CM2015 Programming with Data Midterm Coursework Assignment

part

organisation

• Function to load intents from JSON files
• Function to load and search using regex patterns
• Function to generate responses.

Correct Interaction between main loop and functions (argument passing and return calls).

Students are expected to:

• Write flexible pattern matching statements to recognise multiple instances of the same intent type.
• Utilise basic regex constructs (e.g., \d, \w, ., *, +, ?) for robust pattern handling.
• Build regex patterns that account for variations in user input (e.g., case insensitivity, optional words).
• (Optionally) Implement advanced regex features like grouping, or lookaheads for more nuanced understanding.

• Choose responses randomly from a list of options.
• Applying dynamic strings substitutions with a memory (e.g., store user’s name or colour or some personal information in an object and substitute it in a response).
• Combination of techniques mentioned above.

• Each intent should contain a list of regexbased patterns stored in the JSON file.
• Responses should be written as template strings in the JSON file. These templates can include placeholders (e.g., {name}, {color}) that are dynamically filled at runtime using string substitution based on user input or stored memory.
• The chatbot should implement a function that loads the JSON file at runtime, extracts patterns and responses, and builds internal structures such as pattern2intent and intent2response dictionaries.

• Splitting user input into individual words or tokens to analyse structure and meaning more easily.
• Eliminating characters like commas, or periods that are not essential for intent detection.
• Reducing words to their root form to match patterns more broadly and accurately.

reflection

• Chatbot application (e.g., use-case, domain of operation).
• Describe 3 different test cases that clearly illustrate chatbot behaviour.
• Any other steps to organise data/code that you implemented.
• Describe advanced techniques that you used.

• Be reproducible in the current notebook format including making relevant data sources and libraries accessible and explicit.
• Use proper conventions e.g. relative path vs absolute.
• Be explained or described where libraries are used in relation to their utility/ability to solve a particular problem in an efficient manner.
• Notebooks should be structured with a logical set of processes/procedures including clear, logical headings.
• Not overly verbose e.g. including comments to describe print statements.