Ambiguities, Built-in Biases and Flaws in Big Data Insight Extraction

TL;DR

This paper demonstrates that hierarchical classification methods in big data analysis can introduce systematic biases and ambiguities, affecting the reliability of extracted insights, even with complete data.

Contribution

It reveals how local ambiguity resolution in hierarchical models causes biases, highlighting fundamental flaws in common data reduction techniques.

Findings

Additional white aggregates emerge from local ambiguities.

Systematic bias increases with recursive aggregation.

Local symmetry-breaking decisions skew outcomes.

Abstract

I address the challenge of extracting reliable insights from large datasets using a simplified model that illustrates how hierarchical classification can distort outcomes. The model consists of discrete pixels labeled red, blue, or white. Red and blue indicate distinct properties, and white represents unclassified or ambiguous data. A macro-color is assigned only if one color holds a strict majority among the pixels. Otherwise, the aggregate is labeled white, reflecting uncertainty. This setup mimics a percolation threshold at fifty percent. Assuming direct access of the various proportions of colors is infeasible from the data, I implement a hierarchical coarse-graining procedure. Elements (first pixels, then aggregates) are recursively grouped and reclassified via local majority rules, producing ultimately a single super-aggregate whose color represents the inferred macro-property of the collection of pixels as a whole. Analytical results, supported by simulations, show that the process introduces additional white aggregates beyond white pixels, which could be initially present. These arise from groups lacking a clear majority, requiring arbitrary symmetry-breaking decisions to attribute a color to them. While each local resolution may appear minor and inconsequential, their repetitions introduce a growing systematic bias. Even with complete data, unavoidable asymmetries in local rules are shown to skew outcomes. This study highlights a critical limitation of recursive data reduction. Insight extraction is shaped not only by data quality, but by how local ambiguity is handled. That results in built-in biases. Thus, the related flaws are not due to the data, but due to structural choices made during local aggregations. Though based on a simple model, the findings expose the high likelihood of inherent flaws in widely used hierarchical classification techniques.