Description
This thesis is about creating a simple user friendly Frequently Asked Questions website using natural language processing algorithms. Shashin Patel has explained these algorithms in thesis "Natural language techniques for web language driven interface". The application has the facility to upload a text file. The text file is processed and the text is broken into text chunks using Text tagging and chunking algorithm which is mentioned in the thesis. When the user enters a question, the keywords from the question are collected. The frequency of the keywords is calculated and that keyword frequency helps in determining if the question is in Question Answer table. If the question exists, the application generates the corresponding answer. If the application is not able to find the question in the table, it uses the text chunk table to find the most similar texts using most likelihood estimation algorithm and answer it. Apart from using algorithms referred in the thesis "Natural Language processing algorithms", this application also focuses on why python-scripting language has been used. Also this documentation helps in understanding how Django web framework helps in developing a user friendly, efficient and simple web application. The application is deployed using Google App Engine that supports python scripting language. Django web framework has a library that supports Google App Engine that makes it easier to integrate this tool in the application. Thus, the application can be hosted free and with desired domain name
