Statistical Foundations for a Google Play User-Review Sentiment Index: Signal Fusion, Shrinkage, Distributional Validation, and Dynamic Smoothing
Abstract
The decision problem is whether valence among the users who write reviews of an app has changed enough to warrant investigation of a release, outage, or support issue. We propose a statistical specification for that decision, applied to written Google Play reviews collected under a declared locale, language, sort order, retrieval cap, and window rule. That endpoint returns a selected collection rather than a random sample, so the index describes review writers under the declared protocol and not all app users. What is new is not a sentiment classifier but an auditable interface: an ordinal threshold model is the preferred representation of star ratings, a calibrated two-signal generalized least-squares estimator is retained as an interpretable linear approximation when held-out labels support its conditional-unbiasedness assumption, the estimand is declared before the estimator, and every downstream module--uncertainty, diagnostics, and regularization--is tied to that declaration. The framework separates the latent average of a fixed collected set from the mean of a protocol-defined population of review writers, so measurement and between-review variation are not conflated, and treats helpfulness and recency weights as a different policy target. Shrinkage, cross-app empirical Bayes, histogram and dependence diagnostics, and dynamic smoothing are specified as separate modules with defaults and mandatory sensitivity analyses. A reference implementation, a declared 6000-review three-app collection, and a synthetic stress test exercise the displayed equations: fusion reduced error in clean and small-window scenarios, coordinated contamination caused severe bias and interval failure, and the collected sample reproduces the abstention rules. No real-review validation is claimed.
Related
- Distributional sentiment modeling and anomaly detection for consumer complaint assessment
- Gromov-Wasserstein Barycenter Surrogates: Statistical Methodology, Distributional Limits and Applications
- Anthropogenic Forcing, Climate Change, and the Shape of Warming: Statistical Inference for Distributional Cointegration
- A Statistical Perspective on Knowledge Distillation: Foundations, Classical Methods, and Large Language Model Extensions
- Operator–Spectral χ‑Entropy and Hyperbolic Vacuum Structuring: Statistical Foundations of χ‑Reality
BibTeX
@misc{mandap2026statistical,
title = {{Statistical Foundations for a Google Play User-Review Sentiment Index: Signal Fusion, Shrinkage, Distributional Validation, and Dynamic Smoothing}},
author = {Marco Mandap},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2609.31513},
eprint = {2609.31513},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.31513},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Statistical Foundations for a Google Play User-Review Sentiment Index: Signal Fusion, Shrinkage, Distributional Validation, and Dynamic Smoothing
AU - Marco Mandap
PY - 2026
DA - 2026/10/04
PB - arXiv
DO - 10.48550/arxiv.2609.31513
UR - https://arxiv.org/abs/2609.31513
AB - The decision problem is whether valence among the users who write reviews of an app has changed enough to warrant investigation of a release, outage, or support issue. We propose a statistical specification for that decision, applied to written Google Play reviews collected under a declared locale, language, sort order, retrieval cap, and window rule. That endpoint returns a selected collection rather than a random sample, so the index describes review writers under the declared protocol and not all app users. What is new is not a sentiment classifier but an auditable interface: an ordinal threshold model is the preferred representation of star ratings, a calibrated two-signal generalized least-squares estimator is retained as an interpretable linear approximation when held-out labels support its conditional-unbiasedness assumption, the estimand is declared before the estimator, and every downstream module--uncertainty, diagnostics, and regularization--is tied to that declaration. The framework separates the latent average of a fixed collected set from the mean of a protocol-defined population of review writers, so measurement and between-review variation are not conflated, and treats helpfulness and recency weights as a different policy target. Shrinkage, cross-app empirical Bayes, histogram and dependence diagnostics, and dynamic smoothing are specified as separate modules with defaults and mandatory sensitivity analyses. A reference implementation, a declared 6000-review three-app collection, and a synthetic stress test exercise the displayed equations: fusion reduced error in clean and small-window scenarios, coordinated contamination caused severe bias and interval failure, and the collected sample reproduces the abstention rules. No real-review validation is claimed.
ER -