Input Privacy-preserving Techniques

Progress

The first version of the joint use case for international trade has been delivered.
The intersection has been determined and a synthetic dataset has been created.
The use case was presented in a joint session with the UN-Petlab team. In the discussion a number of important findings were made, such as the need for input control and clarifying why the use case is needed and in detail what the privacy concerns are. It has become clear that there is no governance on a IPPT service.

The first version of the use case PET-based remote analytics service has been delivered. The ML model using Random Forest on the dataset proposed by StatCan to predict the students' score (attribute G3) was trained.

The preparation of the questions for the public consultation is still ongoing. Due to illness, it was not possible to discuss this in the session with the UN pet lab.

Next Steps

Run a test with the Istat/CBS software for the International Trade use case.
Alignment with UN petlab or testing with UN-petlab is possible.
Elaborate findings with UN petlab discussion.

Re-training the ML model after categorizing the student score attribute.

Go through questions for public consultation.

Risks and Issues

Issue	Mitigation
It takes more than just technology and methodology if we want to actually take IPPT into production.	Discussion in the executive board. Involve BSTN in this challenge.

Image result for input privacy-preserving techniques

Meta Academy for the Modernization of Official Statistics

Progress

We have had three benchmarking presentations:

Openscapes from the US and their model for Open Source workflows and process in research
Statistics Canada's virtual data literacy for the federal public service
The Carpentries.org is a NGO that already has a framework for generating and delivering training across the world and includes a network of over 3000 trainings and curriculum on working with data and programming.

Next Steps

We will continue with benchmarking presentations.

Over the course of the summer we will be working with project members to fine tune the vision of the project and alignment of next steps.

Risks and Issues

Issue	Mitigation

Data Governance for Interoperability Framework

Progress

The team has had 5 meetings so far. We have agreed in a concept of interoperability from the point of view of the statistical offices.

Next Steps

During the next meeting which is planned for the 5^th of July, we will be discussing on the elements that must be part of the framework, as the drivers to define the sections of the document. The plan is to distribute this sections among the members of the group to start working in parallel in offline mode. During the meetings, we will be discussing and integrating those contents.

Risks and Issues

Issue	Mitigation

News from the Groups
Blue-skies Thinking	Identifying Topics/Opportunities	CONTINIOUS	The group is working on getting pitches to get new ideas for next year, e.g. on process automation

	Digital Twins	IN PROGRESS
	Statistical Cloud Use	IN PROGRESS
	Mobile survey data collection/rapid survey	IN PROGRESS
		TO IDENTIFY

