boradesanket13 / deep-learning-based-chat-bot

Deep-Learning-Based-Chat-Bot

Dependancies

• numpy
• tensorflow
• scikit-learn

pip install numpy tensorflow scikit-learn

Step 1: Training Data

Step 2: Data Preprocessing

• Label Encoding
• Tokenisation
• Padding

Step 3: Building and Training the Chatbot Model

Layers:

1. Embedding Layer: Maps the input sequence of integers to a dense vector representation.

2. Flaten Layer: Convert the multi-dimensional output of the Embedding layer into a one-dimensional vector

3. Dense layer with ReLU activation: Introduces non-linearity to the model, allowing it to learn complex patterns in the data.

4. Output Layer with Softmax Activation Function: Produces probability scores for each class label

Compile the model using

1. Adam optimizer
2. Sparse categorical crossentropy loss function

Training model using fit method with following parameters

1. input sequences
2. encoded labels
3. epochs

Step 4: Generating Responses

predicted label is then converted back into its original text form using the inverse_transform method of the LabelEncoder