AI Search: Context-Aware Recency Boosting

We’re excited to introduce a new AI-powered enhancement to our search capabilities: Context-Aware Recency Boosting.

Until now, our AI search engine applied a general "recency boost," prioritizing newer articles over older ones, even though our AI Index contains valuable content spanning over 10 years. This worked well for many queries, ensuring patrons had access to up-to-date information. However, we recognize that not all searches benefit from a one-size-fits-all approach.

With this new update, our AI dynamically adjusts the recency boost based on the context of the search query, optimizing the balance between relevance and publication date. Here’s how it works:

: For queries that focus on foundational knowledge or historical perspectives—like "Keynes’ theory of consumption" or "history of carbon capture technology"—the AI turns off the recency boost entirely. This ensures that older, highly relevant articles are included in the results.

: For most queries, especially in fields like medicine, technology, or general research, we maintain our default strategy. This applies a moderate recency boost, favoring newer articles while still including older ones if they are highly relevant.

: For queries explicitly seeking the latest advancements—like "latest innovations in carbon capture technology"—the AI applies a stronger recency boost, prioritizing articles from the past 1-2 years.

This novel approach makes the search algorithm smarter and more in tune with what your patrons are looking for.