Prof. Dr. Felix Naumann


Demo Paper Accepted at NAACL 2021

Julian Risch, Philipp Hager, Ralf Krestel

We are excited to announce that our demo paper titled "Multifaceted Domain-Specific Document Embeddings" has been accepted at the Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2021). The paper is based on the master's thesis by Philipp Hager. A preprint can be accessed here and the code is available on GitHub. Feel free to try out the demo here. A screencast is on YouTube.



Current document embeddings require large training corpora but fail to learn high-quality representations when confronted with a small number of domain-specific documents and rare terms. Further, they transform each document into a single embedding vector, making it hard to capture different notions of document similarity or explain why two documents are considered similar. In this work, we propose our Faceted Domain Encoder, a novel approach to learn multifaceted embeddings for domain-specific documents. It is based on a Siamese neural network architecture and leverages knowledge graphs to further enhance the embeddings even if only a few training samples are available. The model identifies different types of domain knowledge and encodes them into separate dimensions of the embedding, thereby enabling multiple ways of finding and comparing related documents in the vector space. We evaluate our approach on two benchmark datasets and find that it achieves the same embedding quality as state-of-the-art models while requiring only a tiny fraction of their training data.