COST Action enetCollect - CA16105

Minutes of the first Core Group meeting

June 21-22, 2017, Ljubljana, Slovenia

Faculty of Arts, University of Ljubljana, Room 116

Frieda Steurs (WG3), Karën Fort (WG5), Branislav Bedi (WG3), Manny Rayner (Dissemination), Verena Lyding (GH), Lionel Nicolas (Chair), Matthias Kunkel (WG4),  Yaakov HaCohen-Kerner (WG2), Corina Forascu (Outreach), Gabriela Haja (Exploitation), Simon Krek (WG2)

Rodrigo Agerri had to leave early for health reasons, Iztok Kosem was absent due to illness.
Magali Paquot (Exploitation) was joining at specific points trough video conferencing.

WGs Leads

[WG1] R&I on Explicit Crowdsourcing for Language Learning material production 
- Dr Elena VOLODINA, vice Dr Iztok KOSEM -

[WG2] R&I on Implicit Crowdsourcing for Language Learning material production
- Dr Rodrigo AGERRI, vice Dr Simon KREK/Dr Yaakov HaCohen-Kerner -

[WG3] User-oriented design strategies for a competitive solution 
- Mr Branislav BÉDI, vice Dr Frieda STEURS - 

[WG4] Technology-oriented specifications for a flexible and robust solution 
- Mr Toma TASOVAC, vice Mr Matthias KUNKEL -

[WG5] Application-oriented specifications for an ethical, legal and profitable solution 
- Dr Karën FORT, vice Katerina ZDRAVKOVA-

[ 21 June 2017 ]

[ Morning session ]

(1) Introduction

Verena discussed several general points regarding the meeting and enetCollect itself.

Agenda for the meeting

Day 1
Morning:       Current state and strategies
Afternoon:    WG 1-5

Day 2
Morning:         Dissemination, Exploitation and Outreach
Afternoon I:         General matters and next steps
(Afternoon II:        Brainstorming on Marie Curie ITN)

Chair, vice-Chair, Grant Holder and COST Science Officer

Statistics on Members and subscriptions

33 countries now +2 coming
56 MC members
140 persons subscribed to mailing list - and probably interested in enetCollect

Information on Members was collected via a questionnaire, including WG of primary interest.
Subscriptions to multiple WGs are possible.
Current WG subscriptions are currently the strongest on WG1+2+3: 

Current subscriptions to the Outreach, Exploitation and Dissemination coordinations:    

Logic behind COST Action

The approach of COST Actions and the objectives of enetCollect were recalled:

Formal objective of enetCollect 

The formal objectives were reminded

3 Research coordination objectives:

4 Capacity building objectives:

Each objective has evaluation criteria (see proposal).

GP1 budget

The budget for GP1 (first grand period), ending in April 2018, was recalled:

The money planned should be used. Bad planning is not appreciated and can lead to reduction of the following GP budget. Budget shifts to the next GP are not possible.
Shifts between budget categories are possible, e.g. the 2nd meeting (Feb/March 2018) and more STSMs can be used for unspent budget.

Dissemination grants: 2 conferences (anybody presenting) + 3 ITC conferences (ECIs)

STSM : min duration is 5 days
Overall STSM budget was planned to accomodate 
1 planning STSM between WG lead and vice lead
2 ordinary for WG 1, 2, 3
1 for WG 4 and 5
1 for Website
1 for the ITN proposal

Timeline of GP 1

[March] Kick-Off with all MC members
[May] GP 1 started
[June] CG meeting
[September] First action meeting  (Bolzano)
[February/March 2018] Second meeting (probably Iasi)
[April 30, 2018] GP1 ends

Scientific and financial reports should be prepare at the end of each GP.

Activities up to now:

(2) Strategies for the active participation of members.

Lionel presented slides regarding means to motivate members to undertake tasks.
He made a distinction between “templated” and “non-templated” tasks:

Templated tasks 
Templated tasks require a limited amount of work (e.g. 1 week of work), and can be delegated to multiple persons.

MC members
MC members are expected to commit one or two week worth of work to the action and are thus the main available workforce.
Templated tasks are ideal for such amount of workforce availability per members.

