CCKS2016: CHINA CONFERENCE ON KNOWLEDGE GRAPH AND SEMANTIC COMPUTING (全国知识图谱与语义计算大会)
PROGRAM FOR THURSDAY, SEPTEMBER 22ND
Days:
previous day
all days

View: session overviewtalk overview

09:00-09:45 Session 13: 特邀报告(Keynote)
Chair:
Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)
09:00
What Computers Should Know

ABSTRACT. Machines with comprehensive knowledge of the world’s entities and their relationships has been a long-standing vision and challenge of AI. In the last decade, huge knowledge bases (aka. knowledge graphs) have been automatically constructed from web data and text sources, and have become a key asset for search, analytics, recommendations and data integration. This digital knowledge can be harnessed to semantically interpret textual phrases in news, social media and web tables, contributing to natural language processing and data analytics. This talk reviews these advances, discusses recent directions such as acquiring commonsense, and identifies new opportunities and future challenges.

09:45-10:30 Session 14: 特邀报告(Keynote)
Chair:
Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)
09:45
面向基础教育的大数据类人智能答题系统总体设想及其困难与挑战
SPEAKER: Heyan Huang

ABSTRACT. 大数据时代,如何充分利用海量知识资源及其处理能力,提高国家信息化建设的智能水平具有极其重要的战略意义;同时伴随网络信息的飞速发展,互联网正从信息网络演变为知识网络,因此,海量知识资源获取与智能知识问答技术将促使智能信息服务水平有质的飞跃,已成为网络大数据时代研究的热点和难点。类人答题作为智能问答的一种有效验证手段,越来越引起学术界、工业界的高度关注。国家重点研发计划立项支持的“基于大数据的类人智能关键技术与系统”项目,通过研究海量知识获取与深度学习、内容理解与推理、问题分析与求解、交互式问答等类人智能的关键技术,构建大数据环境下面向基础教育的海量知识资源和知识图谱,研制具有类人知识处理能力的智能答题系统,其中涉及自然语言处理、信息检索、机器学习、知识工程等多项人工智能核心技术。本报告将主要介绍该系统的总体设想与方案、研究进展,尤其是其中所面临的难点与挑战,以及未来的研究重点与方向展望。

10:30-11:00茶歇(Coffee Break)
11:00-12:00 Session 15: 论坛(Panel)- 知识图谱与认知智能

TBA

Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)
12:00-13:30午餐(Lunch Break)5号楼大堂二层赏园餐厅
13:30-17:00 Session 16: 工业界论坛(Industrial Forum)- Keynotes
Chairs:
Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)
13:30
Analytic knowledge graph for healthcare

ABSTRACT. Since sequencing of the human genome in 2003, we have dreamed about treating patients more effectively based on their genomic profiles. Such a dream remains elusive. “The fundamental difficulty lies in the complexity of biological systems that have evolved through billions of years.” On the other hand, major progress can be and has been made in “personalized medicine” by applying classic AI machine learning on the massive patient medical data accumulated. In essence, we can uncover new insight from the data to help patients without knowing the why a priori. Such new insight is then added back into the medical knowledge to form a richer knowledge graph for analysis later. Exploiting patient medical data brings another set of management problems, namely the heterogeneous nature of data sources and taxonomies, the enormous size of data volume, and huge analytic processing requirements. At this talk, we will discuss all these issues and show some examples at a major research hospital in New York City.

14:00
拓尔思水晶球——基于动态本体的知识管理工具

ABSTRACT. 随着信息爆炸的大数据时代的到来,信息来源五花八门,各行业领域都需要专业的分析师通过数据分析来解决问题和揭示数据背后的秘密,这也是大数据分析师的工作。在互联网上Yago、Dbpedia、Freebase、百度百科等也建立起了各种面向知识关联的应用和服务,拓尔思水晶球通过对实体概念、实体属性、实体与实体之间关系,建立起基于动态本体的知识管理体系,本体定义基于对象的数据模型, 支持动态的本体定义,数据从多源的数据格式,被转换映射为统一的数据对象,关联现实世界中的人、地点、事物、事件以及之间的关系。结合NLP技术,不仅可以从结构化数据中获取知识,还能从非结构化数据中发现和挖掘知识。本次将从实战的角度,分享通过拓尔思水晶球获取知识、建立知识图谱、挖掘知识内涵的全过程。

14:30
从语义到语用

ABSTRACT. 在人与人的交谈中,要理解一句话的含义,除了理解字面含义之外,还要结合多种语境信息,才能理解用户的话语的真实意图。这些语境信息包括说话者的物理语境,如说话的时间,地点和场所,对话的语言语境,即上下文信息,以及知识语境,即说话者的背景知识,领域知识,用户信息等。利用语境信息来理解话语的含义在语言学上称为语用。本报告除了介绍语用学之外,还进一步提出了语用计算,即把语用学应用到人与机器的对话交互中,包括口语的理解,自然语言的生成,和人机交互框架。

15:00-15:20茶歇(Coffee Break)
15:20-17:00 Session 17: 工业界论坛(Industrial Forum)- Invited Talks
Chairs:
Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)
15:20
Knowledge Graph in Japan: Open data beyond

ABSTRACT. This talk introduces recent activities for building Knowledge Graph in Japan. From Open Data Charter of G8 in 2013, many Japanese organizations have published their own data with open data license. For instance, National Tax Agency (NTA) has published a dataset of corporation numbers covering 4.4 million companies, Kawasaki City published valuable information to support families with small children. However almost all of these data are described with CSV or XML format and there are no links to other datasets. In this talk, I introduce many activities in Japan to build Knowledge Graph from these open data. These activities come from two aspects; one is social events such as LOD Challenge, another is mash-up technology related to Linked Data technologies.

15:40
关联挖掘——图可视化的应用实践

ABSTRACT. TBA

16:00
知识图谱在自动应答系统上的应用和挑战

ABSTRACT. TBA

16:20
发现数据之美——大规模行业知识图谱的构建和应用

ABSTRACT. 在互联网高速发展的大背景下,数据的堆积越来越严重。在大多数企业应用场景中,用户希望一种能够存储、处理和查询海量数据的数据库系统;同时,用户还需要能够对多源异构的数据进行有效地融合组织,使其更好地被分析利用来创造价值。本报告将针对企业在数据整合与应用方面遇到的痛点,结合对海翼知大规模行业图谱的介绍,分享整合通用数据,行业数据以及企业私有数据的一些经验,共同探索更好的大数据消费模式。

16:40
小i机器人在中文语义开放平台的研究与进展

ABSTRACT. 介绍人工智能热点研究领域以及小i机器人在人工智能领域的布局;介绍小i机器人在自然语言处理领域的研究进展——小i语义云平台,旨在为用户提供全面的自然语言处理能力,涉及分词、命名实体识别、新词发现、摘要、主题发现、聚类、分类、情感分析等。

17:00-17:30 Session 18: 闭幕式(Closing Ceremony)
Location: 1号楼三层银杏大厅(Building No. 1, Gingko Hall)