CouncilDataProject / cdp-backend

Data storage utilities and processing pipelines used by CDP instances.
https://councildataproject.org/cdp-backend
Mozilla Public License 2.0
22 stars 26 forks source link

Allow meeting minutes to be processed like a transcript #218

Open conantp opened 1 year ago

conantp commented 1 year ago

Feature Description

Some public meetings aren't recorded via audio OR video. However, most bodies are required by law to produce minutes or other documentation that provides insight into the matters discussed. What if the cdp-backend was able to apply the same processing applied to transcripts to "meeting minutes" in PDF / Doc format.

Use Case

For CDP instances that wish to import a large number of historical events, or for CDP instances serving smaller communities, there may be public meetings that are not recorded. By indexing the minutes of those meetings, users can search and identify meetings where specific matters or legislation was discussed.

Solution

An enhancement to cdp-backend that accepts an EventIngestionModel that includes a minutes_uri but NO video_uri, and processes the event properly. Users are able to search and find sections in the minutes where a given search term was discussed.

Alternatives

It might not make sense to process minutes documents the same way that we process timestamped transcripts. So perhaps having the ability to index / search minutes could be useful even for meetings that are recorded.