Regular members 
Also a source of workforce to consider. Unlike MC members, regular members are not defacto invited to any events.
Invitation to the meetings is usually a good incentive for people to actively contribute. We should distribute "call for interest" via email or a "call for papers", which may be more purposeful.

Collective workforce
Collegiate mechanism to encourage collaboration. 
"Help enetCollect and enetCollect will help you".
Setting it into motion might take a bit of time.

Training schools + hackathons 
Especially relevant to templated tasks.
However, it will only be relevant from the second year onwards.

Shared publications
It is important that all members are credited for their efforts. Small templated tasks alone are usually not enough to make a publication.
However, they usually are defined in bundles that together can constitute the critical mass for a publication.
EnetCollect should establish a publication workflow asap.

Co-organized events
Lionel recorded that our COST Science Officier strongly discourages to organize workshops coordinated with major events for the first part of the Action.
Reasoning behind this is that the Action should first build its own identity to avoid that people participate in the Action to just get to go to the collocated events. 

The pros and cons of organizing enetCollect-related workshops at conferences right from the start has been discussed (risk of diluting the identity of enetCollect vs. bringing interested people together and exchanging information on ongoing).

At the beginning of the Action we should look for means to bring expertise to our WGs (e.g. by inviting specialists).
Define clear strategies for exchanging information between WGs, especially as enetCollect is highly interdisciplinary and open-ended.

Non-templated tasks: 
As opposed to templated tasks, non-templated tasks require a specific attention and time, e.g. leading a WG, and finding active and proficient members.

Strategies for finding prodicient members
Look for members with clear synergies on the task-level with enetCollect, e.g. 
- Make surveys that allow to appreciate the members’ motivation.
- Motivate members’ participation with special (COST) tools and strategies. (e.g. STSMs)
Usual COST tools for creating outputs and deliverables.
Give the possibility to organize visits between Action members.

It isn't a must that the host expert will be a formal member of our COST Action.

ITC conference grants for ECIs
Proefficient members from ITC countries can be funded to go to conferences (travel, subsistences and conference fees) and present enetCollect-related work.
This can constitute a good incentive.

Proposal for an ITN training network;
An ITN aims at rising a new generation of researchers on a given subject by funding an interconnected network of PhD students.

Lionel explained that a COST Action and an ITN are natural allies.
Indeed, a COST Action can greatly benefit from an ITN whereas an ITN can greatly benefit from a COST Action. 

A first attempt is considered for this year. 
Gabriela Haja is leading the proposal writing.

Enforce Marie Curie IF (MCIF) with enetCollect host institutions
Eurac Research organizes a one week training school for MCIF to prepare an Marie Curie proposal.
Lionel and Verena are considering organizing a similar event for enetCollect.

This funding scheme is ideal since it allows to devise multiple proposals in parallel and both the proposal writing and the implementation of accepted projects can help enetCollect's objectives.
Indeed the proposals will detail research ideas related to enetCollect and will thus serve the purpose of the brainstorming phase (research on approaches and requirments, starting month 6). 
The implementation of accepted projects proposals will constitute a great amount of prototypical data which serves the purpose of the implementation phases.

At the same time, the added value for the proposers are numerous. 
Among others the possibility to team up with very compatible hosts (plenty of choice) and greater chances of success due to the proposal writing training (see slides).

2nd Action meeting can serve to pair up interested candidates and interested enetCollect host institutions.
Core Group's wishlist
Lionel was asking CG members about their vision for enetCollect and what, if enetCollect could achieve something in particular, what could it be for each member of the CG.

EnetCollect to build complete NLP resources, as NLP tools performance is heavily hindered by incomplete NLP resources.
Lionel is especially interested in means to build lexica.  

EnetCollect to achieve "ethics by design" - building ethics right from the start when defining the specifications and identify key points and key issues and deal with them right from the beginning - we produce platform and that thy are ethically sound, ethics is a rock on which we are building.
e.g. users are logging in, it´s a key issue, do we need their name, and other personal information? We have to be sure from the specifications that we won´t need and use their name, also regarding the information of their location whether we deed it or not, and wether we are going to use it or not.

