DATE OF PUBLICATION
Ilya Makarov Innokentiy Humonen
Co-occurrence Networks for Word Sense Induction
Word sense induction (WSI) is the unsupervised and knowledge-free task of clustering occurrences of homonymous or ambiguous words by their meanings. This problem has been relevant since the second half of the 20th century  and arises in various natural language processing areas, such as machine translation, sentiment analysis, chatbots, etc. In this work, we applied graph neural networks to the WSI-problem using a co-occurrence network and evaluated it on the RUSSE’2018 task . Proposed approach demonstrates satisfactory results with low consumption of computational resources.
You can ask us a question or suggest a joint project in the field of AI
For scientific cooperation and
For journalists and media