Skip to main navigation Skip to search Skip to main content

EEG: Knowledge base for event evolutionary principles and patterns

  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The evolution and development of events has its underlying principles, leading to events happened sequentially. Therefore, the discovery of such evolutionary patterns between events are of great value for event prediction, decision-making and scenario design of dialog system. In this paper, we propose Event Evolutionary Graph (EEG), which reveals evolutionary patterns and development logics between events. Specifically, we propose to construct EEG by recognizing the sequential relation between events and the direction of each sequential relation. For sequential relation and direction recognition, we explore the effectiveness of 4 categories of features: count-based, ratio-based, context-based and association-based features for correctly identifying sequential relations and corresponding directions. Experimental results show that (1) the framework we proposed is promising for EEG construction and (2) methods we proposed are effective for both sequential relation and direction recognition.

Original languageEnglish
Title of host publicationSocial Media Processing - 6th National Conference, SMP 2017, Proceedings
EditorsHuan Liu, Xing Xie, Xueqi Cheng, Huawei Shen, Weiying Ma, Shizheng Feng
PublisherSpringer Verlag
Pages40-52
Number of pages13
ISBN (Print)9789811068041
DOIs
StatePublished - 2017
Event6th National Conference on Social Media Processing, SMP 2017 - Beijing, China
Duration: 14 Sep 201717 Sep 2017

Publication series

NameCommunications in Computer and Information Science
Volume774
ISSN (Print)1865-0929

Conference

Conference6th National Conference on Social Media Processing, SMP 2017
Country/TerritoryChina
CityBeijing
Period14/09/1717/09/17

Keywords

  • Event evolutionary graph
  • Knowledge base
  • Sequential relation between events
  • Social media

Fingerprint

Dive into the research topics of 'EEG: Knowledge base for event evolutionary principles and patterns'. Together they form a unique fingerprint.

Cite this