Quality over Quantity: Boosting Data Efficiency Through Ensembled Multimodal Data Curation
This addresses the challenge of optimizing model performance through better data curation for machine learning practitioners, though it appears incremental as it builds on existing weak supervision and ensemble techniques.
The paper tackled the problem of inefficient and biased curation of web-crawl datasets by introducing EcoDatum, a learning-driven ensemble method that improved data curation quality, achieving a 28% performance gain over the baseline and ranking first on the DataComp leaderboard with an average score of 0.182 across 38 datasets.
In an era overwhelmed by vast amounts of data, the effective curation of web-crawl datasets is essential for optimizing model performance. This paper tackles the challenges associated with the unstructured and heterogeneous nature of such datasets. Traditional heuristic curation methods often inadequately capture complex features, resulting in biases and the exclusion of relevant data. We introduce an advanced, learning-driven approach, Ensemble Curation Of DAta ThroUgh Multimodal Operators (EcoDatum), incorporating a novel quality-guided deduplication method to ensure balanced feature distributions. EcoDatum strategically integrates various unimodal and multimodal data curation operators within a weak supervision ensemble framework, utilizing automated optimization to score each data point effectively. EcoDatum, which significantly improves the data curation quality and efficiency, outperforms existing state-of-the-art (SOTA) techniques, ranked 1st on the DataComp leaderboard, with an average performance score of 0.182 across 38 diverse evaluation datasets. This represents a 28% improvement over the DataComp baseline method, demonstrating its effectiveness in improving dataset curation and model training efficiency.