公司机器人与信息自动化研究所 天津市智能机器人技术重点实验室

Institute of Robotics and Automatic Information System Tianjin Key Laboratory of Intelligent Robotics

2018年秋季先进机器人与人工智能系列学术讲座（第122期）

Seminar Series：Advanced Robotics & Artificial Intelligence

报告题目：The First Step for AI-based Human-Like Language Understanding - Sentiment Analysis of Text

报告人：Dr. WANG Zhaoxia 

单位：Social and Cognitive Computing Department, Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR), Singapore

时间：2018年11月7日16:00--17:30

地点：公司津南校区13-329

Abstract:

With the advent of the Internet, more and more people prefer to post their feedback, reviews, and opinions through social media platforms, such as Facebook, Twitter, Chinese Weibo, and so on. Sentiment analysis of such data is becoming a fast and effective way of evaluating public opinion and sentiment for addressing practical problems, such as predicting consumer preferences or sensing public emotions towards some products, services or issues. In addition, the current AI systems and robots developed still cannot truly understand human language. Therefore, the motivation of this research is two folds: (i) To Realize Human Like Real-World Language Understanding of Text; and (ii) To Develop Human Like Brand New AI Technologies for Big Data Analysis. As the first step to realize the two goals, in this talk, the presenter will introduce sentiment analysis of text, which includes learning-based, non-learning based and hybrid methods. For learning-based methods, different enhancements, such as feature selection, negation dealing, and emoticon handling are studied to improve the performance of existing machine learning methods; for non-learning based methods or hybrid methods, a series of patentable technologies are proposed to address the limitations of the existing methods.

Biography:

Dr. Zhaoxia Wang is a data and AI scientist at the Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR), Singapore. For more details please refer to the attachment.