### [DC-2019](http://www.dublincore.org/conferences/2019/) Hack Day - Sep 26 2019 in Seoul

Hack Day chairs: Tom Baker and Andra Waagmeester

See https://github.com/dcmi/hackday

[This year's Hack Day](http://www.dublincore.org/conferences/2019/hackday/) focuses on [Application Profiles](http://dublincore.org/groups/application_profiles_ig/), [FAIR data principles](https://www.force11.org/group/fairgroup/fairprinciples), [Wikidata](http://www.wikidata.org), and [Wikibase](http://wikiba.se). See the [Hack Day page](http://www.dublincore.org/conferences/2019/hackday/) for further details.

You are invited to submit pitches related to these or any other metadata-related topics of interest by posting to a [submission form](http://bit.ly/DC2019_hackday_pitch_submission) or by directly editing [this page](https://hackmd.io/i7tC1gC6RXOFtzElnhlpZw?both#).

## Pitches

### 1. Modeling an historically complex periodical on Wikidata (or Wikibase)

Over the past 350 years, the French scientific journal [Comptes Rendus](https://en.m.wikipedia.org/wiki/Comptes_rendus_de_l%27Acad%C3%A9mie_des_Sciences) has undergone complex structural changes. Some have that if we could successfully model its changing structure in a custom Wikibase, this would prove that [Wikibase can be a backbone of a library catalog system](https://twitter.com/phoebe_ayers/status/1149100671586787328).

Pitch: Let's take the bait and try to model Comptes Rendus.

### 2. Render and author application profiles in ShEx through a GUI

Shape Expressions (ShEx) is a language for formally expressing application profiles. In May 2019, Wikidata released a schema extension, which allows collaborative editing of application profiles on the underlying Mediawiki platform. Yet ShEx has a steep learning curve that can seem like a insurmountable barrier to potential profile creators. Might that barrier be lowered by a graphical interface that both renders the profiles and allows the profiles to be edited by authors?

Pitch: Let's explore ways to build such tools, starting with existing tools such as Jupyter Notebooks, [Cytoscape](https://cytoscape.org/), and [yEd Graphical Editor](https://www.yworks.com/products/yed).

### 3. Create a Application profile editor in JupyterLab

Application profiles can be defined using the Shape Expression language. Currently doing so requires a cascade of different tools and resources. [JupyterLab](https://jupyter.org/) is a web-based interactive development environment for Jupyter notebooks, code, and data.

Pitch: Let's develop an example notebook for ShEx-based application profiles.

### 4. Use Wikidata Entity Schemas as a platform for storing, editing and discussing application profiles

The Wikidata project now supports the creation, discussion, and maintenance of Entity Schemas (aka application profiles) in ShEx. In this new context, application profiles are not just static documents but, like the underlying data, may continually evolve.

Pitch: Let's look at how Entity Schemas fit into the Wikidata ecosystem, both as version-controlled snapshots and as living documents.

### 5. Extract schemas (application profiles) in support of analysis in data science

[RDFShape](http://rdfshape.weso.es), a generic RDF validation tool, can extract (or "infer") a ShEx validation schema, or "application profile", for a given set of data and generate a visualization of that schema. Could RDFShape help data scientists leverage existing Wikidata entity schemas and infer new schemas in support of data analysis using popular languages such as R?

Pitch: Let's try to fit schema generation and visualization tools into workflows for data analysis using R.

### 6. Review the available toolchains to create, edit, render and consume Application Profiles.

Can we generate an exhaustive list of tools available applicable to Application profiles and their use cases.

### 7. Create application profiles for
#### a. Scientific names of organisms
Wikidata needs to be useful to the users of scientific names of organisms, such as taxonomists, ecologists, gardeners, museums, doctors, geneticists, conservationists etc. Currently the data model for taxonomy, nomenclature and specimen data on Wikidata conflating taxon concepts with taxon names and so Wikidata is not suitable to hold taxon information in a usable format. For example, Zamia calocoma (Q17348078) and Microcycas calocoma (Q144477) are different names for the same taxon, which gives the impression that Wikidata is modelling names. However, there is a picture of Microcycas calocoma (Q144477), so it must be modeling a taxon. If this were modelled correctly we could do many things that are not currently possible, just one example is that Wikidata could be used to test the rules of international nomenclature.

#### b. Projects and grants
DINGO, the Data Integration and Extension for Grant Ontology, is an ontology expressly designed to provide an extensible interoperable framework for formally conceptualizing and expressing the relevant parts of the research/cultural landscape in relation to funding, such that they can easily be shared between different actors and platforms. It is conceived to have sufficient richness of expression to satisfy complex requirements (see sections Aims of the model and Relation with other ontologies, extensions and integrations of the model, while at the same time being simple enough to be of immediate use also for the simplest use cases.

### 8. Agree on the most minimal possible model for application profiles

Can we devise a simple model for application profiles that captures 90% of the most obvious use cases (e.g., list properties used, with what constraints, for multi-entity profiles)?

Karen, Stefanie, Tom

Working on: http://asch.wiki.gwdg.de/index.php/ASCH_Model

### 9. Use a minimal model for application profiles in easy-to-use spreadsheet or YAML templates

Can we devise a simple [spreadsheet template](https://github.com/dcmi/dcap/blob/master/template_tom/dcap_template.xlsx) that would be straightforwardly convertible into ShEx (or other models and syntaxes)?

### 10. A simple tool for converting SHACL to documentation for dataset

Can we generate a SHACL-based documentation for human friendly page?