	Other	BSTN is preparing two physical events: Serbia 29 June: Cloud for official statistics in Servia with 20 participants in person and another 20 online; there might be potential for a project on cloud use. Newport 12-13 June: three sessions: cloud, use of rapid surveys (app based or otherwise) digital twins
Applying Data Science and Modern Methods	Scoping and planning of work	IN PROGRESS	Members have discussed several areas of work and suggested three task team, namely (i) data editing; (ii) modelling; and (iii) respoonsible AI (see Google doc for more details). The group will conduct three sessions in Newport to discuss the proposal (see Google doc for the tentative agenda).
	Activity 2	NOT STARTED
	Activity 3	TO DEFINE
	Activity 4	TO DEFINE
		TO DEFINE
	Other	16 colleagues have signed up for the group (ABS x 4, Statistics Canada x 2, Statistics Portugal x 2, ONS x 2, INEGI, KSH, IMF x2, Bank of Italy, and Eurostat) (see the list of member)
Capabilities and Communication	Future of work toolkits	IN PROGRESS	We had great response rate to the questionnaire on the Future of Work. 41 responses are received up to date. Sub team should start analysing responses to their respective parts of the questionnaire. We had a very interesting CES session on 21 June 2022 on Future work and future workplace – post-Covid-19 working modalities. Results form panel discussion will be analysed in details in order to shape our future key tasks.
	The Job of the Future	IN PROGRESS	We had great response rate to the questionnaire on the Future of Work. 41 responses are received up to date. Sub team should start analysing responses to their respective parts of the questionnaire. We had a very interesting CES session on 21 June 2022 on Future work and future workplace – post-Covid-19 working modalities. Results form panel discussion will be analysed in details in order to shape our future key tasks.
	Reaching Young People	IN PROGRESS	We had great response rate to the questionnaire on the Future of Work. 41 responses are received up to date. Sub team should start analysing responses to their respective parts of the questionnaire. We had a very interesting CES session on 21 June 2022 on Future work and future workplace – post-Covid-19 working modalities. Results form panel discussion will be analysed in details in order to shape our future key tasks.
	Ethics Managment (Data and Business)	IN PROGRESS	Ethics survey was sent to the countries on 16 May, with the deadline for responses of 15 June. We have received 12 response to the survey and 8 more responses should arrive by 7 July. We will wait for the late responses to start analyzing survey. It was also discussed: maybe questionnaire could be used to turn responses into indicators. To build a list of ethics indicators that should be in place in the statistical office; to have guidance to set up ethics in the countries; to promote ethical literacy in the organisations; business case to include ethics in GAMSO. In terms of In-depth review on data ethics we had a couple of meetings. UK and Canada produced text that summarizes data ethics in their countries, also produced outline of the review, scope and definition, how it relates to ethics and business ethics. With country examples from Canada and UK, Switzerland may also contribute. Other countries were asked to submit their examples for the review, for example Albania. UK started to look at the literature review, and Canada is preparing background introduction. Next call will be on Wednesday 20 July at 13.00 Geneva time.
	Market Research, Digital Marketing & Communication strategies (Strategic Communication Framework follow-up)	IN PROGRESS	Final version of the output document is available on the wiki: Brand and reputation management Home. Team should identify what they will be interested to work on for the remainder of the year. Topics that were identified by this task team: How to Measure success and the impact of our communication Strategies to tackle and anticipate disinformation Next call will be on Thursday 7 July at 14.00 CET.
	HRMT Workshop 2022	IN PROGRESS	The HRMT 2022 workshop will be held in Brusels, Belgium, 11-13 October 2022. Information Notice 1 and invitation letter were prepared and sent to NSIs last week. Agenda for the Workshop is focusing on 4 pillars: 1) organisation, 2) employer, manager and leader, 3) employees and supplement with 4) mix/horizontal /blended/hybrid issues. We will view each pillar from four dimensions: 1) mindset, 2) environment, 3) behavior and 4) skills. There are the following deadlines for the workshop: 29 July Abstract or proposal for intended contribution 16 September Registration 16 September Paper (or executive summary), Presentation 7 October Final versions of presentations 11-13 October 2022 Workshop Next meeting of the OC is planned for Tuesday 19 July at 14.00 Geneva time.
	Other
Supporting Standards	GSIM Review	IN PROGRESS	Work ongoing as scheduled. For details, please check our Github group @ https://github.com/UNECE/GSIMRevision/ The current discussions are focusing on the referential metadata objects (huge part of the GSIM). Storyboards (“GSIM in Action”) have been prepared also for Exchange Group and Business Group to promote GSIM at the ModernStats World Workshop in Belgrade. A sprint will likely to be organised (not yet planned). Once the high-level view is in place, small well-defined issues will be addressed.
