干旱区生态环境知识资源中心(Arid): Machine learning applied to understand perceptions, habits and preferences of lamb meat consumers in the Brazilian semi-arid region

Arid

DOI	10.1016/j.smallrumres.2023.107088
	Machine learning applied to understand perceptions, habits and preferences of lamb meat consumers in the Brazilian semi-arid region
	Silveira, Robson Mateus Freitas; Lima, Debora Fonteles; Camelo, Beatriz Veloso; Mcmanus, Concepta; da Silva, Valdson Jose; Ferreira, Josiel; Costa, Helio Henrique Araujo; Dias, Carlos Tadeu dos Santos; Contreras-Castillo, Carmen Josefina; Paveloski, Alessandro; Vecchi, Laura Bertolaso De; Sales, Arthur Pereira; Sarries, Gabriel Adrian; Landim, Aline Vieira
通讯作者	Silveira, RMF
来源期刊	SMALL RUMINANT RESEARCH
ISSN	0921-4488
EISSN	1879-0941
出版年	2023
卷号	227
英文摘要	The objective of this study was to characterize the profile of the lamb consumer in the Brazilian semi-arid region using a machine learning approach. A total of 347 questionnaires with questions about socioeconomic aspects, consumption habits and preferences, meat preparation, and appreciation due to product quality were applied. A combination of cluster analysis (hierarchical and k-means), canonical discriminant analysis, neural network, and tree analysis were performed. The machine algorithms revealed three clusters: Traditional profile - Mature rural consumers, married or single, low level of education, low salary level, average consumption of meat, who do not pay for quality and do not know the benefits of lamb products; Emerging profile: Young men and women, single, with a high level of education, intermediate salary level, low consumption of meat, pay for quality, and do know the benefits of lamb meat; and Conventional profile: Mature urban consumers who live in the city, married and single, with a high level of education and salary, low consumption of meat, pay for quality and do not know the benefits of lamb meat The neural network confusion matrix for classifying consumers in the profile determined that 75.5% were classified in their group of origin, which validates the construction of the typology of consumers by the cluster analysis algorithms. Age was the main variable to segregate the nodes of the decision tree (P < 0.001): Node 1- < = 22; Node 2 > 22 < 29 and Node 3 > 29 years The use of a machine learning approach was able to reveal three types of consumers and defined patterns that could serve as strategies for increasing consumption and greater insertion of lamb in the meat market, as well as analyzing the perception of Brazilians in relation to the meat quality, where these findings may have implications for labelling schemes, marketing and sales strategies for cooked lamb products.
英文关键词	Sheep meat Semi-arid region Survey Multivariate analysis
类型	Article
语种	英语
收录类别	SCI-E
WOS记录号	WOS:001165241000001
WOS关键词	SHEEP ; QUALITY
WOS类目	Agriculture, Dairy & Animal Science
WOS研究方向	Agriculture
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/398652
推荐引用方式 GB/T 7714	Silveira, Robson Mateus Freitas,Lima, Debora Fonteles,Camelo, Beatriz Veloso,et al. Machine learning applied to understand perceptions, habits and preferences of lamb meat consumers in the Brazilian semi-arid region[J],2023,227.
APA	Silveira, Robson Mateus Freitas.,Lima, Debora Fonteles.,Camelo, Beatriz Veloso.,Mcmanus, Concepta.,da Silva, Valdson Jose.,...&Landim, Aline Vieira.(2023).Machine learning applied to understand perceptions, habits and preferences of lamb meat consumers in the Brazilian semi-arid region.SMALL RUMINANT RESEARCH,227.
MLA	Silveira, Robson Mateus Freitas,et al."Machine learning applied to understand perceptions, habits and preferences of lamb meat consumers in the Brazilian semi-arid region".SMALL RUMINANT RESEARCH 227(2023).