Amir Veyseh, a student of Professors Dejing Dou and Thien Nguyen, hit the ground running this year with first-author publications in two top-ranking AI conferences: the International Joint Conference on Artificial Intelligence (IJCAI) and the Annual Meeting of the Association for Computational Linguistics (ACL). Constituting somewhat of record for CIS graduate students in their first year of study, these recognitions bode well for Amir’s future in research and for the department in general.
Amir’s work focuses on relationships between artificial intelligence and natural language. His IJCAI paper develops a novel deep learning method to predict the relationships between entities mentioned in a natural language sentence. This method also extracts the syntactic structure of the sentence, leveraging a new method to control information flow in the neural network and yielding state-of-the-art performance on typical data sets.
The second paper Amir will be presenting this summer seeks to improve AI’s ability to understand whether events mentioned in natural language actually occur. This paper introduces a new deep learning method to improve AI’s understanding of the factuality of natural language sentences. Using graph convolutional neural networks, this method also achieves excellent performance on benchmark datasets.
While we can’t link to Amir’s work till later this summer, we can take a moment now to offer Amir our hearty congratulations!