Bias in algorithmic filtering and personalization

@article{Bozdag2013BiasIA,
  title={Bias in algorithmic filtering and personalization},
  author={Engin Bozdag},
  journal={Ethics and Information Technology},
  year={2013},
  volume={15},
  pages={209-227}
}
  • E. Bozdag
  • Published 1 September 2013
  • Computer Science
  • Ethics and Information Technology
Online information intermediaries such as Facebook and Google are slowly replacing traditional media channels thereby partly becoming the gatekeepers of our society. To deal with the growing amount of information on the social web and the burden it brings on the average user, these gatekeepers recently started to introduce personalization features, algorithms that filter information per individual. In this paper we show that these online services that filter information are not merely… 

OPIAS: Over-Personalization in Information Access Systems

TLDR
It is demonstrated how each user interaction may affect the selection of subsequent posts, sometimes resulting in the creation of a 'filter bubble.

Modelling opinion dynamics in the age of algorithmic personalisation

TLDR
An opinion dynamics model where individuals are connected through a social network and adopt opinions as function of the view points they are exposed to is devised, finding that algorithmic filtering might influence opinions’ share and distributions, especially in case information is biased towards the current opinion of each user.

Understanding User Beliefs About Algorithmic Curation in the Facebook News Feed

TLDR
This work investigated user understanding of algorithmic curation in Facebook's News Feed, by analyzing open-ended responses to a survey question about whether respondents believe their News Feeds show them every post their Facebook Friends create.

bias goggles: Graph-Based Computation of the Bias of Web Domains Through the Eyes of Users

TLDR
This work proposes the bias goggles model, for computing the bias characteristics of web domains to user-defined concepts based on the structure of the web graph, and exploits well-known propagation models and the newly introduced Biased-PR PageRank algorithm.

Algorithms are not Neutral: Bias in Recommendation Systems

TLDR
It is known that iterative information filtering algorithms in general create a selection bias in the course of learning from user responses to documents that the algorithm recommended, but this systematic bias in a class of algorithms in widespread use largely goes unnoticed.

Algorithms are not neutral: Bias in collaborative filtering

TLDR
In addition to biased data and biased algorithm makers, AI algorithms themselves can be biased, and this source of bias warrants serious attention given the ubiquity of algorithmic decision-making.

The BiasChecker: how biased are social media searches?

TLDR
A tool, called BiasChecker, is introduced that contributes to the check for bias in search results on a social media platform, and takes into account several factors that can interfere with the detection of bias, e.g., the cross-over effect, geolocation, IP address, and language.

Social media personalization algorithms and the emergence of filter bubbles

This article investigates how the technical infrastructure of social media platforms, particularly personalization algorithms, enables filter bubbles to emerge. It identifies several possible

Privacy Concern, Trust, and Desire for Content Personalization

TLDR
Investigating how members of the general public think about personalization concluded that focusing on the benefits rather than the costs of personalization may be a more useful starting point in ongoing debates.

From Editors to Algorithms

TLDR
A content analysis of Facebook’s own patents, press releases, and Securities and Exchange Commission filings is used to identify a core set of algorithmic values that drive story selection on the Facebook News Feed, and finds evidence that friend relationships act as an overall influence on all other story selection values.
...

References

SHOWING 1-10 OF 148 REFERENCES

Social information filtering: algorithms for automating “word of mouth”

TLDR
The implementation of a networked system called Ringo, which makes personalized recommendations for music albums and artists, and four different algorithms for making recommendations by using social information filtering were tested and compared.

User Profiles for Personalized Information Access

TLDR
This chapter surveys some of the most popular techniques for collecting information about users, representing, and building user profiles and discusses in detail user profiles represented as weighted keywords, semantic networks, and weighted concepts.

Challenges in measuring online advertising systems

TLDR
A first principled look at measurement methodologies for ad networks is taken, which proposes new metrics that are robust to the high levels of noise inherent in ad distribution, identifies measurement pitfalls and artifacts, and provides mitigation strategies.

Gatekeepers and other intermediaries

TLDR
Librarians are identifying new roles and developing new competencies for the era of electronic information and, arguably, they are proving their worth in the information transfer process even more strongly than before.

Through the Google Goggles: Sociopolitical Bias in Search Engine Design

Search engines like Google are essential to navigating the Web’s endless supply of news, political information, and citizen discourse. The mechanisms and conditions under which search results are

GroupLens: an open architecture for collaborative filtering of netnews

TLDR
GroupLens is a system for collaborative filtering of netnews, to help people find articles they will like in the huge stream of available articles, and protect their privacy by entering ratings under pseudonyms, without reducing the effectiveness of the score prediction.

The Filter Bubble: What the Internet Is Hiding from You

Author Q&A with Eli Pariser Q: What is a Filter Bubble? A: Were used to thinking of the Internet like an enormous library, with services like Google providing a universal map. But thats no longer

The role of social networks in information diffusion

TLDR
It is shown that, although stronger ties are individually more influential, it is the more abundant weak ties who are responsible for the propagation of novel information, suggesting that weak ties may play a more dominant role in the dissemination of information online than currently believed.

Search Engine Bias and the Demise of Search Engine Utopianism

TLDR
This chapter argues that search engine bias is the beneficial consequence of search engines optimizing content for their users, and the most problematic aspect of search engine biases, the “winner-take-all” effect caused by top placement in search results, will be mooted by emerging personalized search technology.

Infotopia: How Many Minds Produce Knowledge

This book explores the human potential to pool widely dispersed information, and to use that knowledge to improve both our institutions and our lives. Various methods for aggregating information are
...