Talpa Search

Talpa Search, an AI book search engine, revolutionizes the way users find books and media by integrating cutting-edge AI technology with reliable book data. This powerful tool utilizes massive language models to analyze and comprehend queries, cross-referencing against a vast repository of authoritative bibliographic data. Offering diverse search criteria, such as subject, tags, characters, and cover design, Talpa Search prioritizes library collection items while including those outside the collection.

As a complementary service currently available at no cost to libraries utilizing Syndetics Unbound from ProQuest and LibraryThing, Talpa Search remains an experimental venture, with uncertainties regarding its future as a standalone product or Syndetics Unbound add-on.

Developed through a collaboration between LibraryThing, a renowned personal cataloging site, and ProQuest, Talpa Search demonstrates the potential of AI in the book search realm.

 It has some limitations, such as working best finding older titles and not being very reliable in finding recently published ones. Talpa Search is a project of LibraryThing, and its name means mole in Latin. In summary, Talpa Search is an innovative and useful AI-powered library search tool that provides accurate and reliable bibliographic data for books and other media.

How does Talpa Search?

Talpa combines so-called “AI” and authoritative book data. The AI comes from Claude AI from Anthropic and ChatGPT from OpenAI. Our authoritative book data comes from Bowker, Syndetics Unbound from ProQuest, and LibraryThing.

Talpa uses AI technology in two ways:

• First, Talpa queries large language models from Anthropic and OpenAI for books and other media, validating every item against true and authoritative bibliographic data.
• Second, Talpa uses the natural-language abilities of large language models to parse and understand queries, which are then answered with traditional library data. So, for example, a search for “novels about World War II in France” taps into subjects, tags, and character data from LibraryThing.

Unlike chatting with large language models directly, Talpa doesn’t “hallucinate” books that don’t exist—everything is checked against our massive storehouse of true and authoritative bibliographic data.

Surprisingly, perhaps, Talpa’s ability to find books by their cover design isn’t powered by AI at all, but by the effort of thousands of book lovers who have played LibraryThing’s CoverGuess cover-tagging game since 2010!

Talpa Search works on your library holdings. Items in your library’s collection are given precedence in the results. Items not in your collection may still appear lower down in the results.

Can we get Talpa Search?

For a limited time, Talpa is free to all libraries that use Syndetics Unbound from ProQuest and LibraryThing. Read the full announcement on Syndetics Unbound News and Updates. Configure Talpa and get links and buttons for your site on the Syndetics Unbound Admin.

Talpa Search is still an experiment. We don’t know if it is going to become a new product, an add-on, or a standard part of Syndetics Unbound.

Interested in learning more? Drop us a line using the form here and note in the comments that you’re interested in Talpa.

What are you working on?

Talpa is in active development, and we want your feedback! Here are some known limitations.

• Talpa works best finding older titles but can miss on recently published ones.
• Because Talpa is intended for the real libraries that use Syndetics Unbound, results are partially determined by what the library has in its holdings. Items that may not be at the library are lower in the results and have black-and-white covers.

Who made Talpa Search?

Talpa Search is a project of LibraryThing. We make the personal cataloging site LibraryThing.com and the small-institution catalog system TinyCat. Together with ProQuest, we are the co-creators of Syndetics Unbound.

The software merges extensive language models from Anthropic and OpenAI with authoritative book information from Bowker, Syndetics Unbound, and LibraryThing to verify each item against reliable bibliographic data.

Version 1: Talpa utilizes artificial intelligence to analyze and comprehend queries, responding with information sourced from traditional library data. Unlike its counterparts, Talpa ensures accuracy by cross-referencing against a comprehensive database of authentic bibliographic data.

Talpa Search operates by scanning library collections and giving priority to relevant items in search results. Additionally, users can customize the tool and receive links and buttons for their website through the Syndetics Unbound Admin at no cost, for a limited period.

Talpa Search remains in the experimental stage and seeks feedback to enhance its functionality. While it excels in locating older titles, it may not be as reliable in finding recently published works.

Talpa Search is a LibraryThing initiative, with its name derived from the Latin word for mole. In essence, Talpa Search is an innovative AI-powered library search tool that offers precise and dependable bibliographic data for books and other media.