Google Ranking Algorithm Research Introduces TW-BERT

TW-BERT, described in a Google white paper, is a unique framework that boosts search rankings with minimal effort.

Highlights

TW-BERT is an end-to-end query term weighting framework that bridges two paradigms to improve search results

Integrates with existing query expansion models and improves performance
Deploying the new framework requires minimal changes

Term Weighting BERT (TWBERT) is a unique ranking framework that Google has announced. It enhances search results and can be easily integrated into preexisting ranking systems.

This new architecture is a breakthrough that enhances ranking procedures generally, including query expansion, though Google has not stated that it is utilising it.
It’s also quite simple to implement, which increases the likelihood that it will be used.

TW-BERT has many co-authors, among them is Marc Najork, a Distinguished Research Scientist at Google DeepMind and a former Senior Director of Research Engineering at Google Research.

He has co-authored many research papers on topics of related to ranking processes, and many other fields.

What is TW-BERT?

TW-BERT is a ranking framework that assigns scores (called weights) to words within a search query in order to more accurately determine what documents are relevant for that search query.

Query Expansion is another area where TW-BERT shines.

By rephrasing or expanding a search query (for as by adding the term “recipe” to the query “chicken soup”), a better match can be made between the search query and the content.

Bringing Together Two Information Retrieval Models with TW-BERT

The report presents a comparison of two search strategies. Both statistical and deep learning models are considered.

The advantages and disadvantages of each technique are then discussed, with the suggestion that TW-BERT can serve as a middle ground that avoids the drawbacks of both.

TW-BERT Bridges Two Approaches

Standard Lexical Retrieval