Closed jordanpadams closed 4 months ago
@jjacob7734 need to describe more what we expect from this theme, and create sub-tickets or one analysis sub-ticket to start with.
1) We need to extract the relevant user stories to define the test suite. 2) Then for these tests, compare the expected results with the actual results from Sinequa by customizing the way each fields are search for, and some weight (boost) on each of them.
User Stories: https://airtable.com/applgnSo7ROCVIjbd/shrvc3CFrOgqsfPlL/tblapwOzIaDVB1x3d
My summary/taxonomy of user stories: https://docs.google.com/spreadsheets/d/1qk-setU_rJ5zLv-jlrPv7ci5Wdxp2UYwt_RPgXaiS8U/edit#gid=0 (also see the Search Function
column in the AirTable above)
Status: @jjacob7734 looking over story categorization made early on in task. to discuss at breakout
Notes from @jjacob7734 + @jordanpadams breakout discussion on 10/5/23:
The following are example queries from @jordanpadams where SDE/Sinequa does not produce the best results. In discussion with SDE, some of these can be improved by fixing data curation problems.
Data pages are not on top:
Suboptimal duplicate pages are referenced:
Planetary Image Galleries
is a subset / copy of the data that is here: https://photojournal.jpl.nasa.gov/Status: Investigating relevance boosting and how that can be brought into the test suite and success criteria.
Create 2 sub-taks:
📆 October Status: Test suite software in work. On schedule
📆 November status: Test suite software in work. On schedule
📆 December status: In work. Completion delayed 1 sprint. No impact on delivery.
Call this done. Also have sinequa documentation here: https://github.com/NASA-PDS/planetary-data-engine/wiki/SDE%E2%80%90Sinequa
💡 Description