STEAM: Self-Supervised Taxonomy Expansion with Mini-Paths
This work addresses the low coverage issue in taxonomies, which are critical for knowledge-based applications, by providing an incremental improvement over existing expansion methods.
The paper tackles the problem of expanding existing taxonomies with new concept terms to address low coverage, proposing STEAM, a self-supervised model that uses mini-paths from the taxonomy for node attachment prediction, resulting in 11.6% higher accuracy and 7.0% better mean reciprocal rank than state-of-the-art methods on three benchmarks.
Taxonomies are important knowledge ontologies that underpin numerous applications on a daily basis, but many taxonomies used in practice suffer from the low coverage issue. We study the taxonomy expansion problem, which aims to expand existing taxonomies with new concept terms. We propose a self-supervised taxonomy expansion model named STEAM, which leverages natural supervision in the existing taxonomy for expansion. To generate natural self-supervision signals, STEAM samples mini-paths from the existing taxonomy, and formulates a node attachment prediction task between anchor mini-paths and query terms. To solve the node attachment task, it learns feature representations for query-anchor pairs from multiple views and performs multi-view co-training for prediction. Extensive experiments show that STEAM outperforms state-of-the-art methods for taxonomy expansion by 11.6\% in accuracy and 7.0\% in mean reciprocal rank on three public benchmarks. The implementation of STEAM can be found at \url{https://github.com/yueyu1030/STEAM}.