-
Social Media Analysis
-
Good Database
-
Detecting and understanding how the audience is reaching out to a brand, whether positively or negatively.
-
-
TweetFeel – Twitter Search with Feelings
-
TweetFeel gathers real-time Twitter data about whatever search term the user has entered, and then evaluates those tweets for positive and negative feelings.
-
-
Learn faster from Customer Feedback