FAIR Data Digest #7
What is CLARIN, the Common Language Resources and Technology Infrastructure? My 1000ths Wikidata edit and FAIR Jupyter Notebooks
Hi everyone,
last week I told you about European Research Infrastructures Consortia or ERIC in short. Today we gonna look into one of these ERICs that is active in the Cultural Heritage field. I also have to celebrate a tiny anniversary related to Wikidata and I will share a resource about FAIR Jupyter notebooks.
❓ Language is a huge part of our lives, thus naturally there is lots of research around it, and data. For example, recorded language in form of audio or written language in the form of text documents. In this week’s FAIR Buzzword Bingo series I will introduce the Common Language Resources and Technology Infrastructures (CLARIN). This research infrastructure provides data, tools and services related to language material.
🌐 I guess I don’t have to introduce Wikidata to you. It was already featured in this newsletter a few times. Just a few days ago I made my 1000th Wikidata edit. A reason to celebrate! I will briefly mention what my 1000th edit was and how you can contribute to Wikidata yourself. This will also mark the start of the new category Wikidata Spotlight where from time to time I will share what I have edited and why.
🖥️ Whether you are a programmer or not, chances are high that you have heard the term Jupyter Notebooks at least once before. If not also no problem, I will briefly explain it and share a recent resource that explains how you can make your Jupyter notebooks FAIR. Thanks for this tip to my colleague Sally Chambers.
❓ FAIR Buzzword Bingo: CLARIN
There are already countless intertwined research and infrastructure initiatives in the cultural heritage field, even more beyond it. Every week I will focus on one of the many FAIR data-related initiatives, this week: the Common Language Resources and Technology Infrastructure (CLARIN)
Imagine that you are researching something that relates to languages and text processing. You may need voice recordings in a specific language from a specific time or digitized documents such as newspapers or legal texts. Where do you get data and tools to do this? And more importantly, how do search for it in in a uniform way?
CLARIN (Q2986825) provides a research infrastructure offering data, tools and services to support research based on language resources. Via their platform you can find relevant data at one of the many participating institutions. One example is the Virtual Language Observatory where you can search in nearly a million metadata records. For example, when searching “Flemish” you can find The Dutch Song Database, a Flemish parliamentary debate or a video collection about Teaching Dutch in primary and secondary education.
Background
CLARIN, or more precisely CLARIN-ERIC is one of the European Research Infrastructure Consortiums I talked about in the last edition. It was built in 2012 with nine founding members after the initial CLARIN project (Q120669114) between 2008 and 2011. In the meantime there are member consortia from 22 European states, observer consortia from South Africa, Switzerland and the UK, as well as a third-party consortium from the United States.
Within the CLARIN-Plus (Q100319394) project between 2015 and 2017 was dedicated to enhance CLARIN. The initiative became even more mature.
“The project aim was to consolidate and strengthen the CLARIN infrastructure in the following areas: the central technical hub and central office, partnerships with other infrastructures, outreach and governance” - CLARIN-Plus outcomes
The partnership with other infrastructures relates for example to Europeana (that I covered in a previous edition) and DARIAH, an initiative that I will cover in a future edition.
Depending on your needs you can find relevant CLARIN member institutions for your use case. CLARIN member institutions are organized in three types of centres: Technical Centres (B) , Metadata providing Centres (C) and Knowledge Centres (K). Knowing this helps you to estimate what kind of data and services you can expect from a specific member institution.
CLARIN and FAIR data
The most important question for this newsletter: how about FAIR data? The CLARIN network is the combined effort of many different institutions. They agreed on certain standards and make searching for data easy. In the book The Enterprise Data Catalog, Ole Olesen-Bagneux specifies two types of data search: searching for data and searching in data. The former is important to identify relevant data in the first place. And among others, this is where the CLARIN infrastructure provides added value: by providing standardized metadata catalogues!
Additionally, according to the CLARIN information video (that you will find below), they explicitly contribute to easy access by making all resources FAIR, in their case:
Findable through the CLARIN discovery boards
Accessible through single-sign on
Interoperable thanks to CLARIN’s promotion of open science standards
Reusable because users know what they can and cannot do with any dataset
What do you think about CLARIN? Is there something I have forgotten? Below you will find a short YouTube video that introduces CLARIN high level. If you are interested in more details here are some links for further reading
The Tour De CLARIN series that highlights different centers: https://www.clarin.eu/Tour-de-CLARIN
An open bibliography related to CLARIN: https://www.zotero.org/groups/562080/clarin/collections/DNV2LZAN
🌐 Wikidata Spotlight
Anyone can edit Wikidata! You can do it either manually via their website or automatically for data in batch via tools such as QuickStatements. Just a few days ago I started to use QuickStatements for the first time. This makes my tiny anniversary short before even more astonishing for me: I already have performed 1000th edits on Wikidata, manually!
How can you edit Wikidata? Just visit a Wikidata item in the browser and click one of the edit buttons to add statements or change the description. It is that easy! Actually you don’t even have to login. If you do not have an account, your IP address will be shown in the editing log. And yes, there is an editing log! On every item you can click View history to see who did when contribute what to the item. Only for certain protected pages or when you want to use QuickStatements you need an account and at least 50 edits.
You may wonder what my 1000th edit was. While curating information about CLARIN (see section above), I had to indicate somewhere specific Technology Readiness Levels (TRLs). In case you haven’t heard of it before, there are nine such TRLs. They were defined by NASA in the 1970s to estimate the maturity of technologies. Since a decade or so, they are also used in funding programmes of the European Union.
Wikidata only knew the single concept TRL that represents the scale as such (Q1478071). I created one concept for each of the levels such that they can be used independently to annotate data. And TRL-6 was my 1000th edit: Q120669350. I still have to investigate which property should be used best for the TRL levels. So far I have used compliesWith (P5009). What do you think? How would you use these TRL levels on Wikidata?
🖥️ Coding corner
What do you do if you have (FAIR) data? I guess you want to use it to reach some sort of a goal. This means you have to process the data, aggregate it or visualize it. These kinds of workflows and the specific steps you perform (and why you perform them) are crucial. Ideally these workflows should be FAIR too!
Especially for data science, the context and the why of data processing are important. You can use Jupyter notebooks as an interactive environment in which you can mix a documentation of your steps and pieces of code.
“The Jupyter Notebook is an open-source, browser-based tool for creating virtual lab notebooks that document research workflows, code, data and visualisations. It is ideal for interactive data science and scientific computing across disciplines, supporting programming languages including Python, R, Julia and a few others” - DOI: 10.1109/MSR.2019.00077
Just a few days ago, Australian Research Data Commons (ARDC) published a practical guide on how to make Jupyter notebooks FAIR. One of their tips is to use code versioning systems such as git and to obtain DOI identifiers for code, for example via an existing GitHub feature. You can check out their complete guide for more information.
That’s it for this week of the FAIR Data Digest. I hope you found the content interesting. Don’t forget to share or subscribe. See you next week!
Sven