Website de recomendação de produtos para restaurantes utilizando banco de dados em grafos
Loading...
item.page.type
TCC
Date
2024-12-05
item.page.ispartof
item.page.citationsscopus
Authors
Matsumoto, Matheus Farias de Oliveira
Rodrigues, Alexandre Henrique Makita
Rodrigues, Alexandre Henrique Makita
publication.page.advisor
Oliveira, Ivan Carlos Alcântara de
Journal Title
Journal ISSN
Volume Title
publication.page.board
publication.page.program
Abstract
A pandemia de COVID-19 impactou o comportamento do consumidor e a dinâmica operacional da indústria alimentícia, que necessitou de adaptações para sobreviver e conseguir vantagem competitiva. Em resposta a esses desafios, este projeto apresenta a construção de um website de recomendação de produtos para restaurantes utilizando um banco de dados em grafo para otimizar a seleção de produtos, promovendo maior eficiência nas compras e vendas. Inicialmente, a coleta de dados foi realizada por meio de técnicas de web scraping, viabilizando testes iniciais antes da criação do website. Além disso, foi conduzida uma análise comparativa entre as complexidades de consultas realizadas em Cypher e em SQL, destacando as vantagens das consultas em grafo para o contexto do projeto. A solução foi implementada com o banco de dados PostgreSQL, complementado pelas extensões Apache AGE, pg_trgm e a biblioteca spaCy. Também empregou-se metodologias de engenharia de software e o wrapper AGESimpleSQL, criado no decorrer do trabalho. O sistema desenvolvido é capaz de realizar consultas em grafo que permite recomendar cestas de produtos de atacadistas para restaurantes, com base na similaridade entre nomes de produtos e ingredientes, quantidades necessárias para a composição de cada prato e preço dos produtos.
The COVID-19 pandemic has impacted consumer behavior and the operational dynamics of the food industry, which has required adaptations to survive and gain competitive advantage. In response to these challenges, this project presents the construction of a product recommendation website for restaurants using a graph database to optimize product selection, promoting greater efficiency in purchases and sales. Initially, data collection was performed through web scraping techniques, enabling initial testing before the creation of the website. In addition, a comparative analysis was conducted between the complexities of queries performed in Cypher and SQL, highlighting the advantages of graph queries for the context of the project. The solution was implemented with the PostgreSQL database, complemented by the Apache AGE extensions, pg_trgm, and the spaCy library. Software engineering methodologies and the AGESimpleSQL wrapper, created during the work, were also used. The developed system is capable of performing graph queries that allow recommending baskets of products from wholesalers to restaurants, based on the similarity between product names and ingredients, quantities required for the composition of each dish and price of the products.
The COVID-19 pandemic has impacted consumer behavior and the operational dynamics of the food industry, which has required adaptations to survive and gain competitive advantage. In response to these challenges, this project presents the construction of a product recommendation website for restaurants using a graph database to optimize product selection, promoting greater efficiency in purchases and sales. Initially, data collection was performed through web scraping techniques, enabling initial testing before the creation of the website. In addition, a comparative analysis was conducted between the complexities of queries performed in Cypher and SQL, highlighting the advantages of graph queries for the context of the project. The solution was implemented with the PostgreSQL database, complemented by the Apache AGE extensions, pg_trgm, and the spaCy library. Software engineering methodologies and the AGESimpleSQL wrapper, created during the work, were also used. The developed system is capable of performing graph queries that allow recommending baskets of products from wholesalers to restaurants, based on the similarity between product names and ingredients, quantities required for the composition of each dish and price of the products.
Description
Indicado para publicação.
Keywords
banco de dados em grafo , postgreSQL , apache AGE , recomendação de produtos , análise de dados , graph database , postgreSQL , apache AGE , product recommendation , data analysis