Classifing a statement into different categories using DL & NLP.
-
Using different labelled data of various sentences and statements making a model to predict the meaning or intent of the statement
-
The predicted classification of intent can be used a ground intent be various AI-Chatbots and applications for checking the general intent of the user and transfer user to the revelant protocol.
-
I used avaliable data sets of :
1. Converation(intent detection chat-bot data) for greeting and conversation statements 2. Movie Review data set for sattements of review 3. Real and Fake news Data set for News statements(facts) 4. Emotion Data set for satatements defining Emotions
-
I created novel data set combining these dfferent data sets, which was then used to produce a DL (ANN) model.
We have been able to classify differen sentences with an accuracy of 98.88% into our Categories of Conversation, Emotion, News or Review.
We were able to identify the words and the biases of the length for various sentences.
- Word-Cloud giving us an understanding of the most frequently used words in general statement of a kind.
- Lengths of sentences have a crucial part to play in sattement analysis.
This novel model and dataset can be used for further work in training chatbots and other intent classifing applications to get an understanding of the user intent make better automated desicions