|
EXPLAINABLE AI: HOW EXPLAINABILITY IMPACTS THE HUMAN INTERACTION WITH AI |
|---|---|
| รหัสดีโอไอ | |
| Creator | Donghee Shin |
| Title | EXPLAINABLE AI: HOW EXPLAINABILITY IMPACTS THE HUMAN INTERACTION WITH AI |
| Publisher | National Research Council of Thailand |
| Publication Year | 2565 |
| Journal Title | SOCIAL SCIENCE ASIA : Official Journal of National Research Council of Thailand in conjunction with Panyapiwat Institute of Management |
| Journal Vol. | 8 |
| Journal No. | 2 |
| Page no. | 21-Jan |
| Keyword | Algorithmic Privacy, Newsbots, Data Disclosure, Explainable Algorithmic Journalism, Privacy Paradox Explainable AI: How Explainability Impacts the Human Interaction with AI |
| URL Website | http://e-journal.nrct.go.th/ |
| Website title | e-journal |
| ISSN | 2229-2608 |
| Abstract | With the rapid growth in the use and implementation of AI in the journalism industry, concerns on the ethical implications have surfaced recently. The issue of explainable algorithmic journalism has grown rapidly and resulted in a large body of attempts that embody normative qualities such as transparency, privacy, and fairness. This study investigates the effect of explainability on privacy in newsbots to understand how users' information processing leads to data disclosure. We discuss the conceptual mapping of explainability and privacy in algorithms, followed by empirical modeling of how the explanatory heuristics play out in the cognitive processes to which they contribute to user privacy and data disclosure. A mixed-method design incorporating both qualitative and quantitative approaches was used to discover user heuristics and to test the effects of incorporating explanatory cues into newsbots on user privacy. Together, our results reveal that explanatory heuristics exert a key role in users' intentions to disclose more data in algorithm interactions. Explainable newsbots facilitate users' understanding of decision-making in algorithms by evoking transparency and fairness, which serve as the cues for user trust and privacy. To suggest best practices for journalism and identify ongoing challenges for newsbots, we discuss algorithmic information processing and show how the process can be utilized to improve user privacy and trust. |
