A carregar...
Publicação
Influence of knowledge and trust on the perceptions of algorithmic gender bias in artificial intelligence
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 925.04 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Artificial Intelligence (AI) is used more frequently, automating the workplace each day, therefore becoming a part of each person's personal and work life. The AI we know of today needs data to learn from. However, there is a gender data gap, which means there is a gap of women in the data that exists today. Furthermore, when there is bias in society, and this is reflected in the data Artificial Intelligence uses, it can learn to make those same biased-decisions. It is therefore important for people to understand the systems they use can be making decisions that may not be unbiased.
This dissertation therefore aims to understand if knowledge in AI and trust in AI have any influence on how individuals in the tech sector perceive gender bias in AI. Furthermore, to understand if gender has any influence on this as well. To understand this a survey was utilized to gather perceptions on this topic.
The results show that when the knowledge in AI is higher, there is a higher perception of gender bias in AI. Furthermore, when an individual has a lower trust in AI there is also a higher perception of gender bias in AI. Gender also has an impact, as it is shown that men have on average a lower perception of gender bias in AI than women. Furthermore, the effect of trust in AI and knowledge in AI on the perception of algorithmic gender bias is weaker for men than for women.
Descrição
Mestrado Bolonha em Management
Palavras-chave
Contexto Educativo
Citação
Haitsma, Marloes Susan (2024). “Influence of knowledge and trust on the perceptions of algorithmic gender bias in artificial intelligence”. Dissertação de Mestrado. Universidade de Lisboa. Instituto Superior de Economia e Gestão
Editora
Instituto Superior de Economia e Gestão