Matching engine analytics: Intent distribution of final set of matches vs in the community

[arushirai1]

Which intents are commonly found in the final set of matches? Which intents are commonly found in the community? How does the intent distribution in the final set of matches compare to the intent distribution in the community? To start answering questions about the characteristics of the intent distribution in the final set of matches and the community, we will start by counting the occurrence of each intent (giving and seeking separately) in the provided user or match list.

Definition of Done

• Implement metric function to count each intent occurrence the final set of matches
• Implement metric function to count each intent occurrence (giving and seeking sides separately) in the community
• Implement display function to show the final match set intent count distribution by each intent category (tutoring, comedy, etc)
• Implement display function to show the community intent count distribution by each intent category (tutoring, comedy, etc) and side (giving, seeking)
• Write tests for both metric functions
• Integrate metric functions into pipeline (mentor will help with this)
• Validate that the pipeline runs without errors and produces correct metrics

See code pointers going over an existing metric function in #251 .

[arushirai1]

For the display function, create a function in pipeline/load/display_metrics.py