Human computation games for commonsense data verification

Tao Hsuan Chang*, Cheng Wei Chan, Jane Yung Jen Hsu

*Corresponding author for this work

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

3 Scopus citations

Abstract

Games With A Purpose (or GWAP) provide an effective way to collect data from web users. As the data grow over time, it becomes increasingly important to verify the data collected. This research explores the design of two games, Top10 and Pirate & Ghost, for verification of sentences in our Common Sense Knowledge Base (CSKB) collected via the Virtual Pets game. Top10 is a single-player game, in which the player attempts to guess the top answers to a given question. The frequency of common answers is used as the verification. Pirate & Ghost is a multi-player role-playing game, in which players cooperate to navigate to a target. The common routes are used to verify relations among concepts in the CSKB. This paper presents the experiments to evaluate the performance of each game, and shows how the games can be coupled to achiever higher efficiency and precision.

Original languageEnglish
Title of host publicationCommonsense Knowledge - Papers from the AAAI Fall Symposium, Technical Report
PublisherAI Access Foundation
Pages19-20
Number of pages2
ISBN (Print)9781577354840
StatePublished - 2010
Externally publishedYes
Event2010 AAAI Fall Symposium - Arlington, United States
Duration: 11 11 201013 11 2010

Publication series

NameAAAI Fall Symposium - Technical Report
VolumeFS-10-02

Conference

Conference2010 AAAI Fall Symposium
Country/TerritoryUnited States
CityArlington
Period11/11/1013/11/10

Fingerprint

Dive into the research topics of 'Human computation games for commonsense data verification'. Together they form a unique fingerprint.

Cite this