ÀÔÇоȳ» ºäÆäÀÌÁö
[öÇлç»ó¿¬±¸¼Ò/Çѱ¹Æ÷½ºÆ®ÈÞ¸ÕÇÐȸ] Á¦35ȸ ÄÝ·ÎÅ°¿ò ¾È³»
Ä«Å×°í¸® ±âŸ ÀÛ¼ºÀÚ Ã¶Çаú
ÀÛ¼ºÀÏ 2021-03-18 18:06 Á¶È¸¼ö 584

¾È³çÇϼ¼¿ä öÇаú »ç¹«½ÇÀÔ´Ï´Ù.

³»ÀÏ ¿ÀÀü¿¡ ÀÖÀ» Á¦35ȸ ÄÝ·ÎÅ°¿òÀ» ¾Æ·¡¿Í °°ÀÌ ¾È³»ÇØ µå¸®¿À´Ï °ü½ÉÀÖ´Â ÇлýºÐµéÀÇ ¸¹Àº Âü¿©¸¦ ¹Ù¶ø´Ï´Ù.

 

 1. Çѱ¹Æ÷½ºÆ®ÈÞ¸ÕÇÐȸ Á¦35ȸ ÄÝ·ÎÅ°¿ò

 2. ÀϽÃ: 2021³â 3¿ù 19ÀÏ(±Ý¿äÀÏ) ¿ÀÀü 10½Ã~

 3. ¹ßÇ¥: À̹ΰæ (Åػ罺 ´ëÇб³, ¿À½ºÆ¾)

 4. Á¦¸ñ: "Enabling Participatory and Procedurally-Fair AI"

 5. ³»¿ë ¿ä¾à:

As artificial intelligence (AI) is transforming work and society, it is ever more important to ensure that AI systems are fair and trustworthy and support critical values and priorities in organizations and communities. In this talk, I will first present empirical findings on people¡¯s trust and fairness around algorithms that make managerial and resource allocation decisions. My research suggests that techniques for distributive fairness are not sufficient for gaining people¡¯s trust in AI. Addressing this gap, I propose two frameworks for achieving procedurally-fair and participatory AI: a procedural justice framework that lays out considerations for procedural fairness in algorithmic decisions, and a participatory framework called WeBuildAI that enables people to build algorithms for their own communities. I present a case study of this framework with a nonprofit called 412 Food Rescue in which stakeholders used the framework to build a food donation matching algorithm and adjudicate equity and efficiency trade-offs in the algorithm.  

 6. °øµ¿ÁÖÃÖ : ¼­¿ï´ëÇб³ öÇлç»ó¿¬±¸¼Ò, ÀΰøÁö´ÉÆíÇ⼺ ÃÖÀûÈ­ ¿¬±¸´Ü(°æ³²´ë)

 7. ÈÄ¿ø : Çѱ¹¿¬±¸Àç´Ü

 8. ¹æ¹ý: zoomÀ» ÀÌ¿ëÇÑ ½Ç½Ã°£ ¿Â¶óÀÎ ¹ßÇ¥/Åä·Ð

¡´Zoom ȸÀÇ Âü°¡¡µ (ÁÖ¼Ò º¯°æ)

https://kaist.zoom.us/j/87419229048?pwd=VXM3Wi9Rdk1vS25Sam9vWDVCZHFGZz09

Meeting ID: 874 1922 9048

Passcode: 278059