About


Straintek is a startup funded by Google’s DNI to filter abusive user content from news portals. Using Deep Leaning and NLP our system is currently the state of the art in the field [1], [2], [3]. Download our dataset or see a realtime application of our system to youtube and reddit.

[1] J. Pavlopoulos, P. Malakasiotis and I. Androutsopoulos, "Deeper Attention to Abusive User Content Moderation". Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2017), Copenhagen, Denmark, 2017. [PDF]

[2] J. Pavlopoulos, P. Malakasiotis, J. Bakagianni and I. Androutsopoulos, "Improved Abusive Comment Moderation with User Embeddings". Proceedings of the workshop Natural Language Processing Meets Journalism, at the Conference on Empirical Methods in Natural Language Processing (EMNLP 2017), Copenhagen, Denmark, 2017. [PDF]

[3] J. Pavlopoulos, P. Malakasiotis and I. Androutsopoulos, "Deep Learning for User Comment Moderation". Proceedings of the 1st Workshop on Abusive Language Online (ALW1), at the 55th Annual Meeting of the Association for Computational Linguistics (ACL 2017), Vancouver, Canada, 2017. [PDF]