Detaildaten

Detaildaten, in data-science and statistics often called granular or microdata, are data records that encode detailed information at the level of individual units, events, or observations. They contrast with aggregated data that summarize groups or time periods. Detaildaten typically include multiple attributes per record and may be longitudinal, capturing changes over time.

Characteristics include high dimensionality, time stamps, unique identifiers, and fine-grained attributes. They enable analysis of dependencies, patterns, and causal relationships that are not visible in aggregated data. They require careful handling to protect privacy when records pertain to real persons or sensitive entities.

Applications include market research through customer transaction logs, operational analytics from event logs, epidemiology with patient- and time-level data, finance with trade-level information, and sensor networks capturing readings at high frequency. In machine learning, detaildaten serve as training data for predictive models and anomaly detection.

Data protection and ethics: Because detaildaten can reveal personal or sensitive information, they are subject to data protection laws such as the GDPR. Practices include pseudonymization, anonymization, data minimization, access controls, and, where appropriate, differential privacy to limit disclosure while preserving utility.

Challenges: Maintaining data quality across many attributes; handling missing values; integrating data from multiple sources; storage and processing costs; risks of re-identification in anonymized datasets; ensuring proper governance and documentation.

See also: microdata, granular data, longitudinal data, data privacy, differential privacy.