Data

Download and reuse the data

Name: Australian Parliament Registers of Interests
Creator: Kevin Rassool
License: https://creativecommons.org/licenses/by/4.0/

The full dataset in several formats, a field dictionary, a static JSON API, and SQL in the browser. Version v2.2026-10-02, data loaded 2026-10-02.

Downloads

Files for version v2.2026-10-02. The links always point at the latest version; each version also stays in its own folder. Sizes and sha256 checksums are published with the dataset in MANIFEST.json.

Data files, version v2.2026-10-02
File	Contents	Size	sha256	Licence
disclosures_v2.db	SQLite database (all tables)	published with the dataset	published with the dataset	CC BY 4.0
disclosures_v2.csv	CSV, one row per item	published with the dataset	published with the dataset	CC BY 4.0
disclosures_v2.csv.gz	CSV, gzip-compressed	published with the dataset	published with the dataset	CC BY 4.0
disclosures_v2.jsonl.gz	JSON Lines, gzip-compressed	published with the dataset	published with the dataset	CC BY 4.0
README.md	README: field dictionary, method, limitations	published with the dataset	published with the dataset
datapackage.json	Frictionless Data Package descriptor	published with the dataset	published with the dataset
MANIFEST.json	sha256 and size of every file	published with the dataset	published with the dataset

Open in Datasette Lite Kaggle

Datasette Lite runs SQLite in your browser and loads the whole database (the whole file) from the data host. Filter, facet and download any query as CSV.

Field dictionary

The columns of the CSV and of every per-page items.json. Empty cells are nulls.

Published columns (33)
Column	Type	Description
`item_id`	string	Stable id of the disclosed item (sha1 of the document hash and the item's content; see docs/v2/loading.md).
`chamber`	string	`house` (House of Representatives) or `senate`.
`parliament`	integer	Parliament number (43 = 2010-2013 … 48 = 2025-).
`member_id`	string	Stable member id (slug of the member's name, e.g. `tony_abbott`); the same person keeps one id across parliaments and chambers.
`member_name`	string	Member's full name.
`party`	string	Member's party for this parliament (start of term).
`political_bloc`	string	`Labor`, `Coalition` or `Crossbench`, derived from party.
`electorate_or_state`	string	House: electorate. Senate: state or territory.
`statement_date`	string	Date of the member's initial statement (YYYY-MM-DD), where printed.
`section`	integer	Register section number (1-14, the House form's numbering; Senate categories are mapped onto it).
`subsection`	string	Sub-part of the section where the form has one (`2(i)` family trusts, `2(ii)` other trusts), else empty.
`category`	string	Section name: Shareholding, Trust, Real estate, Directorship, Partnership, Liability, Bond/debenture, Account, Other asset, Income, Gift, Sponsored travel/hospitality, Membership, Other interest.
`owner`	string	Whose interest it is: `self`, `spouse`, `dependent_child` or `unknown`.
`entity_name_as_printed`	string	The company, organisation, trust or person the item names, as printed. Empty when the item names no entity (e.g. a house address).
`entity_id`	string	Standardised entity id; items naming the same organisation in different spellings share it. Empty for no entity or a generic term (e.g. `family trust`).
`entity_name`	string	Canonical name of the standardised entity.
`entity_type`	string	Entity type (listed_company, private_company, bank_or_financial, trust_or_fund, association_or_ngo, sporting_body, government_body, political_party, union, airline, media_or_entertainment, education, person, other). Empty for most one-off (singleton) entities.
`entity_asx_code`	string	ASX ticker when the entity matched an ASX-listed company (banks, airlines and media groups included, e.g. `CBA`, `QAN`), else empty.
`entity_match_method`	string	How the name was standardised: `curated` (hand table), `asx` (ASX listed-companies snapshot), `llm` (LLM grouping of variants), `singleton` (one-off name, its own entity) or `generic` (generic term, no entity).
`description`	string	The item as printed on the form (the full cell text).
`location`	string	Location, where the form gives one (real estate, travel).
`purpose`	string	Purpose, where the form gives one (real estate, liabilities).
`is_alteration`	integer	1 if the item comes from a later Notification of Alteration, 0 if from the initial statement.
`change_type`	string	`initial`, `added`, `removed`, `varied` or `unknown`.
`lodged_date`	string	Date the statement or alteration was lodged (YYYY-MM-DD; see date_precision).
`date_precision`	string	`day`, `month` (day set to 01) or `unknown` (no date printed; lodged_date empty).
`page`	integer	Page of the source PDF the item is on (1-based). Senate items: 1.
`extraction_confidence`	string	Extractor's confidence in the item: `high`, `medium` or `low` (low/medium = hard-to-read scan or ambiguous layout).
`source_file`	string	Source document path in the repository (House PDF, or the Senate API JSON payload).
`source_sha256`	string	sha256 of the source document.
`source_url`	string	Where the source document was downloaded from (APH).
`extraction_source`	string	`gemini-api` (LLM transcription of the PDF) or `senate-json` (Senate interests API, no LLM).
`extraction_model`	string	Model(s) that transcribed the PDF; `a+b` means model b handled pages model a refused.

Static JSON API

Every file below is a static file on this site, served with Access-Control-Allow-Origin: *, so you can fetch it from any web page or script. There is no key and no rate limit beyond the host's.

Routes
Path	Contents
`/data/items.json`	every item, columnar and dictionary-encoded (see columns)
`/data/members.json`	members sorted by id, with terms, item count and statements
`/data/entities.json`	entities with 2 or more items, plus every typed or ASX-coded entity
`/data/documents.json`	statements sorted by sha256, with source URL and URL class
`/data/items-extra.json`	item ids (row-aligned with items.json); entity names, types, ASX codes and match methods (aligned with its dictionaries); the published CSV header: what the explorer's CSV download adds
`/data/search.json`	member and entity names for the index-page search box
`/data/summary.json`	every number the overview page prints
`/data/schema.json`	this contract
`/members/<member_id>/items.json`	the member's items as row objects (published columns)
`/entities/<entity_id>/items.json`	the entity's items as row objects (published columns), for entities with 2 or more items

Example:

curl https://interests.kevinrassool.com/members/aaron_violi/items.json

The layout of data/items.json is described in data/schema.json. Its web_bundle_version is 1; it changes whenever the layout changes, so check it before you parse.

Metadata and updates

datapackage.json (Frictionless Data Package descriptor, on the data host)
schema.org Dataset metadata (JSON-LD) is embedded in this page and the overview.
Change feed (RSS): one entry per dataset version.
Kaggle: the CSV with the same README.
Source code and documentation

How to cite

Rassool, K. (2026). Australian Parliament Registers of Interests (v2.2026-10-02) [Data set]. Transcribed from the Parliament of Australia Register of Members' Interests and Register of Senators' Interests. https://interests.kevinrassool.com/

Please also credit the source: Parliament of Australia, Register of Members' Interests and Register of Senators' Interests (aph.gov.au).

Licences

The dataset is released under CC BY 4.0, to the extent we hold rights in it: share and adapt it for any purpose, commercial included, crediting this dataset and the source ("Parliament of Australia website"). The dataset records the facts each statement declares; the statements themselves (linked from every item) are published by the Parliament of Australia under CC BY-NC-ND 4.0, so reuse of the documents follows that licence.

Attribution text: "Australian Parliament Registers of Interests (v2.2026-10-02) by Kevin Rassool, CC BY 4.0, transcribed from the Parliament of Australia website."