Summary of Search Focus Group Discussions

The ideas contained on this page came from the four IRC focus group discussions conducted in March 2016 as well as in follow-up emails on the topic.

• Facets for bibliographic data
• Ability for search to handle diacritics properly
• Ability to handle anchored searches (e.g. “^harry potter” to identify indexed fields that begin with “harry potter)
• Autocomplete / autosuggest functionality that predicts search as the user types
• The ability for Evergreen sites to create custom indexes
• The ability to define search formats based on MARC fixed fields
• Strong advanced search filtering – specifically copy location and format filtering
• The ability to search a lot of MARC data in a keyword search - Keyword searching works very well for vast majority of searches due to good coverage of the record
• Ability to search by ISBN in both numeric and keyword search.
• Ability to search by barcode
• Session-less searching, allows linking to search results
• Separate scope and search depth ranging (i.e you can choose to search a particular location and then search at a depth (branch, system, consortium) in the context of that location)
• Call number shelf browse

• Improved speed
  • In general
  • When applying Copy location / format filters
  • When conducting broad searches.
• Better balance between fuzziness and precision when records are retrieved based on word variations
• Improving search retrieval/relevance through analysis of user behavior, handled in a way that respects user privacy
• Improve relevance for data-rich, higher quality records so that similar records with less MARC data don't consistently rank higher than the longer records.
• True blending of relevance factors, e.g. bib-relevance and popularity providing intelligent bumps that result in better relevance ranking.
• Improve relevance for one-word titles
• Autocomplete / autosuggest improvements
  • Improve relevance-ranking for suggested terms
  • Work properly with screen readers
  • Highlight matching terms with diacritics
• Ability to search public copy notes
• More flexibility in the types of facets that can be defined
  • pub date
  • formats
  • audience
  • fiction vs. non-fiction
  • language
• Did You Mean? Suggestions
• Improved speed for ingesting records
• Improvements to zero-hit searches
  • Drops user into a browse search
  • Suggest a meaningful alternative search
  • Automatically launch a popular, similar search
• No core limit to # of search results
• When search hits maximum records that can be retrieved, it should inform the user not all results were retrieved and provide tips for narrowing the search.
• Intermediate results screen in keyword searches that first show author / subject headings that matches terms entered in author/subject search, and then brings the user to results for that heading.
• Leverage authority cross-references in keyword search results – related search
• Ability to manually tweak metarecord groupings
• Including prefixes in call number searching
• More browse search options
  • Call number (on browse search page)'
  • Publisher
  • Date
  • Journal title
• The ability to pull in external, linked data or canned information to the search results page that is related to the user's search terms. (Example: search for a celebrity on Google, info on the celebrity appears on the right.)
• Search also retrieves public lists (bookbags) on the same topic.
• Intuitive way for users to provide hints for type of material are more relevant (e.g. “for a paper” “for leisure reading” etc.)
• Ability to have a non-keyword search that just sorts by format, location, and date (e.g. "give me the newest DVDs for my library")
• Local synonym list / thesaurus
• Catalog tells users which word variations were used in their search and gives them the choice to search the exact terms only.
• Search augmented with external, curated disambiguation information / related, broader, narrower information.
• Improve metarecord grouping / display so that “group formats and editions” search could be the main result set for most searches.
• Ability to quickly jump from record to other formats/editions of same work
• Using OCLC Work Entities to group metarecords
• Ability for end user to toggle on/off "Search within contents and summary notes?".
• True blending of relevance factors, e.g. bib-relevance and popularity providing intelligent bumps that result in better relevance ranking.
• Single search across library materials, articles, digital repository objects, Overdrive, etc.
• Highlighted search terms
• The ability for the UI show why some records are more relevant than others (relevance scores?
• Easier / more understandable to configure search.
• New sort options for Date Created
• Add ability to sort by shelving location in Shelf Browse
• Search by prefix/suffix

• Weighting for particular fields (title, main author)
• Searching by collections (copy location groups?)
• Sometimes need to use apostrophe to find the correct record (look at normalizers)
• More popular titles floating higher (coming in 2.11)

• Currency
• # of copies
• # of circs, holds (or other activity, popularity)
• Specific fields are more important
• Proximity of search terms
• # of times search term appears in record

• Speed
• Better balance between fuzziness / precision
• Leveraging cross-references
• Intermediate search screen for results
• Improved relevance
• Spelling assistance / Did You Mean?
• Better blending of coverage density with other relevance factors (e.g. activity)
• Make it more understandable to see why system is retrieving / ranking as it does (because then it's easier to adjust it.)

• Improvements for saving and sharing lists
• Easier list management, moving titles between lists
• Usability / language improvements for facets on search results page.
• Integrate GoodReads ratings into search results page
• Carousel for new books