leeroy-jankins/Government-Performance-and-Results-Act

Government Performance and Results Act of 1993 Corpus

Dataset Description

The Government Performance and Results Act of 1993 Corpus is a processed legal and public-administration dataset derived from the Government Performance and Results Act of 1993, commonly abbreviated as GPRA.

GPRA established a statutory framework for strategic planning, annual performance planning, program performance measurement, and performance reporting across the executive branch of the United States Government.

The Act was enacted to improve federal program effectiveness and public accountability by requiring agencies to:

• Establish strategic goals
• Define measurable performance goals
• Connect program activities to intended results
• Measure progress toward those goals
• Report actual performance
• Explain unmet goals
• Use performance information in management and decision-making
• Improve congressional oversight of federal programs
• Strengthen public confidence in the federal government

This repository includes the original source document and multiple processed representations prepared for natural-language processing, semantic search, retrieval, analysis, and language-model workflows.

The repository contains:

• The original source document in PDF format
• Token-based chunks in CSV format
• Token-level records in CSV format
• A normalized text corpus in TXT format

The source text was divided into chunks containing approximately 1,000 tokens, with an overlap of 200 tokens between consecutive chunks.

The normalized corpus was prepared using basic preprocessing that:

• Removes HTML
• Removes Markdown syntax
• Removes symbols and unsupported characters
• Removes identified extraction errors
• Collapses repeated whitespace
• Converts text to lowercase

Source Document

The principal source represented by this dataset is:

• Title: Government Performance and Results Act of 1993
• Abbreviation: GPRA
• Public law: Public Law 103-62
• Statutes at Large citation: 107 Stat. 285
• Bill: S. 20
• Congress: 103rd Congress
• Enactment date: August 3, 1993
• Document type: United States federal statute

The repository PDF represents the source document used to create the processed dataset artifacts.

Users should inspect the source PDF and repository metadata to determine:

• The exact source filename
• The publication source
• The retrieval date
• The PDF page count
• The document checksum
• Whether the document contains only the enacted law or related legislative material

Legislative Background

Before GPRA, federal agencies were not subject to a uniform government-wide statutory framework requiring strategic plans, annual performance goals, and recurring reports comparing planned results with actual performance.

Congress enacted GPRA to address concerns involving:

• Insufficient focus on program results
• Weak connections between agency activities and intended outcomes
• Limited performance information for congressional decision-making
• Difficulty determining whether federal programs achieved their objectives
• Inconsistent strategic planning across agencies
• Inadequate accountability for program performance
• Limited public visibility into federal program results
• Management processes focused more heavily on inputs than outcomes

GPRA introduced a results-oriented management framework intended to shift attention from resources and activities toward measurable program outcomes.

Principal Objectives

The Act was designed to:

1. Improve public confidence in the federal government by holding agencies accountable for program results.
2. Initiate program performance reform through pilot projects involving performance goals, measurement, and reporting.
3. Improve federal program effectiveness and public accountability.
4. Help federal managers improve service delivery.
5. Improve congressional decision-making by providing objective information about program performance.
6. Improve internal management by requiring agencies to plan for and evaluate program results.

Core Requirements

Strategic Plans

GPRA required covered federal agencies to prepare strategic plans addressing their major functions and operations.

Strategic plans generally included:

• A comprehensive mission statement
• General goals and objectives
• A description of how goals would be achieved
• The relationship between long-term goals and annual performance goals
• Identification of key external factors that could affect achievement
• A description of program evaluations used in developing or revising the plan
• A schedule for future program evaluations

Strategic plans were intended to provide a long-term framework for agency decision-making and performance management.

Annual Performance Plans

The Act required annual performance plans covering program activities included in the federal budget.

Performance plans generally identified:

• Annual performance goals
• The level of performance to be achieved
• Objective, quantifiable, and measurable indicators
• Operational processes and resources required to meet goals
• Methods used to verify and validate performance information
• Alternative forms of measurement when quantitative goals were not feasible

Annual plans connected agency strategy with near-term operational and budgetary activity.

Program Performance Reports

GPRA required agencies to report on actual performance compared with the goals established in annual performance plans.

Performance reports generally addressed:

• Actual program performance
• Comparison of actual results with planned goals
• Reasons performance goals were not met
• Plans and schedules for achieving unmet goals
• Explanations when a goal was impractical or infeasible
• Performance trends
• Relevant program evaluation findings

These reports were intended to make agency performance visible to Congress, the President, agency managers, and the public.

