UPDATE：立委愚人节北京讲演时间地点已经确认，感谢中文信息学会孙教授的邀请和安排，也感谢董振东前辈教授的建议和推举：

The loacation is :

虽然在四月一日路过北平，但不是愚人节玩笑 :=), 具体地点和活动细节待确认后随时update

Sentiment Mining from Chinese Social Media in Big Data Age

by Wei Li, Ph.D. Computational Linguistics

In this information age of big data, social media such as WeiBo (Micro-Blog, or Chinese twitter) is more and more influential.  The popularity of mobile devices such as smart phones makes it possible for anyone to share his/her observation, experiences, opinions and sentiments any time anywhere in the social network such as WeiXin (or WeChat). The social media big data from WeiBo, WeiXin, Customer Review sites, Blogs and Forums are like a gold mine of intelligence, yet to be mined.  They are in the form of natural language (Chinese in this case) and contain intelligence of public opinions and consumer sentiments on any topics, brands and products. Automated sentiment mining via Natural Language Processing (NLP) is a must-do if we (or businesses) do not want to be overwhelmed by the information overload.

Dr. Li's talk will present the design philosophy behind such a sentiment mining system which he has designed and led the team to develop.  He will first discuss the value and scope of NLP in sentiment extraction and mining, pros and cons between the rule based system and learning based classification, and different levels of sentiment mining in response to the various information needs.  He will then demonstrate a list of real life Chinese social media hot topics as mined by the system to show the value and future of big data and NLP, in areas like automatic survey and social media listening and monitoring for consumer insights.

大数据时代中文社会媒体的舆情挖掘

李维 博士

随着大数据时代的到来，社会媒体（譬如微博）的影响力日益增强。智能手机等移动设备的普及，使得普罗百姓的见闻、意见和情绪可以随时随地传达（譬如利用微信）。微博、微信、博客、论坛这些社会媒体大数据好像一座座富含情报的金山，等待我们去挖掘。在大数据面前，如果不想被信息爆炸淹没，就必然需要使用自动手段，尤其是可以用来自动抽取挖掘舆情的自然语言技术。

李博士的报告基于他主持开发的客户舆情自动抽取挖掘系统。报告分两大部分。第一部分阐述自然语言技术在舆情抽取中的应用范围，比较统计分类方法与规则系统方法的利弊，以及舆情分析的层级体系。第二部分通过一系列社会媒体热点话题的实例，展示大数据挖掘的价值和前景。

【置顶：立委科学网博客NLP博文一览（定期更新版）】