A carregar...
Publicação
Seleção inteligente de recursos humanos com modelos LLM
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.24 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
O processo de triagem de currículos na área de vendas enfrenta desafios significativos devido à diversidade de formatos, terminologias e níveis de detalhamento. Para superar esta dependência da análise manual, este trabalho investigou a aplicação de modelos de linguagem natural (LLMs) para automatizar a extração e padronização de informações relevantes de currículos.
A metodologia utilizou modelos como GPT-4.1, GPT-4.1 Mini e Gemini 2.5 Pro, além da ferramenta ChatPDF e bibliotecas de apoio para a extração textual. Foram elaborados prompts específicos para estruturar atributos nominais e ordinais de forma consistente. O desempenho dos modelos foi avaliado com métricas como acurácia e erro médio aritmético. Posteriormente, o modelo GPT-4.1, que obteve o melhor desempenho, foi aplicado em um conjunto ampliado de 50 currículos para validação. Os dados extraídos foram submetidos a um modelo classificador, resultando em um Erro Médio Absoluto (MAE) de 0.76, numa escala de 10 pontos, na comparação com dados reais, o que valida a confiabilidade do método de extração para a classificação automática.
Os resultados demonstram que os modelos de linguagem natural são eficazes na extração de dados, destacando-se o modelo GPT-4.1. Conclui-se que o uso de LLMs é uma abordagem promissora para a triagem automatizada, pois reduz o esforço manual e aumenta a consistência das avaliações, tendo sido consolidada em um protótipo web integrador para demonstrar sua aplicabilidade prática.
The process of resume screening in the sales field faces significant challenges due to the diversity of formats, terminologies, and levels of detail. To overcome the reliance on manual analysis, this study investigated the application of natural language models (LLMs) to automate the extraction and standardization of relevant information from resumes. The methodology employed models such as GPT-4.1, GPT-4.1 Mini, and Gemini 2.5 Pro, as well as the ChatPDF tool and supporting libraries for text extraction. Specific prompts were designed to structure nominal and ordinal attributes consistently. The models’ performance was evaluated using metrics such as accuracy and mean arithmetic error. Subsequently, the GPT-4.1 model, which achieved the best performance, was applied to an expanded set of 50 resumes for validation. The extracted data were submitted to a classification model, resulting in a Mean Absolute Error (MAE) of 0.76 compared to real data, which confirms the reliability of the extraction method for automatic classification. The results demonstrate that language models are effective in data extraction, with GPT-4.1 standing out. It is concluded that the use of LLMs is a promising approach for automated screening, as it reduces manual effort and increases the consistency of evaluations , having been consolidated into an integrating web prototype to demonstrate its practical applicability.
The process of resume screening in the sales field faces significant challenges due to the diversity of formats, terminologies, and levels of detail. To overcome the reliance on manual analysis, this study investigated the application of natural language models (LLMs) to automate the extraction and standardization of relevant information from resumes. The methodology employed models such as GPT-4.1, GPT-4.1 Mini, and Gemini 2.5 Pro, as well as the ChatPDF tool and supporting libraries for text extraction. Specific prompts were designed to structure nominal and ordinal attributes consistently. The models’ performance was evaluated using metrics such as accuracy and mean arithmetic error. Subsequently, the GPT-4.1 model, which achieved the best performance, was applied to an expanded set of 50 resumes for validation. The extracted data were submitted to a classification model, resulting in a Mean Absolute Error (MAE) of 0.76 compared to real data, which confirms the reliability of the extraction method for automatic classification. The results demonstrate that language models are effective in data extraction, with GPT-4.1 standing out. It is concluded that the use of LLMs is a promising approach for automated screening, as it reduces manual effort and increases the consistency of evaluations , having been consolidated into an integrating web prototype to demonstrate its practical applicability.
Descrição
Mestrado de dupla diplomação com a UTFPR - Universidade Tecnológica Federal do Paraná
Palavras-chave
Linguagem natural Triagem de currículos Extração de dados Ilm Engenharia de prompts