Performance Indicators

Performance indicators provide measurable values used to assess progress toward a goal.

Indicators may measure:

• Outputs
• Outcomes
• Quality
• Efficiency
• Timeliness
• Cost
• Customer service
• Program reach
• Compliance
• Operational effectiveness

GPRA emphasized objective and measurable performance information whenever feasible.

Program Evaluation

Program evaluation supports systematic assessment of whether a program is operating as intended and achieving its expected results.

Evaluations may address:

• Program design
• Implementation
• Outcomes
• Cost-effectiveness
• Efficiency
• Unintended consequences
• Alternative approaches
• Evidence supporting program continuation or modification

Managerial Accountability and Flexibility

GPRA included pilot provisions intended to test whether managers could receive greater operational flexibility in exchange for increased accountability for measurable results.

Performance Budgeting

The Act also established pilot efforts relating budget presentation to program performance.

These provisions sought to improve the connection among:

• Resources
• Program activities
• Outputs
• Outcomes
• Management decisions
• Congressional oversight

Relationship to Later Legislation

GPRA was amended and expanded by later legislation, most notably the GPRA Modernization Act of 2010.

Later reforms introduced or strengthened concepts such as:

• Agency priority goals
• Cross-agency priority goals
• Quarterly performance reviews
• Chief Operating Officers
• Performance Improvement Officers
• Performance Improvement Councils
• Centralized publication of performance information
• More frequent strategic planning cycles
• Greater coordination between performance and budget processes

This dataset represents the source document included in the repository. It should not be treated as a complete representation of all current federal performance-management requirements.

Dataset Purpose

The dataset is intended to support:

• Retrieval-augmented generation
• Semantic search
• Document retrieval
• Legal and policy research
• Federal performance-management research
• Strategic-planning research
• Public-administration education
• Document-grounded question answering
• Embedding generation
• Information extraction
• Topic classification
• Summarization
• Search indexing
• Knowledge-base development
• Language-model evaluation
• Regulatory comparison
• Program-evaluation research

The overlapping token windows help preserve context when statutory provisions, lists, definitions, or reporting requirements cross chunk boundaries.

Repository Contents

The repository contains four principal artifact types.

Artifact	Format	Description
Original document	PDF	Source copy of the Government Performance and Results Act used to construct the dataset.
Chunks	CSV	Normalized text divided into approximately 1,000-token chunks with 200-token overlap.
Tokens	CSV	Token-level representation of the normalized corpus.
Corpus	TXT	Complete normalized plain-text corpus derived from the source PDF.

Example filenames may include:

government_performance_and_results_act_1993.pdf
gpra_1993_chunks.csv
gpra_1993_tokens.csv
gpra_1993_corpus.txt
README.md

Actual filenames may differ from these examples.

Processing Methodology

Source Acquisition

The source PDF was obtained from an official or authoritative United States Government publication source.

The PDF is retained in the repository as the primary source artifact for the corresponding dataset release.

Text Extraction

Text was extracted from the PDF before preprocessing, tokenization, and chunk generation.

PDF extraction may introduce artifacts involving:

• Page headers
• Page footers
• Page numbers
• Line breaks
• Hyphenated line endings
• Multi-column layouts
• Statutory headings
• Section numbering
• Citations
• Signature blocks
• Character encoding
• Scanned pages
• Marginal annotations

The preprocessing process reduces common artifacts but does not guarantee perfect reconstruction of the original document structure.

Corpus Preprocessing

The extracted text was normalized through a basic preprocessing pipeline.

The preprocessing operations include:

1. Removing HTML tags and encoded HTML artifacts
2. Removing Markdown syntax
3. Removing symbols and unsupported characters
4. Removing identified extraction errors
5. Collapsing repeated spaces, tabs, and line breaks
6. Normalizing whitespace
7. Converting all text to lowercase

The resulting corpus is optimized for machine processing rather than exact visual or typographic reproduction.

Tokenization

The normalized corpus was tokenized before chunk generation.

Tokenization supports:

• Token counting
• Chunk creation
• Context-window management
• Embedding generation
• Retrieval workflows
• Token-frequency analysis
• Language-model input preparation

Token counts depend on the tokenizer used during dataset generation.

For reproducibility, the repository should record:

• Tokenizer name
• Tokenizer version
• Encoding name
• Vocabulary version
• Special-token handling
• Truncation settings
• Padding settings

Chunking

The normalized corpus was divided using a sliding token window.

