RecordSearch Show Pages Userscript. GitHub Gist: instantly share code, notes, and snippets.

Not even the Premier's STEM Researcher of the Year 2025 can avoid a Tasmanian departmental job-cutting drive.

Presentation for the RDA Roadshow, Hobart, 22 September 2026

A collection of tools and examples to help you work with data from galleries, libraries, archives, and museums

A collection of tools and examples to help you work with data from galleries, libraries, archives, and museums

A collection of tools and examples to help you work with data from galleries, libraries, archives, and museums

A collection of tools and examples to help you work with data from galleries, libraries, archives, and museums

The Department of Prime Minister and Cabinet has compiled a large collection of speeches, interviews, press releases etc by Australian prime ministers and made them available through the PM Transcripts site. You can search through the transcripts, and download the underlying XML files using a simple API. Back in 2016 (and again in 2019), I downloaded all the XML-formatted transcript files and shared them through a GitHub repository to make it easier for researchers to use the collection as a whole. Many more transcripts have been added in the years since, so an update was long overdue! Last week I reharvested the complete collection, which now comprises 28,560 transcripts. The only change to the harvesting process was that the file that lists all of the transcript file names is now generated by a batch process and has to be downloaded manually. Once that was saved, I could just plug it in to the old code. As previously, I created a CSV index to the transcripts, and did a little analysis and visualisation. Here, for example, is the number of transcripts by year and prime minister – you can see a lot of Scomo and Albo has been added since my original harvest. I also aggregated the transcript files by prime minister – creating a single text file for each PM, as well as a zip file containing all the XML transcripts. That should make it easier to load up all the words of a particular PM for analysis. I completely overhauled the PM Transcripts section of the GLAM Workbench in the process, updating it to the latest format. There you’ll find all the notebooks I used to process the data, as well as details of the complete dataset. The transcripts are made available by the Commonwealth under a CC-BY licence.

The boundaries of electorates change over time through redistribution, so they have both a historical and spatial dimension. I’ve never been able to track down a complete set of historical federal electorate boundaries. Perhaps it’s out there somewhere and I’ve just been looking in the wrong places. However, as I noted in a previous post, I’ve been poking around in the _Parliamentary Handbook_ recently, and noticed that its electorate pages include an option to view maps of the boundaries as they changed over time through redistribution – all the historical versions seemed to be in there! Map of the electorate of Balaclava in the Parliamentary Handbook showing the dropdown list of dates to select a particular redistribution I started to play around with the API and found you could get a list of identifiers and dates for each boundary change by using the electorate name and state, for example here’s Balaclava: https://handbookapi.aph.gov.au/api/ElectorateMaps/ElectorateMaps?electorate=Balaclava&state=Victoria. Using the identifiers in this JSON response, you can request specific maps like this: https://handbookapi.aph.gov.au/api/ElectorateMaps/ElectorateMap?id=12004. The map request returns a big block of encoded data. After running it through decodify, I realised it was `base64` encoded. Decoding the data revealed the boundaries in KML format. Yay! I looped through all the electorates and boundaries saving all of the individual KML files. I then used kml2geojson to convert the KML files to GeoJSON format. The `description` field in the KML files include some useful structured data in a HTML table, so I parsed the HTML and saved the values as individual properties attached to each GeoJSON `Feature`. I’ve uploaded all the KML and GeoJSON files to a GitHub repository. I’ve also created a CSV index to the map files that contains the following fields: * `electorate` – electorate name * `state` – state in which the electorate is located * `map_id` – map identifier from Handbook * `start_date` – date this boundary came into effect * `end_date` – date this boundary was superceded * `geojson_file` – GeoJson file name Hopefully, this will make it easier for people to find and use federal electoral boundaries. According to its copyright page, content on the Parliament of Australia website is provided under a CC-BY-NC-ND licence.

I’m doing some work at the moment for the RAPID-CDL project, documenting the context of Commonwealth Hansard. As a result I’ve been poking around in various sources of related data, such as the _Parliamentary Handbook_ and Wikidata. I was surprised how much data was now available in the _Parliamentary Handbook_ – not just about parliamentarians, but about parties, elections, electorates, and more. It’s a really rich source for understanding how the composition of the Australian parliament has changed over time. Even better, if you look under the hood you’ll see there’s an API that makes all the data available in machine-readable form. For example, this url retrieves summary information about all parliamentarians in JSON format: https://handbookapi.aph.gov.au/api/individuals/. If you inspect an individual entry in the JSON, you’ll see each MP has a unique identifier in the `PHID` field. These identifiers are also used to identify speakers in Hansard, so it’s possible to make links between the text of Hansard and the biographical details in the Handbook. This same identifier is used in the `P10020` property on people records in Wikidata. However, when I ran a query to find records with this property, only about 800 of the 1,881 parliamentarians showed up. The people were in there, but not all of their records included the identifier. To fix this, I downloaded the data about parliamentarians from the Handbook and loaded it in to OpenRefine. I then reconciled the Handbook and Wikidata entries, using a slightly normalised version of the name and the `PHID`. As I worked through the options returned by the reconciliation process, I checked the Wikidata entries and made some manual edits, such as adding in parliamentary service dates. Once the two datasets were reconciled, I generated a set of QuickStatements to add the `P10020` property where it was missing, setting the value to match the `PHID` identifier in the Handbook. The result is that all the records for Australian federal politicians in Wikidata are now linked to the Parliamentary Handbook and Hansard. This query finds all 1,881 records with the `P10020` property. Now that they’re linked, I should probably do some more checking to see if there is other data from the Handbook that could be added. There’s a WikiProject page that provides a useful overview of Wikidata’s coverage of Australian politicians.

Join us at the RDA Roadshow in Hobart to learn about the latest in global research data innovation and collaboration!

A collection of tools and examples to help you work with data from galleries, libraries, archives, and museums

Build reveal.js presentations in your browser. No build step, no backend, no account — your decks live in your own browser storage.

The Tasmanian Museum and Art Gallery is caught in a job cut dispute, as the state government denies union claims.

The state government admits there are job cuts at the Tasmanian Museum and Art Gallery, but denies the union's claim that 30 per cent of the workforce will be affected.