A large logistics company uses a cloud-based system to manage its operations. This system generates semi-structured data such as JSON files…

Semi-structured data like JSON, XML, and CSV contains inherent structural elements (e.g., keys, tags, columns) that can be parsed to understand context. A schema-based parsing approach leverages this structure to identify fields likely to contain sensitive data. Combining this with machine learning (ML) allows the system to learn and recognize complex patterns and contextual clues that simple keyword searches would miss. This dual approach is highly effective for accurately discovering and classifying sensitive information within large volumes of semi-structured data, minimizing both false positives and false negatives.

A. Manual review is not scalable for a large company's data volumes and is highly prone to human error and inconsistency.

C. Basic text search ignores the data's structure (e.g., key-value pairs), leading to high error rates and an inability to understand context.

D. Simple keyword filtering is a rudimentary form of text search and suffers from the same limitations, lacking the sophistication to parse structured formats accurately.

1. National Institute of Standards and Technology (NIST). (2017). NIST Special Publication 500-299: NIST Big Data Interoperability Framework: Volume 7, Standards Roadmap. Section 4.3.2, "Data Parsing," discusses the need for parsing structured and semi-structured data formats like XML and JSON.

2. Gahi, Y., Guennoun, M., & El-Khatib, K. (2021). A survey on data loss prevention systems. Journal of Network and Computer Applications, 183, 103053. Section 3.2, "Content Analysis Techniques," discusses advanced methods beyond simple keywords, including statistical analysis and machine learning for more accurate data identification. DOI: https://doi.org/10.1016/j.jnca.2021.103053

3. Cloud Security Alliance (CSA). (2017). Security Guidance for Critical Areas of Focus in Cloud Computing v4.0. Domain 5, "Data Security Lifecycle," p. 60. This section emphasizes the importance of data discovery and classification as a foundational step, for which automated and intelligent tools are necessary at scale.