Parameter	Value
Chunk size	1,000 tokens
Chunk overlap	200 tokens
Approximate new content per chunk	800 tokens
Chunking unit	Tokens
Chunking method	Sliding window
Preprocessing applied before chunking	Yes

Each new chunk begins approximately 800 tokens after the beginning of the preceding chunk.

Chunking Illustration

Chunk 1:
Tokens 0 through 999

Chunk 2:
Tokens 800 through 1799

Chunk 3:
Tokens 1600 through 2599

The overlap helps preserve context when:

• A statutory requirement crosses a chunk boundary
• A list continues into the next segment
• A definition appears near the end of a chunk
• A strategic-planning provision spans multiple paragraphs
• A reporting requirement is separated from a qualifying condition
• An agency responsibility is separated from its implementation details

Dataset Structure

Chunks Dataset

The chunks CSV contains sequential overlapping text segments generated from the normalized corpus.

A typical schema may include:

Field	Type	Description
chunk_id	Integer	Unique sequential chunk identifier.
text	String	Normalized text contained in the chunk.
token_count	Integer	Number of tokens in the chunk.
token_start	Integer	Starting token offset in the complete corpus.
token_end	Integer	Ending token offset in the complete corpus.
document_id	String	Stable identifier for the source document.
source_file	String	Filename of the source PDF.

The actual repository schema may contain only a subset of these fields.

Example Chunk Record

{
  "chunk_id": 1,
  "text": "the purposes of this act are to improve the confidence of the american people in the capability of the federal government...",
  "token_count": 1000,
  "token_start": 0,
  "token_end": 999,
  "document_id": "gpra-1993",
  "source_file": "government_performance_and_results_act_1993.pdf"
}

Tokens Dataset

The tokens CSV contains a token-level representation of the normalized corpus.

A typical schema may include:

Field	Type	Description
token_id	Integer	Unique sequential token identifier.
token	String	Token text or decoded token value.
token_index	Integer	Position of the token in the complete corpus.
token_value	Integer	Numeric tokenizer identifier, when available.
chunk_id	Integer	Associated chunk identifier, when available.
document_id	String	Identifier for the source document.
source_file	String	Source PDF filename.

Example Token Record

{
  "token_id": 1,
  "token": "the",
  "token_index": 0,
  "token_value": 1820,
  "chunk_id": 1,
  "document_id": "gpra-1993"
}

Corpus Dataset

The corpus TXT file contains the complete normalized text.

The corpus:

• Is lowercased
• Uses collapsed whitespace
• Excludes HTML
• Excludes Markdown syntax
• Excludes unsupported symbols
• Excludes identified extraction errors
• Does not reliably preserve PDF pagination
• Does not reliably preserve original formatting
• May flatten lists and subsection hierarchies

The corpus is suitable for:

• Full-text indexing
• Keyword analysis
• Corpus statistics
• Retokenization
• Alternative chunking strategies
• Embedding generation
• Topic modeling
• Text classification
• Language-model experiments

Original PDF

The original PDF should be used whenever exact statutory wording or formatting matters.

The PDF is the preferred artifact for verifying:

• Capitalization
• Punctuation
• Section numbers
• Statutory citations
• Paragraph structure
• Page numbers
• Headings
• Enumerated requirements
• Defined terms
• Signature information
• Original wording

Dataset Splits

The dataset is generally provided as a single split because it represents one statutory corpus.

Split	Description
train	Complete set of chunks or token records derived from the source document.

Users may create additional splits using:

• Section-based splitting
• Topic-based splitting
• Holdout provisions
• Retrieval evaluation subsets
• Manually curated question-answer sets
• Strategic-planning and reporting subsets

Random splitting should be used cautiously because adjacent chunks contain overlapping text.

Neighboring or overlapping chunks should remain in the same split to prevent information leakage between training, validation, and test sets.

Loading the Dataset

Load from Hugging Face

from datasets import load_dataset


dataset = load_dataset(
    "YOUR_HUGGING_FACE_USERNAME/gpra-1993-corpus"
)

print(dataset)

Load the Chunks CSV

from datasets import load_dataset


dataset = load_dataset(
    "csv",
    data_files={
        "train": "gpra_1993_chunks.csv"
    }
)

print(dataset["train"][0])

Load the Tokens CSV

from datasets import load_dataset


dataset = load_dataset(
    "csv",
    data_files={
        "train": "gpra_1993_tokens.csv"
    }
)

print(dataset["train"][0])

Load the Corpus TXT

from pathlib import Path


corpus_path = Path(
    "gpra_1993_corpus.txt"
)

corpus = corpus_path.read_text(
    encoding="utf-8"
)

print(corpus[:1000])

Load the CSV Files with pandas

import pandas as pd


df_chunks = pd.read_csv(
    "gpra_1993_chunks.csv"
)

df_tokens = pd.read_csv(
    "gpra_1993_tokens.csv"
)

print(df_chunks.head())
print(df_tokens.head())

Retrieval-Augmented Generation

The chunk dataset is designed for retrieval-augmented generation workflows.

