Learning for Unlimited Human Language
Human language is notoriously complex due to the multitude of ways people can express the same meaning. My research focuses on machine learning methods to understand the seemingly unlimited number of expressions in human language. In this talk, I will present our recent work addressing two aspects of this problem:
#1 - “Neural Network Models for Paraphrase Identification, Semantic Textual Similarity, Natural Language Inference, and Question Answering” (published at COLING 2018 - Best Paper Award) We analyze several neural network designs and their variations for sentence pair modeling and compare their performance extensively across eight datasets. We provide a systematic study and show that (i) encoding contextual information by LSTM and inter-sentence interactions are critical, (ii) Tree-LSTM does not help as much as previously claimed but surprisingly improves performance on Twitter datasets, (iii) the Enhanced Sequential Inference Model is the best so far for larger datasets, while the Pairwise Word Interaction Model achieves the best performance when less data is available.
#2 - “A Word-Complexity Lexicon and A Neural Readability Ranking Model for Lexical Simplification” (published at EMNLP 2018) Current lexical simplification approaches rely heavily on heuristics and corpus level features that do not always align with human judgment. We create a human-rated word complexity lexicon of 15,000 English words and propose a novel neural readability ranking model with a Gaussian-based feature vectorization layer that utilizes these human ratings to measure the complexity of any given word or phrase. Our model performs better than the state-of-the-art systems for different lexical simplification tasks and evaluation datasets. Additionally, we produce SimplePPDB++, a lexical resource of over 10 million simplifying paraphrase rules, by applying our model to the Paraphrase Database.
<strong font-size:14px;background-color:#ffffff;"="" style="color: rgb(102, 102, 102); font-family: "Microsoft YaHei"; font-size: 14px;">Biography:
Wei Xu is an assistant professor of Computer Science and Engineering at the Ohio State University. Her research lies at the intersections of machine learning, natural language processing, and social media. She received her PhD in Computer Science from New York University where she was a MacCracken fellow. Between 2014 and 2016, she was a postdoctoral researcher at the University of Pennsylvania. She recently received the NSF CRII Award, Criteo Faculty Research Award, CrowdFlower AI for Everyone Award, Best Paper Award at COLING 2018, as well as research funds from DARPA. She is serving as an area chair for NAACL 2019, EMNLP 2016 and 2018; the publicity chair for EMNLP 2019, NAACL 2016 and 2018; and a workshop co-chair for ACL 2017.