EnetCollect as hub to link people together, reach various audiences, including producers as well as consumers.

1. a big European proposal 
2. something on a local level for teachers and learners at schools - what they do now, what tools they use; finding out how to put a solar panel in a situation that is already there, if information is wasted, how can we use the information, how can we give feedback to the same teacher and learner to reuse the information - capacity to exploit the potential in online language learning.

An overview of different countries (technology, methodology, etc.) used - mapping a situation; making available all kinds of digital learning materials; make available as many corpora and language materials for teachers because up until now there has been a big divide between the traditional teachers and e-teachers; to build something larger on a EU network with interesting language materials that everyone can use - enhance achieve best practice over different countries, enhancing the production of LL materials, make sure that there is some kind of a portal on a web where this material is available.

Other attendees have not yet expressed themselves on this subject.

[ Afternoon session ]

(3) Report on WG1 activities and planning

WG1 is led by Elena Volodina (Leader) and Iztok Kosem (vice-Leader). 
Elena prepared a presentation.

Elena provided background information about herself and Iztok (see slides). She recalled WG1 tasks and then focused on the first GP tasks, i.e. on the state-of-the-art review.  To tackle such a task, they intend to:
- have an STSM in Ljubljana during fall 2017 (potentially in sync /combination with WG2),
- have groups of WG1 members setting and updating a “reading list” and “relevant events”,
- invite to September meeting speakers that are representative of different stakeholders groups,
- perform a survey during September meeting,
- make a call for paper between September meeting and next meeting on the subject of “Case studies/workflows using explicit crowdsourcing” 

A first draft of how WG1 intends to tackle the second and the third year objectives was then provided (see slides) .

WG1 currently has around 36 members with 18 having marked WG1 as their WG of primary interest. 

The presentation then switched on Action-related questions: 
    when will the invitations for September meeting be sent out ?,
    whom to invite and on what grounds should we decide who to reimburse ? 

There has also been discussion on the need to define the difference between Implicit and Explicit crowdsourcing.
All the discussions on these four questions are reported in the Varia section of these minutes (see section 12 below).

(4) Report on WG2 activities and planning

WG2 is led by Rodrigo Agerri (Leader), Simon Krek (vice-Leader) and Yaakov HaCohen-Kerner (vice-Leader). 
Rodrigo wasn’t able to attend the CG meeting. Both Simon and Yaakov prepared presentations.

Simon presented himself, his background in lexicography  and his experience with COST Actions (see slides). 
He then introduced a large Slovene project related to L1 teaching of Slovene for pupils with strong dialect background and the lexicography project proposal that came out of enel network (ELEXIS - European Lexicographic Infrastructure). 
Finally, he presented his institution the Jozef Stefan Institute which itself has several initiatives on Open Technologies, Open Educational Resources and Open Learning.

Yaakov presented himself and his research focus on text classification and clustering. He suggested to use crowdsourcing to annotate single-labeled and multi-labeled datasets.
Practical ideas related to implicit crowdsourcing include the use of games, quizzes and competitions to enable the annotation of single and multi-labeled datasets and a fashion to the GWAP approach (game-with-a-purpose).
For example, we could ask learners to label news items or images which itself is very alike the ESP game of Luis von Ahn. 
There are many possibilities and approaches for Crowdsourcing out there by now. The creating of state-of-the-art reviews will also serve to select approaches of interest and discard others.

(5) Report on WG3 activities and planning

WG3 is led by Branislav Bedi (Leader) and Frieda Steurs (vice-Leader). 
Both of them had prepared presentations.

Branislav started by presenting himself, his institution, his background as language teacher, his interest in human-agent interaction, embodied conversational agents and multimodality (see slides).
He also presented pictures of a task-based game called “virtual Rejkjavik”.

He then changed the focus to the actual composition of WG3 with 30 participants, 18 indicating WG3 as WG of primary interest. 
The participants can be grouped into 4 areas of interest: pedagogy, language, technology & design, research.