A typical workflow includes:

1. Load the chunks CSV.
2. Generate an embedding for each chunk.
3. Store embeddings in a vector index.
4. Generate an embedding for the user question.
5. Retrieve the most relevant chunks.
6. Supply the retrieved chunks to a language model.
7. Generate an answer grounded in the retrieved text.
8. Return source metadata with the answer.

Example Embedding Preparation

from datasets import load_dataset
from sentence_transformers import SentenceTransformer


dataset = load_dataset(
    "csv",
    data_files={
        "train": "gpra_1993_chunks.csv"
    },
    split="train"
)

model = SentenceTransformer(
    "sentence-transformers/all-MiniLM-L6-v2"
)

embeddings = model.encode(
    dataset["text"],
    normalize_embeddings=True,
    show_progress_bar=True
)

print(embeddings.shape)

Recommended Retrieval Prompt

You are a federal performance-management research assistant.

Answer the question using only the supplied Government Performance and
Results Act context. Distinguish requirements contained in the original
1993 statute from requirements introduced through later legislation or
implementation guidance. Identify the relevant statutory section when
available. State when the supplied context is insufficient.

Context:
{retrieved_chunks}

Question:
{question}

Answer:

Major Subject Areas

The corpus addresses subject areas including:

• Congressional findings and purposes
• Results-oriented government
• Strategic planning
• Agency mission statements
• General goals and objectives
• Performance goals
• Performance indicators
• Annual performance plans
• Program performance reports
• Program activities
• Program evaluation
• Measurement methodologies
• Data verification and validation
• Managerial accountability
• Managerial flexibility
• Performance budgeting
• Pilot projects
• Congressional oversight
• Office of Management and Budget responsibilities
• Agency implementation
• Performance information
• Public accountability
• Federal program effectiveness

Intended Uses

Appropriate uses include:

• Federal performance-management research
• Public-administration education
• Retrieval-augmented generation
• Semantic search
• Document-grounded question answering
• Strategic-planning analysis
• Performance-measurement analysis
• Program-evaluation research
• Federal management research
• Legal and policy information retrieval
• Embedding-model evaluation
• Search-relevance testing
• Topic modeling
• Text classification
• Summarization
• Knowledge-graph development
• Statutory comparison

Example Questions

The dataset may support questions such as:

• What was the purpose of GPRA?
• What must an agency strategic plan contain?
• How are annual performance goals established?
• What is a performance indicator?
• What information belongs in an annual performance plan?
• What must an agency report when it fails to achieve a performance goal?
• How does GPRA address program evaluation?
• What role does OMB have under GPRA?
• How does GPRA connect budgeting and performance?
• What are managerial accountability and flexibility waivers?
• What pilot projects were established by GPRA?
• How does GPRA support congressional oversight?
• What is the difference between an output and an outcome?
• How did GPRA change federal management?
• How does the GPRA Modernization Act differ from the original statute?

Out-of-Scope Uses

The dataset should not be treated as:

• Legal advice
• An official interpretation of GPRA
• A replacement for the enacted statute
• A replacement for the current United States Code
• A replacement for the GPRA Modernization Act of 2010
• A replacement for current OMB guidance
• A replacement for agency-specific performance policies
• A definitive statement of current reporting deadlines
• A complete description of current federal performance-management practice
• A basis for formal budget or program decisions
• A substitute for program evaluation
• A substitute for agency legal or policy review

The dataset should not be used by itself to make legal, budgetary, regulatory, programmatic, or administrative decisions.

