AIDA significantly (1) speeds up, (2) reduces cost and (3) improves accuracy of first level document review.
Filter out 80-90% of non-responsive documents while retaining responsive ones (90-99% recall),
reducing your collection size and vendor hosting and review fees.
Generate responsiveness as well as privilege predictions.
Achieve high recall and precision levels from just 25% of the documents required by industry leaders in TAR systems.
Flexibility
AIDA is unique in that it is highly customizable and adaptable to individual cases.
Unlike traditional TAR systems that often require mass re-annotation of the documents when the document request changes, AIDA can adapt to many types of changes rapidly without additional review time and cost.
Defensibility
AIDA provides not only accurate but highly defensible predictions. The statistical analysis AIDA provides adheres to the commonly accepted standard in the TAR industry.
When AIDA collects the seed examples from the case team, it automatically reserves a subset for evaluation and continuously monitors progress via a metrics dashboard accessible to the team.
Security
AIDA can be hosted in house or within a secure virtual private cloud. All data is encrypted
with AES-256 at rest and in transit. We perform automated network security testing
regularly.
The Founders
Igor Labutov
PhD, Cornell, Carnegie Mellon
Igor received his PhD in Computer Engineering from Cornell University, and was a postdoc at the Machine Learning Department at Carnegie Mellon University,
where he built conversational agents that learn from natural language interactions with users.
Bishan Yang
PhD, Cornell, Carnegie Mellon
Bishan received her PhD in Computer Science from Cornell University, and was a postdoc at the Machine Learning Department at Carnegie Mellon University, where she built machine learning systems that make sense of text documents.
Relevant Publications
"Learning to ask for conversational machine learning"
Shashank Srivastava, Igor Labutov, Tom Mitchell.
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and
the 9th International Joint Conference on Natural Language Processing, EMNLP 2019.
View Publication
"Look-up and Adapt: A One-shot Semantic Parser"
Zhichu Lu, Forough Arabshahi, Igor Labutov, Tom Mitchell.
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and
the 9th International Joint Conference on Natural Language Processing, EMNLP 2019.
View Publication
"Learning Semantic Parsers from Natural Language Supervision"
Igor Labutov, Bishan Yang, and Tom Mitchell.
Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018.
View Publication
"Multi-Relational Question Answering from Narratives: Machine Reading and Reasoning in Simulated Worlds"
Igor Labutov, Bishan Yang, Anusha Prakash, and Amos Azaria.
Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018.
View Publication
"Zero-shot Learning of Classifiers from Natural Language Quantification"
Shashank Srivastava, Igor Labutov, and Tom Mitchell.
Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018.
View Publication
"Leveraging knowledge bases in LSTMs for improving machine reading."
Bishan Yang and Tom Mitchell.
Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL 2017.
View Publication
"A Joint Sequential and Relational Model for Frame-Semantic Parsing."
Bishan Yang and Tom Mitchell.
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017.
View Publication
"End-to-End Learning for Structured Prediction Energy Networks."
David Belanger, Bishan Yang, and Andrew McCallum.
Proceedings of the 34th International Conference on Machine Learning, ICML 2017.
View Publication
"Embedding entities and relations for learning and inference in knowledge bases."
Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 3rd International Conference on Learning Representations, ICLR 2015.
View Publication