講題： Computational Linguistics: The Road Toward Natural Language Understanding and Generation
Computational linguistics is the study of computer systems for understanding and generating natural language. The ultimate goal is to construct machines or programs that could perform natural language processing (NLP), understanding (NLU) and ultimately generation (NLG). In this talk, we will start from the introduction of machine learning to show how it is applied to learn word/sentence representation to solve basic NLP tasks (such as sentiment analysis, named entity recognition, and correference resolution) as well as to be used in advanced NLU tasks (such as machine translation, question answering, and reading comprehension). We then illustrate how these techniques could be applied for opinion extraction from online reviews and event search on the Web. Finally, we will discuss the limitations of natural language understanding when data are trained purely from linguistic forms and the challenges of natural language generation.