Branislav reminded the work plan for WG3 and described how they intend to tackle this GP objective, i.e. how to address the task of creating the state-of-the-art overview regarding online language learning platforms. 
Part of it will be addressed by asking WG3 members to canvas and detail the solutions existing in their own country or for their own language. A person (yet to be determined) will then combine the different inputs and create the first version of the overview. 

It has been noted that since there can be a great number of existing solutions there will probably be missing countries or languages in the first version but that they can be included in a subsequent version over the years.   

In her own presentation, Frieda stressed out the importance of the user-orientation. In order to build the network and create the state-of-the-art review, it is important to identify reference persons for each user profile/country (including stakeholders from companies) and ask them for references within their countries. With that objective in mind, a dedicated questionnaire could be created.

EnetCollect should naturally fit with the CALL community since new trends and methods related to corpus linguistics, data mining, computational modelling or even artificial intelligence are nowadays very present in CALL-related events (e.g. seee EUROCALL 2017 themes).
Frieda mentioned that the Institute voor de Nederlandse taal:, is also to serve as a collection and consolidation point for teaching materials.

Last but not least, she pointed out that we should also consider users with restricted abilities (blind, deaf, etc.) and understand what language learning platforms there are for them and how they use them.
Pointer to new technology: platform from TECOL (KU Leuven)

(6) Report on WG4 activities and planning

WG4 is led by Toma Tasovac (Leader) and Matthias Kunkel (vice-Leader). 
Toma wasn’t available for the CG meeting. Matthias prepared a presentation.

Matthias started by presenting some background information about Toma and him (see slides).

He reminded that WG4 has been devised as a support for WG1+2+3. Therefore it needs input from the other WGs. If not, then the WG4 will have a boring project. 
He highlighted the need to establish proper communication channels with the other WGs and recommended that each WG should have a designated contact person for WG4. 
This person should help with the communication in both-ways: from WGX to WG4 and from WG4 to WGX and thus understand the needs of both WGs for which he/she plays this “interface” role.

Matthias reminded that there is no budget for software developments: enetCollect is a network project and not a R&I project. 
In that perspective, enetCollect in meant to create ideas, knowledge and contacts/consortia and not concrete outputs. Likewise, WG4 is meant to create blue-prints to be later implemented in funded projects and not within WG4. 

The other WGs should provide their input with the user perspective in mind and WG4 will study it from a software development perspective (feasibility, cost, existing technologies etc.). 
There is a need to gather the input from the other WGs in a structured fashion such as a templated wiki or similar (no random emails). WG4 will create a solution for the meeting in September and present it to the other WGs.

(7) Report on WG5 activities and planning

WG5 is led by Karën Fort (Leader) and Katerina Zdravkova (vice-Leader). 
Katerina wasn’t able to attend the CG meeting. Karen prepared a presentation.

Karen started by presenting the current members of WG5 and the provided background information regarding herself and Katerina. 
She presented her profile, experience and research interests, especially with regards to Game-With-A-Purpose (GWAP), crowdsourcing and ethics. 
She then presented Katerina’s profile, experience and research interests (see slides). 

Karen highlighted the need to go beyond simple and general guidelines and to target “ethics by design”-> include ethical considerations from the start. She reminded that enetCollect's target groups in terms of crowd could include fragile groups (e.g. migrants, children, etc.) and that, in order to retain the crowd, we must make sure that the crowd has the means to trust us.

As for WG5, Karen mentioned the need to to establish proper communication channels with the other WGs and recommended that each WG should have a designated contact person for WG5.
This person should help with the communication in both-ways: from WGX to WG5 and from WG5 to WGX and thus understand the needs of both WGs for which he/she plays this “interface” role. 
There is also a need to foster participation of other WGs-members to WG5 even as passive members in order to share and receive information on the topic of ethics etc.
WG5 priority is to raise awareness on ethical and legal issues, provide guidelines, involve external experts to evaluate guidelines, etc. It is foreseen that the business perspective will come later. 

Karen mentioned the probable need of an STSM between WG4 and WG5 . Frieda mentioned that, in order to be truly inclusive, we should consider in enetCollect, if the web content accessibility guidelines (REF?) are a technical or a ethical concern. For that purpose, Matthias recommended to include direct user evaluations as guidelines might fall short on such aspect. 

