<![CDATA[This research aims to determine the direction of research in the analysis of mixed data types. The world is currently filled with increasingly diverse data, especially in terms of data types, which are not only numerical or categorical but can be both (mixed). In Data Mining, the analysis of mixed data poses significant challenges because numerical and categorical data exhibit different properties. The research methodology employed in this study utilizes the PICOC framework (Population, Intervention, Comparison, Outcome, Context) to collect and review relevant literature. The primary findings from this comprehensive literature survey reveal that a majority of the research related to mixed data is published in reputable journals Q1, indicating sustained interest in the topic of mixed data analysis. Clustering models emerge as the most frequently used models in the field of mixed data analysis. However, it's noteworthy that accuracy metrics remain the predominant evaluation benchmark, often leading to comparisons with the ideal clustered data. The management of mixed data typically involves normalization techniques, specifically normalizing the scale to amalgamate the two types of data. The conclusion drawn from the results of the literature review is the necessity to develop unlabeled mixed data, encompassing both the model and metrics required to assess the outcomes. Additionally, this research emphasizes the significance of a comprehensive development model, ranging from feature selection to evaluation models. Therefore, the analysis of mixed data types remains a field with ample opportunities for exploration and potential innovation. This potential is particularly evident in the areas of dynamic model development and the ability to handle structured and extensive data.]]>
