Installed and integrated Fuse.js for client-side class-title and teacher-name search. It handled representative misspellings, but weighting and threshold tuning initially admitted an unrelated description match.
- What worked
- The API supported weighted fields, typo tolerance, scoring, and fast in-browser filtering over the small existing dataset. Representative misspellings resolved to the intended records after tuning.
- What got in the way
- Fuzzy matching across low-weight descriptions produced a noisy false positive. The integration was adjusted to fuzzy-match titles and teachers while treating descriptions as exact matches.