Simon reminded that enetCollect is a network project whose main aim is not to make concrete implementation but at building up the ideas, knowledge and contacts/consortia for enabling concrete implementation at a later point.

[ 22 June 2017 ]

[ Morning session ]

(8) Introduction

Verena explained that since the strong authentication feature has not yet been activated on EURAC side by the time of the CG meeting, expenses have to be claimed through a paper document. 
CG members will thus fill the form on the e-cost platform, print the automatically generated document, sign it manually, send a scanned version of the signed document by mail and finally send the signed document via regular mail.
Verena and Lionel will keep CG members updated regarding the fine-grained details (e.g mail addresss) via an e-mail the week after the meeting.

(9) Report on the activities of the Dissemination Coordination

Manny acts as the Dissemination coordinator.
The report focused on the subject of discussions that happened during the first online meeting of the dissemination coordination, hereafter referred to as DC, that took place the 19th of May. 
The minutes of the meeting are accessible at the following address.

Manny pointed out that COST Actions are highly bottom-up activities, which aim at creating communities.
In order to exploit the community to create results two solutions are suggested:
    (1) getting funding (e.g. Switzerland offers funding for COST-related research)
    (2) establish collaborations with the aim to produce shared publications
Regarding collaborations, several ideas were presented ("evaluate my system", "annotate data",  "port my app to a new language", etc.) and example cases were shown (see slides).

WGs representation
It has been mentioned that it would be good to have for each WG one identified member of the DC acting as a contact point.  

It has been pointed out that there is an urgent need to collect information in one place asap and avoid relying too much on emails which is too time consuming and error prone.  
This is especially true currently regarding the question about how to join.

For the website, a two step approach is being implemented:
- Rina Zviel-Girshin offered to make a student contest to create a first (static) version of the website, deadline is the 25th of June. 4/5 teams have candidated.
- Winners of the contest will then be invited to the meeting in Bolzano and to an STSM aiming at adapting the first static version on top of a dynamic CMS which would provide more tools for the workflow of the action. For example, allowing content to be edited by different members (after properly affecting who does/changes what) 

The dynamic CMS is likely to be ILIAS, a LMS which development and maintenance is overseen by WG4 vice-leader Matthias. Matthias pointed out that there might be some difficulty when adapting the static version on the front-end of ILIAS. 
Matthias already agreed to set up a temporary ILIAS installation (that will later be moved to EURAC) to start gathering information. He also recommended to reserve the domain names. Lionel will reserve the domain names in the next days.

Internal Dissemination – Collaboration fostering
For collaboration fostering, the need of a website based on some CMS has been once again pointed out. The website should be a point of reference and collection of information of other MC´s works, ideas, projects, suggestions etc.  Internal exchange and detection of collaboration possibilities mainly works by means of information on the webpage and presentations at annual Action meetings. When registering to such a website, partners should have the possibility to define or use predefined labels described their interest in enetCollect. A mean to describe the partners they are looking for. 
Suggestion by Karen: apply gamification to foster participation (gathering points, etc.).

Social media
During the online meeting, the choice of possible “regular” social media has been reduced to Twitter and youtube channels. One DC member is interested in participating to the Twitter whereas another one mentioned her interest for the youtube channel.  
Regarding scientific social media, Researchgate, Academia, Google scholar and Mendeley have been mentioned. It appeared clear that they can be helpful to increase visibility of enetCollect-related publication, but that they probably should not be enforced. In general a distinction between 'enetCollect-funded' publications, i.e. presentations and articles that are presented at conferences by means of Dissemination meetings and ITC conference grants, and publication related to the enetCollect thematic has to be made. 'enetCollect-funded' publications need to be published on the enetCollect webpage with a clear acknowledgment of the Action and COST, while related publications can benefit from an optional label, indicating their relatedness to enetCollect. In order to approach the publication procedures in a coordinated fashion, there would be a need to create guidelines, e.g. for assigning the proper labels.  

