Pré-processamento, transformação e análise da base de dados de uma empresa atacadista de material elétrico em Curitiba

Abstract

The present work primarily addresses the preprocessing, transformation, and analysis of the database of a wholesale electrical materials company in Curitiba. The proposed theme holds significant relevance, given the company’s current challenge in obtaining reliable and easily accessible information from its product portfolio. This difficulty results in considerable time spent searching for such information. Therefore, we performed manual preprocessing of the database, followed by natural language processing (NLP), to handle and transform the data in a way that would allow manipulation through machine learning (ML) algorithms for automatic categorization. The results obtained included improved performance of the tested ML algorithms and the presentation of a processed, standardized, and reliable database in a CSV file format after the entire processing. Finally, we propose developing a recommendation system based on the treated data for future work.