College of Information Studies

Permanent URI for this communityhttp://hdl.handle.net/1903/1631

The collections in this community comprise faculty research works, as well as graduate theses and dissertations.

Browse

Filters

2023 - 20242

Settings

Subject: search.filters.subject.Crowdsourcing

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    REVISITING SHAKESPEARE'S WORLD: OPTIMIZING DATA OUTCOMES AND INVESTIGATING CONTRIBUTOR DYNAMICS
    (2024) Wang, ZhiCheng; Van Hyning, Victoria; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    In this study, we present our work processing data output from Shakespeare's World (2015-2019), an early transcription project hosted on the Zooniverse online crowdsourcing platform. We refined the dataset to make it more amenable to low-code tools such as OpenRefine, enabling easier exploration and reuse. Utilizing the cleaned dataset, we also explored Shakespeare's World volunteers’ contribution patterns. By documenting our process of cleaning the outcome dataset, we provide steps and insights that may be useful for other transcription projects working with data derived from the Zooniverse platform. In addition to offering one plausible way to clean and analyze Zooniverse outcome data, our study also reveals the significant contributions from both anonymous and registered Shakespeare’s World volunteers; the challenges in maintaining participation over the project’s lifespan; and how the original aggregation protocol, which was designed specifically to combine multiple transcriptions by Shakespeare’s World volunteers, resulted in fewer successfully transcribed lines than expected. These findings have broader implications for project design, volunteer engagement, and data management practices in online crowdsourced transcription projects.
  • Thumbnail Image
    Item
    The People's Choice: PAIRing User-Centered Design With Crowdsourcing To Combat Misinformation on TikTok
    (2023) Grover, Saransh; Hassan, Naeemul; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Social Networking sites like Facebook, Twitter, YouTube, and TikTok have created a rampant increase in user-generated content online. Moderation and validation of misinformation on these platforms are still significant challenges. One approach to address misinformation on social media has been to crowdsource the validity of content through the platform users. However, research conducted on crowdsourced fact-checking has focused largely on traditional and text-based sources. In addition, it has yet to focus on user-centered design to understand how users of platforms would create tools to mitigate misinformation. This thesis addresses these knowledge gaps by understanding approaches to using crowdsourcing to combat misinformation on TikTok, the fastest-growing social networking site with over one billion monthly active users. By using TikTok as a case study, I conduct a thematic analysis of content on the platform to understand how users currently counter claims and misinformation and then conduct participatory design sessions with TikTok users to identify limitations, improvements, and potential solutions. Based on these findings, I present a set of design guidelines referred to as the PAIR approach that outline key considerations for a crowdsourcing platform combatting misinformation on a social networking site such as TikTok.