It has discussed the possibility of fostering the publications of project paper in national conferences by all members. For example, with regards to NLP, the spanish partners could target SEPLN, the French ones TALN etc.
Publications in International conference would however be authored by CG members. The first likely one is LREC, which deadline is the 25th of September. Lionel will prepare something in the coming weeks and keep the CG informed.   
CALL conferences should also be targeted both at international and national levels.

Shared Calendar
It has been pointed out that there is a need for creating a shared calendar to highlight relevant upcoming event. Lionel will take care of that.

(10) Report on the activities of the Exploitation Coordination

Magali acts as the Exploitation coordinator. She presented her report remotely.
The report focused on the subject of discussions that happened during the first online meeting of the exploitation coordination, hereafter referred to as EC, that took place the 8th of June. 
The minutes of the meeting are accessible at the following address.

WGs representation
As for the other coordinations, it has been mentioned that it would be good to have for each WG one identified member of the EC acting as a contact point.  

Overlap with WG5
During the discussion, it became clear that the intiatives regarding the better ways to exploit the data produced by enetCollect overlapped with WG5 objectives. 
The EC will thus not pursue on this subject for the time being.

Marie Curie Fellowships training school 
As Eurac Research has started a program for training MC applicants in the proposal writing, we were considering to organize a similar training school in the context of enetCollect. 
The advantages of such an initiative are described in L&V’s presentation on day 1 and CG reaction to it reported in the corresponding part of the minutes.

Proposal sharing
Given that the enetCollect partners have access to different funding programs at their local, regional and national levels, we see it useful to encourage the sharing of project proposals, their specific and generic parts, as well as fostering the cooperation on proposal writing. 
In the end, such proposal are unlikely to compete with one another.

Since this approach is fairly unusual, the different way to make it happen has been discussed. For the time being, it became clear that a low-tech approach coordinated by a single person is preferable. 
Magali is up for the role. If there is a significant uptake by members, more elaborated workflow will then be considered. 

Collective workforce
Within the enetCollect proposal we have introduced the idea of a ‘collective workforce’ among Action partners, which is meant to formalize ways in which Action participants can receive help from each other on small tasks (e.g. reviewing a paper) and in return helping others, without necessarily have to ‘pay back’ to the same person. The logic behind is similar to the concept of ‘Time bank’ idea.

 It has been expressed that such an initiative might not bring a decisive added value. Therefore, if implemented in practice, a very fluid and easy to use workflow should be established. 

Extra: COST-related funding
During the dissemination slot, Manny indicated that they submitted a project proposal to a Swiss COST-related funding scheme. Lionel also pointed out that Bulgarian partner also have the possibility to obtain additional funding through a national COST-related scheme and encouraged CG members to double-check with their CNC if such national funding isn’t available. He also recommended to see at institution level, especially for partners working in large institutions.

(11) Report on the activities of the Outreach Coordination

Corina acts as the Outreach coordinator. 
The report focused on the subject of discussions that happened during the first online meeting of the dissemination coordination, hereafter referred to as OC, that took place the 9th of June. 
The minutes of the meeting are accessible at the following address.

WGs representation
As for the other coordinations, it has been mentioned that it would be good to have for each WG one identified member of the OC acting as a contact point.  

Erasmus+, key action 2
Corina introduced the funding scheme Erasmus+, with a focus on the Key Action 2, Strategic Partnerships for higher education. 
This funding scheme is especially relevant for enetCollect as it provides funding for mid-size projects (usually ~300 000 €, depending on the national agency it is submitted to) an can involve several partners across Europe and even non European countries. 

Another positive aspect, is that proposals for such a funding requires a limited amount of content (~10 pages), despite having several tenths of pages for administrative forms. 
Compiling such a proposal could thus be of interest for several partners and a “spamming” approach could be devised to submit several proposals every year. 
It has been mentioned that the right time to plan proposals will probably be after month 18th when the “brainstorming” phase will start.

OC purposes
Under outreach we understand specific activities that aim at expanding the community within and around enetCollect.

