SocialBERT -- Transformers for Online SocialNetwork Language Modelling
This work addresses the need for more context-aware language models in online social networks, though it is incremental as it builds on existing BERT architecture.
The authors tackled the problem of incorporating social network information into language models by developing SocialBERT, which integrates knowledge about an author's network position into BERT, resulting in up to a 7.5% improvement in probabilistic modeling quality for specific authors.
The ubiquity of the contemporary language understanding tasks gives relevance to the development of generalized, yet highly efficient models that utilize all knowledge, provided by the data source. In this work, we present SocialBERT - the first model that uses knowledge about the author's position in the network during text analysis. We investigate possible models for learning social network information and successfully inject it into the baseline BERT model. The evaluation shows that embedding this information maintains a good generalization, with an increase in the quality of the probabilistic model for the given author up to 7.5%. The proposed model has been trained on the majority of groups for the chosen social network, and still able to work with previously unknown groups. The obtained model, as well as the code of our experiments, is available for download and use in applied tasks.