Data Quality

The dataset was designed to provide:

• A preserved source PDF
• A normalized text corpus
• Structured token records
• Overlapping token-based chunks
• Machine-readable CSV files
• Context windows suitable for retrieval
• Compatibility with common NLP libraries
• Traceability to the source document

Potential limitations include:

• PDF extraction errors
• Missing or damaged characters
• Broken statutory citations
• Loss of page structure
• Loss of capitalization
• Loss of punctuation or symbols
• Flattened headings
• Merged paragraphs
• Split words
• Repeated headers or footers
• Tokenizer-dependent token counts
• Duplicate text created by overlap
• Incomplete provenance metadata
• Differences between the original statute and current law

Effects of Preprocessing

Lowercasing

Lowercasing improves normalization but removes distinctions involving:

• Acronyms
• Proper nouns
• Defined terms
• Headings
• Capitalized statutory titles
• Emphasized text

Symbol Removal

Removing symbols may affect:

• Statutory citations
• Section identifiers
• Parenthetical references
• Hyphenated terms
• Dates
• Percentages
• Enumerated lists

Whitespace Collapsing

Whitespace normalization may flatten:

• Paragraph boundaries
• Indented subsections
• Lists
• Tables
• Multi-column layouts
• Signature blocks

Error Removal

Removal of extraction errors improves general corpus quality but may remove valid text if it resembles a known extraction artifact.

Chunk Overlap

The 200-token overlap creates intentional duplication across adjacent chunks.

Users should account for this duplication when:

• Calculating term frequencies
• Measuring corpus size
• Training language models
• Creating dataset splits
• Aggregating retrieved passages
• Evaluating retrieval diversity
• Computing document statistics

Validation Considerations

Original Statute and Current Law

The original 1993 Act should not automatically be treated as a complete statement of current federal performance-management law.

Later amendments and guidance may materially affect current requirements.

Exact-Language Verification

The original PDF or current United States Code should be consulted whenever exact wording, section numbering, or legal interpretation matters.

Strategic and Annual Requirements

Models should distinguish among:

• Long-term strategic goals
• Annual performance goals
• Performance indicators
• Program activities
• Program evaluations
• Actual results
• Explanations for unmet goals

Output and Outcome Measures

An output measures a product or service produced by a program.

An outcome measures the result or effect associated with the program.

These concepts should not be treated as interchangeable.

Chunk Leakage

Adjacent and overlapping chunks should remain in the same dataset split to prevent information leakage.

Citation Accuracy

Dates, section numbers, public law references, and statutory citations should be verified against the source PDF.

Limitations

Historical Source

The source statute was enacted in 1993. Federal performance-management law and practice have evolved significantly since enactment.

Loss of Formatting

The normalized corpus does not reliably preserve:

• Page layout
• Statutory indentation
• Subsection hierarchy
• Tables
• Footnotes
• Marginal notes
• Signature blocks
• Original typography

Incomplete Semantic Preservation

Removing symbols and formatting may alter the interpretation of:

• Section references
• Lists
• Defined terms
• Citations
• Dates
• Parenthetical qualifications
• Cross-references

Tokenizer Dependence

The 1,000-token chunk size and 200-token overlap reflect the tokenizer used during dataset construction.

A different tokenizer may produce different:

• Token counts
• Chunk boundaries
• Numeric identifiers
• Decoded token values

No Guarantee of Legal Currency

The dataset is a research artifact and does not guarantee that every provision remains current.

Limited Scope

The dataset does not comprehensively include:

• GPRA Modernization Act of 2010
• Current OMB performance guidance
• Agency strategic plans
• Agency performance plans
• Agency performance reports
• Federal budget guidance
• Program evaluation standards
• Evidence Act requirements
• Agency-specific performance-management policies

Related Authorities and Resources

Users may benefit from reviewing GPRA together with:

• GPRA Modernization Act of 2010
• Foundations for Evidence-Based Policymaking Act of 2018
• Chief Financial Officers Act of 1990
• Government Management Reform Act of 1994
• Federal Managers’ Financial Integrity Act of 1982
• Digital Accountability and Transparency Act of 2014
• Relevant provisions of Titles 5, 31, and 39 of the United States Code
• Office of Management and Budget Circular A-11
• Agency strategic plans
• Agency performance plans
• Agency performance reports
• Federal budget documents
• Government Accountability Office evaluations of GPRA implementation

These materials are not necessarily included in the repository.

Ethical and Operational Considerations

The dataset contains public legal and public-administration information. It is not intended to contain:

• Classified information
• Controlled unclassified information
• Personally identifiable information
• Proprietary information
• Procurement-sensitive information
• Confidential agency performance data

Downstream applications should nevertheless be evaluated for:

• Incorrect statutory interpretation
• Reliance on superseded requirements
• Hallucinated performance obligations
• Confusion between outputs and outcomes
• Misrepresentation of agency responsibilities
• Failure to distinguish law from guidance
• Inadequate disclosure of uncertainty
• Failure to cite authoritative sources
• Use in high-impact decisions without qualified review

Formal performance, budget, legal, and program-management decisions should be reviewed by qualified personnel using current authoritative sources.

