Project Specification
- Project Type: Academic
- Project Mode: Independent
- Project Status: Complete
- Project Duration: 3 Months
- Project Involvement
- Supported multiple corpuses
- Authored hyperparameter tuning
- Incorporated dataset partitioning
- Implemented multilayer perceptron
- Integrated Long Short-term Memory
- Incorporated Recurrent Neural Network
- Engineered chatbot response accuracy
- Integrated spell-checking and correction
- Developed apostrophised concatenation
- Software Applications
- Google Colaboratory
- Supported Platforms
- macOS
- Linux
- Windows
- Download: Google Colab
- Documentation: .pdf .docx
Synopsis
A chabot extended from the implementation provided by Mathew Inkawhich
which presents to be a use-case of recurrent sequence-to-sequence neural network models, for conversational artificial intelligence platforms.
The chatbot features three corpuses that compile as it's vocabulary selection, a series of tuned neural network algorithms alongside recurrent neural network,
a spell-check and correction algorithm, multiple measures to validate the accuracy of the configured model, training and testing dataset partitioning,
apostrophised word concatenation, and visual output for the models configurative and performative statuses.