Case Study: CareerBuilder improves multilingual search and recommendations with Basis Technology Rosette

CareerBuilder - Customer Case Study

CareerBuilder, a global leader in human capital solutions and one of the largest job boards in the U.S., needed to improve search and recommendation quality across massive English and multilingual resume and job-listing data. After moving from FAST to Apache Solr, the company faced the challenge of matching the right people to the right jobs with high precision and recall, while handling linguistic variations in languages including English, German, Greek, Spanish, and Chinese. Basis Technology’s Rosette text analytics services were part of the language analysis stack they evaluated to solve this problem.

Basis Technology’s Rosette delivered the best results in head-to-head testing, coming closest to CareerBuilder’s FAST baseline and outperforming standard stemming approaches in precision and overall accuracy. Its lemmatization, decompounding, language identification, and part-of-speech tagging helped CareerBuilder improve search relevance and power better content-based recommendations, including stronger recall for compound-word languages without sacrificing precision. The impact was significant: CareerBuilder reported that Rosette became a key part of its search infrastructure, supporting a better user experience for over 24 million unique visitors per month.

CareerBuilder

Trey Grainger

Director of Engineering, Search & Analytics