Structured Evidence Inventory for "What Can Electricity Smart Meter Data Be Used For? An Intent-Oriented Taxonomy"

This dataset accompanies the manuscript What Can Electricity Smart Meter Data Be Used For? An Intent-Oriented Taxonomy. It provides a structured, machine-readable inventory of the evidence used to support the taxonomy developed through a concept-oriented narrative review of the analytical uses of electricity smart-meter data.

The inventory connects three complementary elements: the publications considered in the analytical synthesis, the substantive claims formulated in the taxonomy sections, and the evidential relationships between individual publications and claims. It supports inspection of the literature underlying the taxonomy and makes the authors’ synthesis and classification decisions more transparent.

Taxonomy covered by the dataset

The taxonomy comprises six blocks and seventeen topics:

• A. Understanding demand

  • A1: Demand Characterization

  • A2: Load Profiling

  • A3: Appliance-Level Decomposition

• B. Modeling expected demand

  • B1: Load Forecasting

  • B2: Baseline Modeling

• C. Acting on energy systems

  • C1: Demand Response

  • C2: Prosumer and Distributed Energy Resource Operations

• D. Evaluating events and performance

  • D1: Event Impact Evaluation

  • D2: Anomaly, Fraud, and Fault Detection

  • D3: Performance Assessment and Recommissioning

  • D4: Energy Poverty and Socio-Economic Analysis

• E. Supporting strategic decisions

  • E1: Tariffs and Markets

  • E2: System Planning

• F. Enabling smart-meter analytics

  • F1: Data Engineering

  • F2: Data Infrastructure

  • F3: Data Governance

  • F4: Data Availability

File format

The dataset consists of three semicolon-delimited UTF-8 CSV files:

1. sources.csv

2. claims.csv

3. taxonomy_citations.csv

The first row of each file contains the field names. Identifiers are case-sensitive and should be preserved exactly. An empty field indicates that the corresponding information is unavailable or inapplicable.

1. sources.csv

This file contains one record for each publication represented in the evidence inventory.

Fields

• key: Unique publication identifier corresponding to the BibTeX citation key used in the manuscript. It is the primary key of sources.csv and links publications to taxonomy_citations.csv.

• title: Full title of the publication.

• year: Bibliographic year assigned to the publication. For publications awaiting definitive issue assignment, it corresponds to the year in which the final peer-reviewed version was made available online.

• publication_type: Formal publication format, represented through a controlled vocabulary.

• study_type: Primary methodological design or scholarly function of the publication within the evidence inventory, represented through a controlled vocabulary.

Controlled vocabulary for publication_type

• journal_article: Article published in a peer-reviewed scholarly journal.

• conference_paper: Paper published in peer-reviewed conference proceedings.

• book_chapter: Scholarly contribution published as a chapter in an edited volume.

• report: Research, technical, institutional, or policy report issued by an identifiable organization.

Controlled vocabulary for study_type

• empirical_study: Study that analyzes observed, experimental, survey, or operational data to produce original empirical findings.

• methodological_study: Study whose main contribution is the development, adaptation, validation, or comparison of an analytical or computational method.

• review: Study that synthesizes an existing body of literature, including narrative, systematic, scoping, and other structured review designs.

• dataset_descriptor: Publication whose primary contribution is the documentation, release, or characterization of a dataset.

• dataset_review: Publication that identifies, catalogues, compares, or evaluates multiple datasets and their potential applications.

• policy_analysis: Study focused primarily on legal, regulatory, governance, institutional, ethical, or public-policy questions.

The assigned study_type represents the publication’s principal function in the inventory. A publication may contain elements associated with several designs while receiving one primary classification.

2. claims.csv

This file contains the substantive claims identified in Sections A1–F4 of the taxonomy. Claims summarize definitions, methods, findings, applications, challenges, and contextual propositions supported by the reviewed literature.

Fields

• claim_id: Unique claim identifier composed of the taxonomy topic code and a sequential number. For example, A1-C01 identifies the first recorded claim in Topic A1. It is the primary key of claims.csv and links claims to taxonomy_citations.csv.

• topic: Taxonomy topic to which the claim belongs, expressed through one of the codes A1–F4 defined above.

• claim: English-language formulation of the substantive claim supported by the cited literature.

• claim_type: Function performed by the claim within the analytical synthesis, represented through a controlled vocabulary.

Controlled vocabulary for claim_type

• definition: Defines an analytical intent, concept, output, scope, or conceptual boundary.

• method: Describes a method, model, analytical procedure, technical approach, or data-processing strategy.

• finding: Reports or synthesizes an empirical result or recurring observation from the literature.

• application: Describes a practical use, operational implementation, decision context, or area in which the analysis can be applied.

• challenge: Identifies a limitation, barrier, risk, unresolved problem, or methodological difficulty.

• context: Provides conceptual, institutional, regulatory, technical, or historical context required to interpret the analytical topic.

Each claim receives one primary claim_type according to its main function in the manuscript.

3. taxonomy_citations.csv

This file represents the many-to-many relationships between the claims in claims.csv and the publications in sources.csv.

Each row records the evidential role performed by one publication in relation to one specific claim. A publication may support several claims, and a claim may be supported by several publications.

Fields

• claim_id: Foreign key identifying a claim in claims.csv.

• key: Foreign key identifying a publication in sources.csv.

• role: Evidential function performed by the publication in relation to the specified claim, represented through a controlled vocabulary.

The combination of claim_id and key identifies a publication–claim relationship. Evidential roles are claim-specific: the same publication may perform different roles in relation to different claims.

Controlled vocabulary for role

• core: The publication directly and substantively supports the claim. It constitutes central evidence for the corresponding statement.

• supporting: The publication complements, corroborates, illustrates, or extends the evidence supplied by the core sources.

• contextual: The publication provides relevant conceptual, methodological, legal, technical, or historical background while serving a secondary evidential function for the specific claim.

Relationships among the files

The three files form a relational evidence structure:

• sources.csv identifies and characterizes the publications.

• claims.csv records the substantive propositions supported by the literature.

• taxonomy_citations.csv connects publications and claims and specifies the evidential role of each connection.

The relational links are:

• sources.key → taxonomy_citations.key

• claims.claim_id → taxonomy_citations.claim_id

Together, these relationships allow users to:

• identify the publications supporting each claim;

• examine the claims supported by an individual publication;

• distinguish core, supporting, and contextual evidence;

• inspect how evidence is distributed across taxonomy topics;

• extend, validate, or compare the proposed taxonomy.

Classification principles

Publications and claims were organized according to the primary analytical intent represented by the output of each study. Classification considered four related questions:

1. What is the direct output of the analysis?

2. What type of quantity is examined or estimated?

3. What reference gives that output its meaning?

4. What action, evaluation, or decision is the output intended to support?

Topic assignments, claim formulations, study types, and evidential roles represent the authors’ structured interpretation of the publications within this conceptual framework.

Scope of the evidence inventory

The contemporary evidence base covers peer-reviewed English-language publications published or made available online in final peer-reviewed form between January 2020 and 30 June 2026, irrespective of their subsequent assignment to a journal issue. Earlier publications are included when they provide methodological foundations or influential conceptual context.

The inventory covers the substantive taxonomy sections A1–F4. References used exclusively for general introductory background, review methodology, or concluding synthesis fall outside its analytical scope.

The dataset supports transparency and inspectability in the accompanying concept-oriented narrative review. It provides a structured record of the relationship between the review’s substantive claims and their supporting literature and can facilitate future validation, extension, updating, or comparison of the proposed taxonomy.