Makhortykh, Mykola; Urman, Aleksandra; Ulloa, Roberto; Kulshrestha, Juhi (24 May 2022). Can an algorithm remember the Holocaust? Comparative algorithmic audit of Holocaust-related information on search engines (Unpublished). In: Connected Histories 2022: Memories and Narratives of the Holocaust in Digital Space: 1. EHRI-AT-Konferenz. Wien. 23.05.-24.05.2022.
Full text not available from this repository. (Request a copy)Web search engines play an important role in today's media environment. By responding to the large number of queries (approximately 3.8 million per minute for Google alone), these algorithmic systems shape social reality by informing their users about current and historical phenomena. However, there are many questions concerning the quality of information provided by the search engines in light of existing evidence of them being subjected to bias (e.g., gender or race bias) as well as other forms of malperformance (e.g., retrieval of inaccurate information).
In our paper, we look at how search algorithms represent information about the Holocaust. The importance of this specific subject is two-fold: on the one hand, algorithms can help human societies across the globe to fulfill moral obligations to remember victims of mass atrocities by prioritizing reliable information sources and promoting historical facts. However, algorithms can also undermine Holocaust remembrance by promoting denialist narratives and offensive content, in particular considering the intense use of online media by Holocaust deniers.
To understand what information about the Holocaust is prioritized by the search algorithms and whether their performance is subjected to malperformance, we conduct a systematic audit of the six world’s largest search engines (Google, Bing, Yahoo, Baidu, Yandex, and DuckDuckGo). Using an automated agent-based auditing approach to control for search personalization and randomization, we collected top 20 text search results for multiple (n=160) agents using the query “Holocaust” in February 2020 and March 2021. Then, we used content analysis to identify the source type for each result as well as the quality of information provided by it. By doing so, we aim to understand how different algorithms represent the Holocaust and whether there is consistent malperformance (e.g., recurring prioritization of denialist content) which can affect the way the Holocaust is remembered.
Item Type: |
Conference or Workshop Item (Paper) |
|---|---|
Division/Institute: |
03 Faculty of Business, Economics and Social Sciences > Social Sciences > Institute of Communication and Media Studies (ICMB) |
UniBE Contributor: |
Makhortykh, Mykola, Urman, Aleksandra |
Subjects: |
000 Computer science, knowledge & systems 300 Social sciences, sociology & anthropology 900 History |
Language: |
English |
Submitter: |
Mykola Makhortykh |
Date Deposited: |
23 Jun 2022 14:35 |
Last Modified: |
05 Dec 2022 16:20 |
Uncontrolled Keywords: |
holocaust, search engine, algorithms, algorithm audit, bias, memory, web search |
URI: |
https://boris.unibe.ch/id/eprint/170655 |
Actions (login required)
 |
Edit item |