A carregar...
Publicação
Retrieving user interactions and personalizing content through recommender systems
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.83 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Designers and product creators can struggle to express their vision in the
early stages of a project, resourcing to abstract ideas.
A Machine Learning platform to aid authors in
getting visuals and gather user preferences is in
demand. The tool presented helps in this process, by
giving users graphic representation of their ideas,
which are entered via keywords. The app is aware of
users’ inputs and learns as it cycles through the
process. Using reputable models from Open AI,
Meta and in-house built recommender systems, a
Streamlit application was created to compile a
package of useful functionalities to creators.
Descrição
Palavras-chave
Machine learning Web application streamlit Recommender systems Text mining Database constrution Image similarity Cnn