product-search-scoring

Vietnamese-aware product search uses accent-insensitive text normalization combined with keyword scoring across multiple fields—name, brand, type, and tags. This Skill provides a 3-layer pipeline: text normalization to handle Vietnamese diacritics, keyword processing for relevance ranking, and query building with sorting, enabling fast, accurate results even with regional language variations.

Multi-field search ranking combines scoring across denormalized product attributes with caching to reduce query overhead and improve response time. This Skill implements keyword-scored ranking with pagination support, category and price filters, and tag-based matching, delivering optimized search performance across large product datasets.

Yes. This Skill handles Vietnamese text normalization for accent-insensitive matching while filtering stop words to improve relevance. It processes queries across product name, brand, type, and tags simultaneously, ensuring accurate results regardless of diacritical marks or common word variations.

Scoring-based ranking assigns relevance weights to keyword matches across multiple product fields, then applies category, price, and tag filters to narrow results. This Skill combines normalized keyword matching with structured filtering and pagination, balancing precision and performance for e-commerce search.

Keyword caching reduces redundant normalization and scoring calculations on repeated queries, while highlighting improves user experience by marking matched terms in results. Use both when search volume is high or when user-facing search results need visual feedback on match locations.