# Things Learnt From eResearch 2024

**URL:** <https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854>\
**Category:** Technical\
**Tags:** eresearch\
**Created:** [31 October 2024 00:56 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854 "2024-10-31T00:56:18Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![TommyGatti](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/tommygatti/32/2353_2.png) [@TommyGatti](https://forum.access-hive.org.au/u/TommyGatti)\
**Post date:** [31 October 2024 00:56 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/1 "2024-10-31T00:56:18Z")

</div>

Just a general area for writing down cool things learnt/found/leads to follow up on after eResearch 2024. Feel free to add to this, everyone 😃

## Model Release

- CSCS spack public GitHub to see how they use `spack` and associated CI/CD
- Spack benchmarking
- [hpci](https://github.com/Garvan-Data-Science-Platform/hpci) - CI integrated with HPC jobs

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [31 October 2024 03:59 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/2 "2024-10-31T03:59:12Z")

</div>

Usual soulless mega-corp disclaimer, but this is quite cool:

> **[NotebookLM | Note Taking & Research Assistant Powered by AI](https://notebooklm.google/)**
>
> Use the power of AI for quick summarization and note taking, NotebookLM is your powerful virtual research assistant rooted in information you can trust.

There was a demo given where a notebook had a bunch of PDFs added to it and then used as the source for the LLM. Notable features:

- Domain aware LLM (knows the difference between science and poetry?)
- Includes referencing in overviews etc
- Can make an “engaging podcast” summary of material. Seems kinda dystopian, but as a podcast fan I can see the appeal of having an engaging conversation that summarises a bunch of dry material.

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [31 October 2024 04:01 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/3 "2024-10-31T04:01:26Z")

</div>

Only mentioned in passing, but looks very interesting and wondering if anyone has used it, or had any opinions

> **[Data Version Control · DVC](https://dvc.org/)**
>
> Open-source version control system for Data Science and Machine Learning projects. Git-like experience to organize your data, models, and experiments.

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [31 October 2024 04:13 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/4 "2024-10-31T04:13:39Z")

</div>

In Varvara and James’ [BoF about data and metadata standards](https://conference.eresearch.edu.au/data-and-metadata-standards-for-web-publication-of-research-data/) there was a decent amount of time spent talking about the differences and strengths/weaknesses of [ROCrate](https://www.researchobject.org/ro-crate/) and [frictionless.io](http://frictionless.io).

My take-away

Frictionless:

- simpler
- lower barrier to entry
- more human readable/writeable

ROCrate:

- Relies heavily on linked data approach to existing schema ([schema.org](http://schema.org))
- Means higher barrier to entry
- Also means more interoperability because of common schema
- Better suited to machine written workflows

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [31 October 2024 04:22 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/5 "2024-10-31T04:22:44Z")

</div>

Talk about [hpci](https://github.com/Garvan-Data-Science-Platform/hpci) also mentioned a couple of other similar tools

> **[GitHub - SvenMarcus/hpc-rocket](https://github.com/SvenMarcus/hpc-rocket)**
>
> Contribute to SvenMarcus/hpc-rocket development by creating an account on GitHub.

> **[Exascale Computing Project CI / Jacamar CI · GitLab](https://gitlab.com/ecp-ci/jacamar-ci)**
>
> HPC focused CI/CD driver for the GitLab custom executor. https://ecp-ci.gitlab.io/

Might be some interesting queueing related stuff @ben

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [31 October 2024 05:44 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/6 "2024-10-31T05:44:30Z")

</div>

Saw a really interesting [lightning talk by Alexis Tindall](https://conference.eresearch.edu.au/learning-from-our-users-reminding-us-of-the-value-of-keeping-humans-in-the-loop/) about the importance of keeping a user focus in service provision.

It was informed by attending this course on human centred problem solving:

[https://www.adelaide.edu.au/hr/leadership-and-talent/developing-at-adelaide/human-centred-problem-solving](https://www.adelaide.edu.au/hr/leadership-and-talent/developing-at-adelaide/human-centred-problem-solving)

---

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [2 November 2024 06:48 UTC](https://forum.access-hive.org.au/t/things-learnt-from-eresearch-2024/3854/7 "2024-11-02T06:48:16Z")

</div>

There was also a [great presentation about FileSender](https://conference.eresearch.edu.au/aarnets-filesender-evolution-capabilities-and-future-innovations/)

> **[AARNet FileSender](https://www.aarnet.edu.au/filesender)**
>
> A powerful and purpose-built application for researchers to send their files anywhere

which is a great service from [AARNet, Australia’s Research and Educational network provider](https://www.aarnet.edu.au/who-we-are).

FileSender is accessible to anyone with an Australian institutional university account. It’s free, and there are very few limits on the size of files. It’s like a dropbox download link but better!

You can also issue vouchers to collaborators who can’t access the service, allowing them to transfer files to you.
