Systems Thinking¶
Primary Goals¶
- Save time
- Gain context
- Deepen understanding
The initial goals are to create a tool that allows viewing individual messages and metadata in a way that allows the reader to view the conversation in different ways. Culling out the noise (cutting some messages) and focusing on the signal of the conversation.
Collection of Messages (aka Topic or Thread)¶
It would be nice to have a generic way to refer to the collection that doesn't evoke preconceived notions based on the messaging platform.
Possibilities: - collection - set - topic - thread - conversation - discussion - series - chain - sequence
For now, I'll use "conversation" as a placeholder.
Metadata¶
Conversation¶
- title
- id
- first_message_author
- first_message_timestamp
- estimated_read_time
- last_post_timestamp
- calculated_conversation_length
- number_of_messages
- number_of_unique_authors
Individual message¶
- id
- author_name
- author_id
- timestamp
- message_length
- messaage_text
- response_to_message_id
- contains_link
Author¶
- id
- name
- username
- additional_privileges (bool)
Shape of data¶
This allows visualization of the data.
- Timeline of messages
- Tree of messages
- Cluster messages by author
- Cluster message by keyword
Calculated Info and Metrics¶
Initial set of calculated metrics
Conversation Metrics¶
- messages_per_day
- median_timestamp
- 75_timestamp
- 90_timestamp
- 25_timestamp
Blocks¶
- Ingest messages
- Display conversation metadata
- Display authors
- Display frequent authors
- Display message
- Display message metadata
- Display shape of data
Ingest messages¶
Plug-in support for different messaging platforms
- Discourse
- Discord
- Slack