‘Vertical Reality Glasses’: Fixing Typos in Search Queries
summary
Describes building a typo-correction system for search queries, focusing on handling frequent user misspellings and service-specific vocabularies. Covers why generic spellcheckers fail for domain-specific lexicons and approaches to adapt correction algorithms to a search service’s own terms and usage patterns.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 3:19Empty search results problem
- 8:22Typo correction model
- 17:46Error model details
- 23:20Model training
- 29:46Language model
- 34:04Results and architecture
- 37:17Future plans
- 40:12Q&A
related talks
09mSpell Correction & Query Segmentation
41mGetting Started with Query Understanding #bbuzz
31mAdvanced search for your e-commerce website
30mSearch on Steroids — A Story of Semantics
24marchive – Andrei Lopatenko
14mA Case on Building a Multilingual Business Matching System With OpenSearchService
21mBring AI-Based Search to Your Web App
30mSemantic Search and Neural Information Retrieval