Coarse-to-Fine: Hierarchical Multi-task Learning for Natural Language Understanding
This work addresses the challenge of efficiently leveraging diverse corpora for natural language understanding, offering a method to improve task performance by modeling task relevance, though it is incremental in building on existing multi-task learning approaches.
The paper tackles the problem of learning generalized text representations for multiple natural language understanding tasks by proposing a hierarchical multi-task learning framework with a coarse-to-fine paradigm, which groups similar tasks to boost performance and reduce negative impacts from irrelevant ones, achieving superior results on 13 benchmark datasets across five tasks.
Generalized text representations are the foundation of many natural language understanding tasks. To fully utilize the different corpus, it is inevitable that models need to understand the relevance among them. However, many methods ignore the relevance and adopt a single-channel model (a coarse paradigm) directly for all tasks, which lacks enough rationality and interpretation. In addition, some existing works learn downstream tasks by stitches skill block(a fine paradigm), which might cause irrationalresults due to its redundancy and noise. Inthis work, we first analyze the task correlation through three different perspectives, i.e., data property, manual design, and model-based relevance, based on which the similar tasks are grouped together. Then, we propose a hierarchical framework with a coarse-to-fine paradigm, with the bottom level shared to all the tasks, the mid-level divided to different groups, and the top-level assigned to each of the tasks. This allows our model to learn basic language properties from all tasks, boost performance on relevant tasks, and reduce the negative impact from irrelevant tasks. Our experiments on 13 benchmark datasets across five natural language understanding tasks demonstrate the superiority of our method.