	Core Ontology for Official Statistics phase 2	IN PROGRESS	Work ongoing as planned. The discussion is organised according to the issues defined on the Github platform. For details, please check our Github group @ https://github.com/linked-statistics/COOS
	GSBPM Task	IN PROGRESS	Task Team has bascially completed its work. The output of the work will be presented at our ModernStats World Workshop in Belgrade, includig a group activity managed by the Task Team members. If valuable input is received from the workhop, the team might have 1 or 2 additional discussions. Apart from that, the finalisation of the outputs is the next step and the Task Team will conclude its work. The output includes a final proposa for the tasks produced by the Task Team, collection on country examples (collected at the beginning of the work) and some proposals for the GSBPM revision Task Team (planned to be launched next year). For details on main Task Team output, please access the Google doc @ https://docs.google.com/document/d/1hzG4uSOyOnxTMoj4ti5zEGtLl82e2eJhAN4_RZKknc0/edit
	SDMX-DDI-GSBPM	IN PROGRESS	The Task Team was launched at the beginnig of May but after a few meetings, we had a reassessment of what is the reasonable outcome of the work. Based on the results of the first meetings, more alternatives were open on the level of detail and the ways of producing the mappings between SDMX and DDI. After some reassessment and rescoping, the team agreed to first write down a few sentences on how SDMX and DDI can help users in practice, on the level of GSBPM sub-processes. The first exercise will focus on 3-4 sub-processes then the team will compare the outcome and aligh the future work accordingly.
	GSBPM overarching processes	NOT STARTED	Task Team is expected to start soon after the GSBPM Tasks Task Team finished its activities. The Supporting Standards Group will soon discuss the options on when and how to launch this activity.
	CSPA capacity building	NOT STARTED	There has been no progress with this Task Team since the last EB report. After the workshop in Belgrade, this Task Team will have to be reassessed. I still need to contact Emanuele Baldacci on some lessons on the CSPA catalogue. If there is still no appetite for the work to start and no chair, I will come back to the Executive Board with a proposal on the Task Team's future.
	ModernStats World Workshop 2022	IN PROGRESS	Our workshop takes place between 27-29 June in Belgrade. Thanks to the hard work of InKyung Choi, our Serbian colleagues and the Organising Committee of the workshop, all is set for the meeting. For more details on the workshop, please visit: ModernStats World Workshop 2022 We expect approx 50-55 people in person and approx. 10 more to connect online. The Supporting Standards Group will evaluate the lessons learned from the workshop at its next meeting on 7 July. I will also brief the Executive Board on the outcome of the workshop in the next report.
	Other	Highlights for the Executive Board: The current top priority is the ModernStats World Workshop that takes place between 27-29 June in Belgrade. The fate of the CSPA capacity building Task Team is still uncertain. I will brief the Executive Board on the proposed future of the group at the next reporting. To the best of my knowledge, the Supporting Standards Group has no champion.
Machine Learning 2022	Global Data Squad	IN PROGRESS	We have re-launched a research project looking at ML for AIS with the UN Task Team on AIS. We are recruiting ML group members to work together on identifying spatial clusters within shipping vessel locations data, representing relevant areas within and around shipping ports globally. We have had strong interest in taking part in the group and aim to start work next month.
	Workstreams	IN PROGRESS	The seven Theme Groups have been meeting regularly in May and June to exchange experience and knowledge. Colleagues in the model retraining and quality of training data have been working together on joint documents which bring together their insight on the topics. The Imagery group has been running peer review discussions of academic papers as well as a study group for those new to Earth Observation. Each group holds regular presentations of work from across the official statistics community. Recent presentations include ones from the IMF on its CTS coding tool (model retraining group) and matching big data sources (text classification); and ones from Statistics Canada on MLOps for the IT Infrastructure group; and triaging enquiries using multilingual transformers model for the text classification group.
	Hackathon(s)	IN PROGRESS	The Group has been busy preparing for the ML sprint at the international data science meeting, ONS, UK, July 12-14. Seventeen ML Group members will gather for activities in three different areas - model retraining, quality of training data and web scraping data. The activities will help advance the existing work of the three theme groups. There will also be an opportunity to promote the group's work to members of other international working groups who will be attending the wider meeting.
	Capacity Building	IN PROGRESS	We are developing plans for further Coffee and Coding sessions for the autumn, These would be in online workshop format, with opportunity for members to work on their own code.
	Communications	IN PROGRESS	We issued a newsletter in May containing the latest news and opportunities from the ML for official statistics space. We have also been promoting similar material on the members website and on the group's new discussion forum, Slack.
	Other	.