Word Sense Disambiguation

Unsupervised, Knowledge-Free, and Interpretable Word Sense Disambiguation

This page contains a description of a word sense disambiguation (WSD) system based on the word sense inventories induced by the JoBimText project. The distinctive feature of this WSD model is that it is at the same time unsupervised, knowledge-free, and interpretable.

Description

Interpretability of a predictive model is a powerful feature that gains the trust of users in the correctness of the predictions. In word sense disambiguation (WSD), knowledge-based systems tend to be much more interpretable than knowledge-free counterparts as they rely on the wealth of manually-encoded elements representing word senses, such as hypernyms, usage examples, and images. We present a WSD system that bridges the gap between these two so far disconnected groups of methods. Namely, our system, providing access to several state-of-the-art WSD models, aims to be interpretable as a knowledge-based system while it remains completely unsupervised and knowledge-free. The presented tool features a Web interface for all-word disambiguation of texts that makes the sense predictions human readable by providing interpretable word sense inventories, sense representations, and disambiguation results. We provide a public API, enabling seamless integration. 

The system performs extraction of word and super sense inventories from a text corpus. The disambiguation models are learned in an unsupervised way for all words in the corpus on the basis on the induced inventories. The user interface of the system provides efficient access to the produced WSD models via a RESTful API or via an interactive Web-based graphical user interface. The system is available online and can be directly used from external applications. The code and the WSD models are open source. Besides, in-house deployments of the system are made easy due to the use of the Docker containers.12 A prominent direction for future work is supporting more languages and establishing cross-lingual sense links.

Resource

• The source code of the system, featuring instructions on in-house deployment with docker.
• The original publication at the EMNLP'2017 conference.
• System architecture and documentation of the API. 

 Screenshots