COMP474_Project

Build an intelligent agent that can answer university course and student related questions, using a knowledge graph and natural language processing

describe each file

• RDFs: contains all the schemas
• Triples folder: contains generated and written Triples
• data folder: contains courseMaterial and students_grades.csv
• courseMaterial folder: contains COMP335 and COMP474 folders
• COMP335 folders: contains Lectures, Worksheet, Assignments
• COMP474 folders: contains Lectures, Worksheet
• Roboprof.py: main file calls functions from AutomatedKnowledgeBaseConstruction.py, KnowledgeBaseTopicPopulation.py, etc creating the triples in .ttl format.
• AutomatedKnowledgeBaseConstruction.py: contains all the functions necessary to run this project such as building the Knowledge Base and selecting the right query to create outut files.
• Namespace.py contain all namespaces used for Triples creation
• KnowledgeBaseTopicPopulation.py contains functions for extrating topics using SpaCy
• Queries folder holds each query as required in the project in separate files.
• Data holds course materials, Concordia's datasets and students data.

steps to execute code for building knowledge base

• Run the "Roboprof.py" file

steps to setup Fuseki Jena

• Run the exe file for the server.
• Upload dataset from project to the server. Dataset can befound in folder Triples/MergedTriples.ttl

steps to execute RASA chatbot

• In terminal, enter "rasa train" to train the model.
• In another terminal enter "rasa run actions"
• Enter "rasa shell" to start communicating with the chatbot.