Chatbot-with-deep-learning

Implementation

• Read the description of the movie from the dataset.
• Convert the CSV into JSON.
• Collect intents require to train the model.
• Separate the pattern and response based on the data collected.
• Use tokenization that will grab the words from the sentence.
• create a bag of words that will represent an any given pattern(inputs).
• The neural network only understand numeric value rather than a word that's we need to convert them into numeric encoding.
• we create 1 hot encoding which will contain the 1 or 0 based on the word exist or not in the sentence.
• activation=softmax tells the probability of each neuron in the list (helps to finds the response).
• 8 fully connected hidden layer.
• Train the model using a changing number of epoch and batch sizes.
• Input the text and output will be a response based on the prediction and also the genre of the movie.

Architecture

INPUT DATA          :arrow_right:              HIDDEN LAYER         :arrow_right:            HIDDEN LAYER         :arrow_right:         OUTPUT DATA
45 input neurons     :arrow_right:     8 fully connected neurons     :arrow_right:     8 fully connected neurons     :arrow_right:     6 neurons ("Softmax").

Requirements

Install python package required is present Here

Output

Author(s)

Hritik Jaiswal