Library Metasearch
####Create a Unified Interface to Browse All Internal and External Reference Descriptions
##Background
The AMNH library holds an extensive collection of publications and archives, most of which can be found in catalogs available online (Sierra). Additionally, the Museum's publications are available in full text in the library's digital repository (DSpace). The Digital Special Collections (Omeka) holds thousands of images from digitized photographic negatives, archives and rare books. To support archive materials and primary resources, the library has been creating entity records of museum expeditions and expedition personnel, departments, permanent halls, and temporary exhibitions (xEAC). These entity records provide general descriptions offering context and sometimes links to related resources in the museum.
An example of a Google Instant Answer card with myriad related resources
All of these resources combined provide a rich history of the museum itself through its staff, scientific research and expeditions or exhibitions and special events. To find this material efficiently, a user has to know where to search for a particular asset. For instance, a search for digitized images should be done using the Omeka public interface. Primary materials (like field books and other expedition related items) are found in our ArchivesSpace catalog of archives and special collections. The library's online catalog in Sierra contains the majority of the library’s holdings, but descriptions are limited.
Links to external databases can be entered manually in any one of these systems but it would be great to virtually browse all of our descriptions using a single interface. Similar to the example from a Google search pictured above, the library wants to dynamically pull data and generate a visually appealing "instant answer" or info sheet / card that displays relevant images and basic data with links to fuller descriptions. Information can be generated from metadata in the catalogs and from content from digitized publications - maybe it's possible to grab relevant images and text from museum publication PDFs? How should we store search logs, tracking information or analytics about searches? It's up to you, we need you to show us what can be done!
Example search terms to use:
- Hyman, Libbie Henrietta, 1888-1969
- Whitney South Sea Expedition of the American Museum of Natural History (1920-1941)
- Akeley, Carl Ethan, 1864-1926
- Hayden Planetarium
All of these separate systems have their own public interfaces for searching and retrieving information and their own REST APIs. While some data may be duplicated in systems (e.g., finding aid data in ArchivesSpace also has a corresponding bibliographic record in Sierra), a user has to choose one of the systems to navigate to find what they are looking for. Many times, a user will repeat the same search on another system's portal page. It is important to describe the material using the most appropriate standards for the format; however, a user has to understand our particular library landscape in order to search the right tool.
Of course any search tool you build and its results pages should have a user-friendly display (mobile-friendly, responsive) and be designed to be visually appealing, as you'll be organizing a lot of information and we want it to be easy to read and navigate. A user should eventually be able to drill down to the specific data source systems in order to see further descriptions (descriptive metadata, basically) on a particular asset, but we'd like you to first design the tool for finding general, not detailed information. We are reserving the "keyword in context" results for the Snippet Search challenge, though you could collaborate with a team and make it a "module" to the library metasearch (see Solutions below).
This challenge absolutely requires review and collaboration with AMNH API Portal hackathon teams to discuss working with API wrappers around DSpace, Omeka, BHL, and others as documented on the Online Resources And Data Sets page.
-
The basic search interface and results page. We'd like to see what a unified search and results page interface would look like. Show us what you've got!
-
Faceted searching & search by concept. Create an interface with dynamically generated results that allow users to search by concept / category such as: Person, Expedition, Exhibition, Halls, Temp Hall, Departments. This can use the [controlled list of names]((https://docs.google.com/spreadsheets/d/1uk8RjsOo1F9ErQEn_F-4ymHhCVFPgRWEg_65jHbMry8/edit?usp=sharing) and other resources listed in the AMNH API Portal Challenge.
-
"You May Also Be Interested In..." See if you can come up with a way to generate related search queries - maybe this can be built through capturing search histories or maybe by inferring relationships based on content of materials being sourced? Maybe there's a way to refine within a search - can the search strings be saved to develop some learned behavior? People who searched "X" also searched "Y"? We'd love to see what you come up with for this.
-
Incorporate the "snippet" mode of viewing information into your search results. Display snippets of search result information with links to expand and display a larger view of additional content, or otherwise link to external documentation. This would involve collaborating with teams who are working on the Snippet Search challenge.
-
Is there something we missed? What else can you think of? What would make a "metasearch" more useful or powerful?
- Online Resources And Data Sets: Contains all relevant information for this challenge!
- AMNH API Portal Challenge
- Snippet Search Challenge
