Query expansion based on user behaviour

Summary

Grab developed a query expansion framework to resolve search issues such as typos, multi-language terms, missing alternatives, and word spacing. The corpus is generated through manual annotation for the top thousand fat-head queries and session-based data mining of user rewrite patterns. Offline pipelines extract adjacent query rewrites occurring within 30 seconds where only the second query generated a click, limiting candidates to three per query for serving efficiency. In production, expanded keywords pass through a query understanding, multi-recall, and result fusion flow with automated pruning of underperforming pairs. Online A/B testing across six Southeast Asian countries showed improvements in click-through and conversion rates, with data-mined expansions outperforming manual annotations alone.

Takeaways

1. 1

   User rewrite mining heuristics extracted adjacent query pairs within 30-second windows where the initial query had no click and the second query received a click.

2. 2

   Expansion candidates were capped at three per query to maintain serving efficiency across query understanding, multi-recall, and result fusion stages.

3. 3

   A/B testing across six countries demonstrated that data-mined query expansions produced higher click-through and conversion rate uplifts than manual annotations alone.