Dr. Xia Hu
Associate Professor at Rice University
ChatGPT in Action: An Experimental Investigation of Its Effectiveness in NLP Tasks
Dr. Yixin Cao
Assistant Professor at Singapore Management University
Trustworthy Natural Language Processing with Knowledge Guidance
Pre-trained language models (PLMs) have achieved great success in many downstream tasks, such as Question Answering and Machine Translation. However, even if the network models are getting deeper and deeper, they may still make ridiculous mistakes due to the lack of commonsense or factual knowledge. Using knowledge graphs (KGs) is one of the most effective ways for machines to organize and explore various types of knowledge. In this talk, we will discuss the limitation of current entity-centric KGs and propose our proposed multi-modal event KG, which has a stronger expressive ability. We will describe how we build a pipeline to extract multi-modal events from unstructured data and organize them under the guidance of cross-modal ontology. We will also show the preliminary exploration of trustworthy knowledge reasoning and natural language generation. Finally, we will talk about the remaining challenges to shed light on future research.