Computational Linguistics and Lunch at UPenn by our postdoc Qian Yang!

Identifying and Connecting Salient Information via Semantic Text Representations

Abstract: On the web, the staggering amount of data can often appear overwhelming to us and lead to information overload. To solve this problem, one approach is to create human-readable summaries of longer texts and obtain compressed versions that still provide useful information to the user. Another direction is to extract machine-readable semantic representations of a text, which enable downstream information management applications and systems, especially better search interfaces to find relevant information. In this talk, I will introduce our three contributions towards these goals. The first contribution is a novel evaluation metric for textual summaries based on whether they capture salient information. Specifically, we propose PEAK, the first method to automatically assess summary content using the Pyramid method that also generates the pyramid content models. The second contribution is a novel method to present salient information in a text to a reader. Specifically, we rely on a deep learning-based sentence ranking method based on semantic text representations. The third and final contribution is a novel method to connect semantic triple representations so as to make them more easily searchable and to enable better information management applications. 

The upcoming CLunch (Computational Linguistics and Lunch) at UPenn will take place on Monday, March 26th, at 1.30.

Bio: Qian Yang is a postdoc at Penn State University, hosted by Prof. Rebecca J. Passonneau. Before that, she obtained her Ph.D. from Tsinghua University in 2017, advised by Prof. Gerard de Melo. She was also a visiting student at RWTH Aachen University in 2016, hosted by Prof. Hermann Ney. Her research focuses on a wide range of topics in natural language processing (especially text summarization, open information extraction, knowledge graph, machine translation), deep learning, and the intersection of AI and HCI. She is currently seeking a postdoc position starting from May 1st, 2018.