Rotom: A Meta-Learned Data Augmentation Framework for Entity Matching, Data Cleaning, Text Classification, and Beyond

Deep Learning revolutionizes almost all fields of computer science including databases and data management. However, the high demand for high-quality labeled training data is slowing down deep neural nets wider adoption. To this end, data augmentation (DA), which generates more labeled examples from existing ones, becomes a common technique. Meanwhile, the risk of creating noisy/unnatural examples and the large space of hyper-parameters make DA less attractive in practice. We introduce Rotom, a multipurpose data augmentation framework for a range of data management and mining tasks including entity matching, data cleaning, and text classification. Rotom features InvDA, a new DA operator that generates natural yet diverse augmented examples by formulating DA as a seq2seq task. The key technical novelty of Rotom is a meta-learning framework that automatically learns a policy model for selecting and combining examples generated by different DA operators, whereby combinatorially reduces the search space of hyper-parameters. Our experimental results show that Rotom can effectively improve a models performance by combining multiple DA operators, even when applying them individually does not yield performance improvement. With this strength, Rotom outperforms a previous deep entity matching system with <9% training data, achieves new state-of-the-art results in low-resource data cleaning, and improves against two recently proposed data augmentation techniques from NLP.