Responsible AI in the Enterprise by Adnan Masood PhD | Heather Dawe MSc

Author:Adnan Masood, PhD | Heather Dawe, MSc
Language: eng
Format: epub
Publisher: Packt Publishing Limited
Published: 2023-07-30T00:00:00+00:00

Auditing checklists and measures

Along with compliance standards and code reviews, quantifying the results for model bias is a critical step in building accountable ML systems. In this section, we will provide a list of some of these checklists and measures.

Datasheets for datasets

Datasheets for datasets is an initiative aimed at improving transparency, accountability, and an understanding of the datasets used in the development and training of ML models. Introduced by Timnit Gebru, an AI ethicist and former co-leader of Google’s ethical AI team, as well as Kate Crawford and others, this initiative proposes using a standard way to report datasets, which its creators refer to as datasheets33. Their rationale was inspired by the electronics industry, where datasheets provide important information about the components being used:

“In the electronics industry, every component, no matter how simple or complex, is accompanied with a datasheet that describes its operating characteristics, test results, recommended uses, and other information.”

For ML, datasheets for datasets outlines essential information about a dataset, including its purpose, composition, collection process, data preprocessing, intended uses, limitations, and ethical considerations. By providing this information, datasheets help developers, researchers, and users understand the context in which a dataset was created, its potential biases, and the appropriateness of its use for specific tasks.

By promoting transparency and encouraging developers to think critically about the data they use, datasheets for datasets contributes to the responsible and ethical development of AI and ML systems. It can help identify potential issues related to fairness, accuracy, and bias, and encourage the creation of more robust, reliable, and equitable AI models.

An example of such a datasheet in use could be for traffic data used for smart city initiatives, such as the one developed and used by Bristol City Council in the UK. This datasheet provides essential information about the dataset, which contains urban traffic patterns to help develop intelligent transportation systems. This datasheet outlines the dataset’s purpose, composition, collection process, preprocessing steps, intended uses, limitations, and ethical considerations. The dataset includes data from various sources, such as traffic cameras, vehicle counts, and GPS data from smartphones. It has been preprocessed to ensure data quality and protect user privacy. While the dataset is intended for use in developing traffic prediction models and traffic management systems, it comes with certain limitations, such as potential biases and data quality issues.

The datasheet will highlight the importance of ethical considerations, such as data anonymization and adherence to local data protection regulations, ensuring transparency and fostering responsible AI development in the urban transportation domain.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.