A model for trust under a suitcase word perspective.

• Ano de defesa: 2017
• Autor(a) principal: Cartolano Junior, Etienne Américo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Behavioral decision; Decisão comportamental; Informação (Confiabilidade); Simulação; Simulation; Suitcase word; Trust; Trust model
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/3/3141/tde-16082017-151433/

Trust is a social phenomenon fundamental to relationships and a building block of our society. People experience it daily, such as in a borrowing between friends, in an e-commerce transaction, in a mother-son relationship, in a connection between autonomous agents, or to show faith in God (\"In God we trust\"). In the specific case of Biodiversity domain, trust is one of the pillars of the Citizen Science projects, which are helping to solve the lack of biodiversity data by engaging citizens to work as volunteers to address this problem. Measuring and simulating levels of trust on these projects might reveal or anticipate losses; for example, the disposal of data because a deficit of trust on the technical capacity of the volunteers, opening an opportunity to manage and improve it. However, trust is a hard concept to define. The word \'trust\' may carry different meanings, such as honesty, security, integrity, competence, etc. and this is an attribute of the \'suitcase words\'. Adopting the \'suitcase\' perspective would change the way as we define, model, and simulate trust, once people would identify, decode, and simulate many meanings of trust with a single approach. In this scenario, the main objective of our research was to verify the hypotheses 1) that trust is a suitcase word, and 2) that trust can be modeled and simulated under a suitcase word perspective. A network analysis of the Web of Science citation database was able to confirm the hypothesis that trust is a suitcase word, since a distribution analysis of articles showed that trust occurs across a wide range of disciplines, and since co-occurrence maps of keywords showed that trust meanings from these disciplines may be significantly different. To verify the second hypothesis, we proposed a framework to manage trust with three components: 1) a suitcase model to identify different meanings of trust, which is the main purpose of this work, 2) a procedure to detail trust situations in terms of the suitcase model, and 3) a behavioral decision model of confidence, which was required for our simulation, since trust and control play complementary roles in the development of confidence, and consequently, to generate a confident behavior to cooperate. In our suitcase model the decision to trust (or distrust) the trustee depends on the trustors\' general capacity to take risks (= trustfulness) and on the assessment of trustee\'s interests and capacity to behavior as the trustor expects (= trustworthiness). In a practical and workable way, trustworthiness was considered a function of the trustor\'s expectations (expected evidence) and the trustee\'s previous behavior (collected evidence) for each situation. We proposed a formalism to the suitcase model, and then replicated the PlayGround simulator to modify it and incorporate our model. The new simulator, the PlayGround 2.0, was used to run a case study using trust situations from Citizen Science projects. Our main goal with this case study was to test the hypothesis that trust can be simulated under a suitcase perspective. A successful simulation would plot agents in the field reacting differently according to each situation. Results were as expected, what demonstrated the comprehensive utility of our model, with potential to handle different meanings of trust in the context of Citizen Science in the Biodiversity domain.