Training Data Improvement by Automatic Generation of Semantic Networks for Bias Mitigation
Roman Englert,
Jörg Muschiol
Issue:
Volume 6, Issue 1, March 2022
Pages:
1-7
Received:
20 February 2022
Accepted:
16 March 2022
Published:
29 March 2022
Abstract: The significance of Bias Detection has increased appreciably, due to the increased application of AI. Although syntactic bias is well explored with statistical techniques, there remains semantic bias challenge like for example, Google’s face recognition which excludes colored people. Human expertise is required to detect semantic bias, e.g., for the application of the root-out-bias method. We propose a further automatization to this laborious method, based on the Training Data Improvement for Bias Mitigation (TDIBM). The concept, is to automatically construct a Semantic Network (SN) from the domain description of the training. For the semantic network nouns are extracted. As a second step, synonyms and semantically similar nouns are searched, e.g. in dictionaries, and added to the SNs. As a result, the SN contains nouns that enhances the given domain, with previously unknown knowledge. This SN can be used to check with, e.g., the root-out bias method, whether the training sample is biased, or not. Should the training sample be biased, then the corresponding nouns from the SN can be added to the training sample set to mitigate the bias. The newly developed method, TDIBM is evaluated twofold: Firstly, with the description of the COMPAS system, which is a case management and decision support tool used by U.S. courts to assess the likelihood of a defendant becoming a recidivist. Secondly, an autonomous driving domain is applied, to investigate accidental driving of a Tesla car. Here TDIBM detected among many new features, including one to solve ambiguous scene interpretations for autonomous driving vehicles.
Abstract: The significance of Bias Detection has increased appreciably, due to the increased application of AI. Although syntactic bias is well explored with statistical techniques, there remains semantic bias challenge like for example, Google’s face recognition which excludes colored people. Human expertise is required to detect semantic bias, e.g., for th...
Show More
A Review on Emerging Trends and Technologies in Library
Som Nepali,
Rajesh Tamang
Issue:
Volume 6, Issue 1, March 2022
Pages:
8-15
Received:
4 March 2022
Accepted:
21 March 2022
Published:
31 March 2022
Abstract: Library service through mobile technology is a recent trend in library service. Mobile technology and its development have given rise to the excitement of faculty and student fraternity. This type of infrastructure needed by libraries to provide such services. The awareness about technologies like Mobile based services, Augmented Reality, Gamification, Internet of Things applications. The method opted for the study is descriptive and the tool used for collecting information is web survey. Literature related to emerging trends in libraries was collected from similar projects and other related articles from web. The project was started by analysing the technological developments in various libraries, kinds and new trends that emerged recently. Relevant data regarding the topic is also collected from scholarly publications and online databases to review the benefits, usage and the importance of emerging technology trends in libraries. This topic deals with new developments and techniques that are evolving in libraries. Some of the new trends are identified in this work. The limitation of the study is that, as many innovations are being introduced and also developed, only a selected number of technologies are included in the study.
Abstract: Library service through mobile technology is a recent trend in library service. Mobile technology and its development have given rise to the excitement of faculty and student fraternity. This type of infrastructure needed by libraries to provide such services. The awareness about technologies like Mobile based services, Augmented Reality, Gamificat...
Show More
