Meeting Minutes for 06/07/2022

Discovery Committee Meeting
June 7, 2022
Minutes

Demo of new features, bug fixes, and documentation for the upcoming release (Minute 12:38)

• Release Notes

Discussion of development priorities for the next release (Minute 48:00)

Road Map 

• Every time we have a release the following week is for development priorities for the next two or three quarters.
• Third-quarter Pascal is going to be working on some reorganization for the full reindex process. 
• Search documentation will be created.
• Pika copyright will be updated and submitted to the US Copyright office.
• Chris will be working on some Google Analytics upgrades.
• Adam, Ashley, and C.J. are working on accessibility standards and planning for Pika.  

Other discussion topics 

• Maintenance Mode will be used for tonight’s deployment (Minute 00:23)
  • Some sites will have account handling setup work and be put in a maintenance mode
  • While the work is being done regular users will see a status page to let them know the catalog is in maintenance mode. This will prevent people from logging in during the account configuration work.
  • This is only for Discovery Partners that use pins.
• Potential changes for advanced search logic (Minute 01:02)
  • Pascal is sure the advanced search logic does not work in a way that people think it would work. 
  • Pascal would like to do something new from scratch instead of trying to fix something that is not designed correctly. 
  • The original search form was based on VuFind. With all searching modifications made over the years have made the advanced search unuseful. 
  • Pascal called for ideas to help create useful advanced searches.
    • Gayle commented that her impression was that it worked better at one time and asked if we could go back and change for better functionality or are we beyond that? 
    • Pascal commented that he thinks the advanced searching has not had serious attention.
      •  Basic searching is not generally basic and it is actually quite complicated searching.
        • It is doing a weighted search of many different types of fields based on relative importance. 
    • For the advanced search, you would probably want to build your own more complex search that is not based on the basic search query that is currently used. 
  • There was a chat request to be able to do some more "and" and "or" with the facets
    • Pascal responded that it is a feature request that he has not come up with a way to build a good interface to have both options.
  • Adam suggested that members could look at their database or search interfaces to see what types of functions are in those advanced searches as we try to think through functionality.
  • Pascal thinks we should reverse that concept and consider the current basic search as the advanced search and the future basic search should be the one that people can control.  
  •  Pascal mentioned that one of the obstacles for search is grouped works since all the searching is against the grouped work. Currently, all the searches do not give you the book with certain properties. The search is giving you the grouped work that has related records that have certain properties.   
• Scoped format boosting updates for search result relevancy (Minute 7:16)
  • Another problem with relevancy searching is that Pascal has a path to fixing the issue that he was not able to implement a change yet. 
  • There is an issue with grouped works, especially for consortiums with catalogs for multiple libraries.
    • One element with grouped works is boosting relevancy by recognizing that a grouped work has many editions of a format that indicates that a work is more relevant to a search.  Every format has a format boost. The problem is when many libraries own many editions of a title the format boost is so outsized to everything else that it will show up in searches where it may not be relevant.
    • The solution that Pascal has in mind is the format boost is needed but not as a universal number for your search scope, especially in the case of limited search scopes like the academics. Most of the records for those many editions in a work are not contributing to the scope so it is really not justified that those records are contributing to the format boost for academic libraries. 
    • The format-boosting scoping will help with a lot of search problems across the board.
  • Nina asked Pascal to click on show editions and describe why the published date is not in any chronological order. Nina always wondered why it is not in chronological order and wondered if it is not helpful to be in that order?
    • Pascal explained that it is a design feature. For a fiction work, it is sorted by availability to the patron.
    • For non-fiction titles, it is in order by the edition number and edition statements. When someone clicks on the place hold button the system puts a hold on the most recent edition.
• Subject Searches (Minute 43:50)
  • When you do a subject search with a single-word phrase, you have many irrelevant results.
  • The way the query works is that it is doing a stemming search. Stemming is when most of the fields look for the common version of a word. For example, stories are stemmed to stori. Various forms of a verb will be reduced. If you have a search word that is both a noun and an adjective the stemming brings up even more irrelevant results.
  • The format boost can also affect the results and include more irrelevant results as mentioned earlier.

The next meeting is Tuesday, September 06, 2022