Collective Intelligence and Crowds

Permanent URI for this collection


Recent Submissions

Now showing 1 - 5 of 6
  • Item
    Democratic Replay: Enhancing TV Election Debates with Interactive Visualisations
    ( 2018-01-03) Plüss, Brian ; De Liddo, Anna
    This paper presents an online platform for enhancing televised election debates with interactive visualisations. Election debates are one of the highlights of election campaigns worldwide. They are also often criticised as appearing scripted, rehearsed, detached from much of the electorate, and at times too complex. Democratic Replay enhances videos of election debates with a collection of interactive tools aimed at providing a replay experience centred around citizens' needs. We present the system requirements, design and implementation, and report on an evaluation based on the ITV Leaders' Debate from the 2015 UK General Election campaign.
  • Item
    It’s All News to Me: The Remix
    ( 2018-01-03) Boon, Miriam
    This paper examines the automation of editorial curation of online news and blog articles based on reader ratings. Websites usually provide no guidelines on how to evaluate and rate articles; the NewsTrust project explores how doing so could improve rating precision. Building on and expanding from existing, but incomplete, research, I describe simulations of article comparison to determine how many reader ratings are necessary to distinguish between articles.
  • Item
    A Multi-Scale Correlative Approach for Crowd-Sourced Multi-Variate Spatiotemporal Data
    ( 2018-01-03) Gorko, Thomas ; Yau, Calvin ; Malik, Abish ; Harris, Matt ; Tee, Jun Xiang ; Maciejewski, Ross ; Qian, Cheryl ; Afzal, Shehzad ; Pijanowski, Bryan ; Ebert, David
    With the increase in community-contributed data availability, citizens and analysts are interested in identifying patterns, trends and correlation within these datasets. Various levels of aggregation are often applied to interpret such large data schemes. Identifying the proper scales of aggregation is a non-trivial task in this exploratory data analysis process. In this paper, we present an integrated visual analytics environment that facilitates the exploration of multivariate categorical spatiotemporal data at multiple spatial scales of aggregation, focusing on citizen-contributed data. We propose a compact visual correlation representation by embedding various statistical measures across different spatial regions to enable users to explore correlations between multiple data categories across different spatial scales. The system provides several scale-sensitive spatial partitioning strategies to examine the sensitivity of correlations at varying spatial extents. To demonstrate the capabilities of our system, we provide several usage scenarios from various domains including citizen-contributed social media (soundscape ecology) data.
  • Item
    Coordinating Advanced Crowd Work: Extending Citizen Science
    ( 2018-01-03) Crowston, Kevin ; Mitchell, Erica ; Østerlund, Carsten
    Crowdsourcing work with high levels of coupling between tasks poses challenges for coordination. This paper presents a study of an online citizen science project that involved volunteers in such tasks: not just analyzing bulk data but also interpreting data and writing a paper for publication. However, extending the reach of citizen science adds tasks with more dependencies, which calls for more elaborate coordination mechanisms but the relationship between the project and volunteers limits how work can be coordinated. Contrariwise, a mismatch between dependencies and available coordination mechanisms can be expected to lead to performance problems. The results of the study offer recommendations for design of crowdsourcing of more complex tasks.
  • Item
    Computer-Mediated Deception: Collective Language-action Cues as Stigmergic Signals for Computational Intelligence
    ( 2018-01-03) Ho, Shuyuan Mary ; Hancock, Jeffrey T.
    Collective intelligence is easily observable in group-based or interpersonal pairwise interaction, and is enabled by environment-mediated stigmergic signals. Based on innate ability, human sensors not only sense and coordinate, but also tend to solve problems through these signals. This paper argues the efficacy of computational intelligence for adopting the collective language-action cues of human intelligence as stigmergic signals to differentiate deception. A study was conducted in synchronous computer-mediated communication environment with a dataset collected from 2014 to 2015. An online game was developed to examine the accuracy of certain language-action cues (signs), deceptive actors (agents) during pairwise interaction (environment). The result of a logistic regression analysis demonstrates the computational efficacy of collective language-action cues in differentiating and sensing deception in spontaneous communication. This study contributes to the computational modeling in adapting human intelligence as a base to attribute computer-mediated deception.