Licensing

The original federal statute is a work of the United States Government and is generally not subject to copyright protection within the United States.

The repository’s derived materials may include:

• Chunk organization
• Token records
• Corpus formatting
• Metadata
• Documentation
• Processing scripts
• Dataset configurations

These derived materials may be subject to the repository’s stated license.

The metadata uses license: other because the source document is a United States Government work and the repository may apply separate terms to the derived artifacts.

Users are responsible for evaluating any licensing requirements applicable to redistribution, derivative works, or use outside the United States.

Citation

When using this dataset, cite both the dataset repository and the underlying statute.

Dataset Citation

@dataset{gpra_1993_corpus,
  author       = {DATASET_AUTHOR},
  title        = {Government Performance and Results Act of 1993 Corpus},
  year         = {DATASET_PUBLICATION_YEAR},
  publisher    = {Hugging Face},
  url          = {DATASET_REPOSITORY_URL},
  note         = {Tokenized and chunked corpus derived from Public Law 103-62}
}

Statutory Citation

Government Performance and Results Act of 1993,
Pub. L. No. 103-62, 107 Stat. 285 (1993).

Dataset Maintenance

The dataset should be reviewed periodically to:

• Confirm the source PDF
• Record the source URL
• Record the source retrieval date
• Add a source checksum
• Correct extraction errors
• Improve preprocessing rules
• Validate token counts
• Record tokenizer metadata
• Detect missing pages
• Remove unintended duplicate records
• Preserve statutory section identifiers
• Add page-level provenance
• Document schema changes
• Version the preprocessing pipeline
• Rebuild chunks when the source changes
• Compare the original statute with current codified law

Material changes should be documented in repository commits or release notes.

Recommended Future Enhancements

Potential enhancements include:

• Statutory section metadata
• Current United States Code citations
• Page numbers
• Source character offsets
• Source token offsets
• Exact PDF provenance
• Document checksums
• Tokenizer name and version
• Preprocessing-version metadata
• Preserved-case corpus
• Raw extracted text
• Section-aware chunking
• Paragraph-aware chunking
• Semantic chunking
• Heading-aware chunking
• Manually curated question-answer pairs
• Retrieval evaluation benchmarks
• Strategic-planning topic labels
• Performance-measurement labels
• Output and outcome classifications
• GPRA Modernization Act cross-references
• Sparse retrieval indexes
• Dense retrieval indexes

An expanded chunk schema could include:

Field	Type	Description
chunk_id	Integer	Unique chunk identifier.
document_id	String	Stable source-document identifier.
section	String	Relevant statutory section.
subsection	String	Relevant statutory subsection.
topic	String	Strategic planning, performance measurement, reporting, or another subject.
page_start	Integer	First source page represented.
page_end	Integer	Last source page represented.
token_start	Integer	Starting corpus token offset.
token_end	Integer	Ending corpus token offset.
token_count	Integer	Number of tokens in the chunk.
text	String	Normalized chunk text.
source_file	String	Source PDF filename.
source_url	String	Official source location.
source_date	String	Enactment or publication date.
retrieved_at	String	Source retrieval date.
checksum	String	Hash of the source or chunk content.
tokenizer	String	Tokenizer or encoding used.
preprocessing_version	String	Version of the preprocessing pipeline.

Reproducibility Metadata

For reproducible dataset generation, the repository should record:

• Source URL
• Source filename
• Source checksum
• Source retrieval date
• PDF page count
• Extraction library
• Extraction-library version
• Preprocessing script
• Preprocessing version
• Tokenizer
• Tokenizer version
• Encoding
• Chunk size
• Chunk overlap
• CSV schema
• Dataset generation date
• Python version
• Dependency versions

Disclaimer

This dataset is provided for research, education, experimentation, and software-development purposes.

It is not an official publication of the United States Government and is not affiliated with, endorsed by, or sponsored by Congress, the Office of Management and Budget, the Government Accountability Office, or any federal agency.

The dataset does not constitute legal, budgetary, program-management, performance-management, or policy advice.

Users are responsible for validating all material information against the original statute, current codified law, current OMB guidance, and other applicable authoritative sources.