Re: Next meeting JAN-10 ; Season's Greetings

Hi All.
Welcome to the new year and hope it goes pleasantly and productively.

## Meetings

For kicking off DPVCG activities, we have scheduled meetings to continue 
in the usual time slot i.e. WED 15:00 WEST / 16:00 CEST - starting on 
JAN-10. If this time slot doesn't work for you - please reply to me with 
a proposed time slot and I will try to accommodate a schedule that works 
for as many people as possible.

## Agenda for the coming year

In the first meeting (JAN-10), we will be looking to set the agenda for 
the coming year. I encourage everyone to participate in this process by 
sharing their ideas, visions, criticisms. Through this, I hope we start 
on a strong footing and keep the momentum going throughout the year.

## DPV v2

In the last few meetings, we discussed having the next iteration of DPV 
be v2 instead of e.g. v1.1 to signal the significance of scope change 
(inclusion of non-personal data, legal basis, etc.) Moving ahead, I 
would like to aim to have a release every 6 months e.g. by expanding 
taxonomies to represent AI, Risk Management, and ISO Consent/Notice 
standards, further enrichment/refinement of taxonomies through published 
authoritative documents. As part of setting the agenda, we can also 
discuss setting potential dates for this work.

## DPV Documentation - new features

The work on the documentation generator is almost complete. The 
remaining parts require people to go through it and then we decide 
whether to move it to the main DPV repo. Below I list changes so people 
can go through them before the next meeting(s). The use of harshp.com is 
because this is the 'dev' version and the resources at w3id.org/dpv are 
intended to be used as stable/production.

### Visual changes

- Hierarchical views to see the taxonomy as an interactive list/tree 
e.g. https://harshp.com/dpv/dpv/#vocab-purposes
- Term description table e.g. https://harshp.com/dpv/dpv/#PrivacyNotice 
shows the term description, parent concepts with path to the top 
concept, properties/relations it can be used with, examples (if defined 
with the concept as subject), and a link to the section it is documented in
- So if someone enters the term IRI/URL directly - they are shown a more 
useful summary including other relevant terms and examples

### Examples

- separate pages for detailed guidelines and examples on specific 
concepts e.g. entities https://harshp.com/dpv/dpv/modules/entities (work 
in progress) - this is to avoid the main DPV spec page becoming even 
more gigantic
- examples are easy to declare and embed e.g. using just the ID (E0018) 
the example at https://harshp.com/dpv/examples/#E0018 is inserted into 
the HTML. If this example is updated, it gets updated everywhere it is 
used.
- All examples are also listed on the single page 
https://harshp.com/dpv/examples/
- Examples of examples - 
https://harshp.com/dpv-x/guides/consent-27560#purpose

### Multi-Lingual

- multi-lingual support: there is a drop-down just after the vocabulary 
details on the top of the page which takes you to e.g. German (DE) 
https://harshp.com/dpv/dpv/index-de.html
- On this page, the term labels used are in German if verified, German 
with suffix (machine translated) if unverified, and in English with 
suffix (translation missing) if unavailable.
- Multi-lingual labels are also defined in RDF with the same suffixes 
(except those missing).
- Thanks to Tobias / TRAPEZE for helping start this.

### Location and Jurisdictions

- location taxonomy https://harshp.com/dpv/loc/ with links to external 
taxonomies (ISO, EU Vocab - this one is nice because it has 
multi-lingual labels for locations already)
- legal taxonomies per jurisdiction e.g. EU 
https://harshp.com/dpv/legal/eu/ shows the available laws, authorities, 
and adequacy decisions and all legal terms are consolidated under a 
common page at https://harshp.com/dpv/legal/
- laws moved under jurisdiction e.g. DPV-GDPR is now EU-GDPR and located 
under /legal/eu/gdpr https://harshp.com/dpv/legal/eu/gdpr/

### Serialisations

- default serialisation is now RDFS+SKOS
- OWL serialisations are put in the same folder with suffix -owl, 
multiple formats (e.g. ttl, json-ld, omn).
- OWL RDF is generated by simply replacing the SKOS concepts using OWL 
concepts based on 
https://www.w3.org/2006/07/SWD/SKOS/skos-and-owl/master.html
- OWL assertions e.g. disjoint classes, functional property can be 
specifically and easily added by using SPARQL 
https://github.com/coolharsh55/dpv/blob/dev/code/200_serialise_RDF.py#L220
- Question of which IRI to use for OWL RDF is pending

### Spreadsheet management

- documentation generation code has been rewritten from scratch - so all 
the technical debt accumulated from building over small scripts at the 
start is now replaced with new bugs!
- The main benefit of doing this is the immense flexibility in producing 
RDF/HTML e.g. 
https://github.com/coolharsh55/dpv/blob/dev/code/vocab_schemas.py is the 
place where each value in a column in a spreadsheet is passed to the 
specified python function to figure out what triples to produce.
- This allows 'different' layouts such as the spreadsheet declaring 
legal basis x rights mappings to be easily converted into triples/RDF 
without affecting the rest of the code.

### Support for Reasoning/Inferencing

- support for ad-hoc triple insertion using 'hooks' e.g. using SPARQL 
queries to declare all direct/indirect concepts linked to 
dpv:SpecialCategoryPersonalData as its instances 
https://github.com/coolharsh55/dpv/blob/dev/code/vocab_management.py#L743 
- so in the spreadsheet we don't need to track which categories are 
special or not as this will be done automatically in the RDF and HTML. A 
non-trivial example where this can be immensely useful - suggesting 
suitable ISO standards for specific 'top' tech/org measures and then 
using this method to have all subcategories of those measures also have 
a link to the standard.
- this code can be trivially supplemented by other reasoners to produce 
DPV variations with inferred triples e.g. skos:broaderTransitive or 
rdf:type assertions
- The idea here is that the RDF data can be directly utilised without 
preprocessing e.g. import dpv.ttl (or separarely dpv-inferred.ttl) and 
you're ready to go

Regards,
Harsh

On 22/12/2023 10:53, Harshvardhan J. Pandit wrote:
> Hi.
> In light of the impending holiday season, we will be taking a pause in 
> DPVCG meetings, and reconvening on January 10th (15:00 UTC,WET / 16:00 
> CET). If you forsee not attending this time slot in the new year and 
> want alternatives, I will be sharing a poll in Jan first week.
> 
> We are ending the year on a high note - and I want to personally thank 
> all the participants and users of DPV/DPVCG who form the 'community' for 
> their time, thoughts, and contributions.
> 
> Last year around this time we reached a significant milestone in the 
> form of DPV v1, and have continued the momentum in building a promising 
> outcome based on addressing practicalities and innovations such as AI - 
> which is leading to a potential DPV v2 and several guidelines and 
> specifications. I look forward to continuing this effort in the new year 
> with a refreshed vigour and renewed motivation.
> 
> Wish everyone a pleasant and joyful holiday.
> Regards,

-- 
---
Harshvardhan J. Pandit, Ph.D
Assistant Professor
ADAPT Centre, Dublin City University
https://harshp.com/

Received on Wednesday, 3 January 2024 16:02:42 UTC