🏆 FVNWL Team Takes Two Prizes at the European Statistics Awards! 🏆
Hong-Hanh, Quang-Tien, and the team FVNWL, have achieved outstanding results in the European Statistics Awards for the Web Intelligence Classification Challenge! They proudly secured:
- 2nd Place in the Accuracy Award (€5,000)
- 3rd Place in the Innovativity Award (€1,000)
The Challenge: Making Sense of Job Ads
The core challenge was to automatically classify millions of online job advertisements into standardized categories. This is crucial for analyzing labor market trends but is complicated by inconsistent job titles, multiple languages, and rapidly evolving job roles that make manual classification impossible.
Our Solution: Generation-Assisted Retrieval
We created a novel “Generation-Assisted Retrieval” framework, treating classification as a smart search problem. Our process was:
- Hybrid Search: We first found potential job categories based on the ad’s overall meaning (semantic search). We then re-ranked these results using a precise word-by-word match on the job title, which we hypothesized was the most critical piece of information.
- LLM-Powered Analysis: The top results were fed to two Large Language Models (LLMs). They didn’t just pick a code—they also explained their reasoning and provided a confidence score, making the system highly interpretable.
- Smart Ensembling: We combined the LLMs’ predictions using a custom Digit-wise Hierarchical Ensembling method, which intelligently voted on each digit of the final job code to maximize accuracy.
If you are interested in our solution, explore our report and Github: