Sentiment Analysis
About Project
-
Developed a machine learning model for sentimental analysis of human emotions using recurrent neural networks and LSTM(long short-term memory)
-
First part includes testing different models on millions of data to decide best model
-
Improved model by summarizing input text
-
The purpose of this model is to provide accurate feedback of human sentiment
Steps of my project-
Dataset Collection
-
Collected millions of data for training sentiment analysis model by web scrapping and finding datasets online
-
Labelled all the positive sentences as 1 and all the negative sentence as 0
Trying various models
-
Wrote python script to test various possible inbuilt models(sklearn, keras, etc.) and compared the accuracy of all those models
-
In conclusion from few research papers and testing various models, LSTM neural network gave the best result for sentiment analysis
Coding LSTM Algorithm
-
Coded LSTM Algorithm on some movie tweet samples to analyze sentiments
Improved Model
-
Improve accuracy and reducing time complexity
-
Removed stop words and summarized the sentences