Detalhes bibliográficos
| Ano de defesa: |
2021 |
| Autor(a) principal: |
Seminotti, Malomar Alex
 |
| Orientador(a): |
Rieder, Rafael
|
| Banca de defesa: |
Não Informado pela instituição |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade de Passo Fundo
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Computação Aplicada
|
| Departamento: |
Instituto de Ciências Exatas e Geociências – ICEG
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
http://tede.upf.br:8080/jspui/handle/tede/2320
|
Resumo: |
Recommender systems are used in several domains, such as content suggestion or product purchase, helping in decision-making processes. These systems often analyze content stored in databases, and unstructured datasets such as images and videos are underexplored. With this in mind, this work presents a recommender system with the ability to indicate products, in real-time, based on people’s behavior, considering the analysis of video images using computer vision and artificial intelligence techniques. The inferences for sales suggestions take monitoring images of a pre-defined area, with products that users observe during a visit to the store. We conducted a pilot study with 15 volunteers, in a laboratory environment, due to the restrictions imposed by the Covid-19 pandemic. The application identified all people correctly 100%, without generating false positives while preserving the participants’ identities, not named them any time. The study showed that the application could register customers’ behavior and buying habits to the store owner, offering an efficiency of 73.4% in identifying the products observed. This fact can help retailers giving similar products at more affordable prices, offering advantages to the customer to add sales value, improve the availability of products at the points of sale, in addition to helping to train their employees. |
