Identifying Insufficient Data Coverage for Ordinal Continuous-Valued Attributes

Asudeh, Abolfazl; Shahbazi, Nima; Jin, Zhongjun; Jagadish, H. V.

doi:10.1145/3448016.3457315

Cited by 22 publications

(14 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Coverage. The notion of data coverage has been studied across different settings in [4,5,7,8,37,46,52,70] as a metric to measure representation bias . At a high level, coverage is referred to having enough similar entries for each object in a data set.…”

Section: Datamentioning

confidence: 99%

“…It may be inappropriate to treat a 35-yo as young but a 36-yo as old. In [8], the authors extend the notion of coverage to continuous space for identifying representation bias. We will discuss the detailed techniques in Section 4.1.…”

Section: Identification Of Representation Biasmentioning

confidence: 99%

“…Generating Proper Warning Signal. Generating proper signals for the trustworthiness of the analysis [8] occurs when querying about a particular data point that might potentially be concerning due to belonging to an underrepresented subpopulation. Whether to consider the outcome and how to take action is left to the model user.…”

Section: Nomentioning

confidence: 99%

“…Following a similar definition of coverage discussed earlier in section 4.1, they extend the notion of coverage to continuous space for identifying representation bias in [8]. A query point in a continuous data space is covered if there are enough number of data points in its neighbourhood.…”

Section: Continuous Attribute Spacementioning

confidence: 99%

“…At every iteration, for every arm, UCB computes confidence intervals for the expected reward, and selects the arm with the maximum upper-bound of reward to be explored next. Finally, they argue that reward of obtaining a tuple from a group is proportional to how rare this Algorithm 19 Upper Confidence Bound [55] Input: Group counts 𝑞 1 , • [8] generates a trust signal for any arbitrary point in the query space. The warning signal states whether the query point is covered or not.…”

Section: Data Collectionmentioning

confidence: 99%

See 4 more Smart Citations

Representation Bias in Data: A Survey on Identification and Resolution Techniques

Shahbazi¹,

Li²,

Asudeh³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The grand goal of data-driven decision-making is to help humans make decisions, not only easily and at scale but also wisely, accurately, and just. However, data-driven algorithms are only as good as the data they work with, while data sets, especially social data, often miss representing minorities. Representation Bias in data can happen due to various reasons ranging from historical discrimination to selection and sampling biases in the data acquisition and preparation methods. One cannot expect AI-based societal solutions to have equitable outcomes without addressing the representation bias. This paper surveys the existing literature on representation bias in the data. It presents a taxonomy to categorize the studied techniques based on multiple design dimensions and provide a side-by-side comparison of their properties. There is still a long way to fully address representation bias issues in data. The authors hope that this survey motivates researchers to approach these challenges in the future by observing existing work within their respective domains.

show abstract

Section: Datamentioning

confidence: 99%

Section: Identification Of Representation Biasmentioning

confidence: 99%

Section: Nomentioning

confidence: 99%

Section: Continuous Attribute Spacementioning

confidence: 99%

Section: Data Collectionmentioning

confidence: 99%

See 3 more Smart Citations

Representation Bias in Data: A Survey on Identification and Resolution Techniques

Shahbazi¹,

Li²,

Asudeh³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Towards Assessing Data Bias in Clinical Trials

Criscuolo¹,

Dolci²,

Salnitri³

2022

Heterogeneous Data Management, Polystores, and Analytics for Healthcare

View full text Add to dashboard Cite

Fairness & friends in the data science era

2022

View full text Add to dashboard Cite

The data science era is characterized by data-driven automated decision systems (ADS) enabling, through data analytics and machine learning, automated decisions in many contexts, deeply impacting our lives. As such, their downsides and potential risks are becoming more and more evident: technical solutions, alone, are not sufficient and an interdisciplinary approach is needed. Consequently, ADS should evolve into data-informed ADS, which take humans in the loop in all the data processing steps. Data-informed ADS should deal with data responsibly, guaranteeing nondiscrimination with respect to protected groups of individuals. Nondiscrimination can be characterized in terms of different types of properties, like fairness and diversity. While fairness, i.e., absence of bias against minorities, has been widely investigated in machine learning, only more recently this issue has been tackled by considering all the steps of data processing pipelines at the basis of ADS, from data acquisition to analysis. Additionally, fairness is just one point of view of nondiscrimination to be considered for guaranteeing equity: other issues, like diversity, are raising interest from the scientific community due to their relevance in society. This paper aims at critically surveying how nondiscrimination has been investigated in the context of complex data science pipelines at the basis of data-informed ADS, by focusing on the specific data processing tasks for which nondiscrimination solutions have been proposed.

show abstract

Identifying Insufficient Data Coverage for Ordinal Continuous-Valued Attributes

Cited by 22 publications

References 37 publications

Representation Bias in Data: A Survey on Identification and Resolution Techniques

Representation Bias in Data: A Survey on Identification and Resolution Techniques

Towards Assessing Data Bias in Clinical Trials

Fairness & friends in the data science era

Contact Info

Product

Resources

About