We can distinguish between Action-internal expansion (increasing the number of its members) and Action-external expansion (reaching out to individuals and communities that can be interested in making use of enetCollects outputs).
The boundaries between Action-external outreach and dissemination / exploitation activities is not clear cut.

At the beginning of the project, outreach will mainly concern Action-internal outreach. 
Starting the end of year two to three results should start to become available which allow to also focus on Action-external outreach.

The OC will help enetCollect to follow COST policies with regards to gender, senority and nationalities. 
The WGs are still building up and WG leads have to gain an overview of the available and missing competences within their WG. The outreach group intend to provide the ‘service’ to facilitate the outreach activities.

[ Afternoon session ]

(5) Varia

September meeting (schedule, WG meeting overlap, etc.)
It has been mentioned the need to have a mean allowing members to present their institutions and their interests and to get to know one another in general. 
Two solutions have been proposed: a poster session and a kind of “speed dating”.  The poster session seemed an actionable option. The “speed dating” option is interesting but would require a proper planification. It is not yet sure if it will be used for the first meeting.

Thanks to the different discussions of the CG meeting, it also became clear that, at least for the first annual meeting, it will not be adequate for WG meetings to overlap. 
Indeed, there is a strong indication that members will participate to more than one WG in average, and for this first meeting, members still don’t have a clear understanding of what the different WGs are about. They should thus have the possibility to attend each WG meeting.
This presents a challenge since we only have two days with at least the first morning being busy with the MC meeting and the overall Action introduction.  Different possibilities accommodating MC meeting, Action meeting, WG Meetings, CG meeting and a poster session have been discussed. L&V will make a draft and submit towards mid-July.

Since this overlapping issue might persist over time, it has pointed out that annual meetings could favor organizational matters while independent WG-Meetings could have more space for research matters.
In that perspective, the first meeting in September will most likely aim at kick-starting the participation of the MC members and distributing tasks while the second one in Feb/March could be longer and thus provide  more space for purely research-related matters. 

Invitation to the September meeting
The meeting in September will mostly be restricted to the MC while the one in Feb/March will include a greater number of regular members. 
The number of invitations for regular members is still to be establish and depends on the number of MC members that will not attend. 
L&V will send invitations to the meeting towards  the beginning of July with a rather short deadline. 
A gross number of free invitations will thus be available towards end of July.

For this one meeting, a transparent policy for assigning the free invitations has not been defined. 
Future meeting should have such a transparent meeting. WG-leaders are invited to send L&V the list of persons they would like to invite with some motivation of how interesting it is for them to have them over. 
L&V will then balance WG needs and then assign the free invitations.

February/March meeting place.
Partners from Iasi in Romania offered to organize the meeting in February/March. 
No other CG members has so far made a different propositions. Since the place and dates for this meeting are expected to be voted during September meeting, a call for interest on that subject should be released beforehand. 
Probably in early August.  

Yaakov reminded that the meeting should happen early enough to have the time to perform the travel reimbursements and spend the left-over budget with other means if such reimbursement is not enough to wipe the remaining budget.
Elena mentioned that Sweden is not as expensive as people usually think and could thus be an option for future meeting.

Dissemination meeting + ITC conference grants
It has been foreseen in the budget the means to send people to conference to present enetCollect or enetCollect-related results. 
However, because of the start overload and since conference deadline are usually 6/8 months before the actual conference. 
It is unlikely that such a presentation will happen before the end of the first grant period. Therefore, such budget will probably be used for different purposes.

For the next grant period, it is already quite clear that LREC will be targeted.
Deadline is September 25th. Lionel will prepare a draft. 
The shared calendar mentioned earlier during the report on the dissemination coordination should help doing such a planning.

Suggestions for conferences:
The 3rd International Conference on the Globalization of Second Language Acquisition and Teacher Education
submission deadline; July, 28
dates: 30-31 Oct, Hawaii
19th International Conference on Language Teaching and Second Language Acquisition
submission deadline: June, 25
dates: 28-29 Aug, Paris, France
AAAI Human Computation and Crowdsourcing Conference 2017
submission deadline (demo/work-in-progress): June, 30 
dates: October 24-26, Canada
* a list of Crowdsourcing conferences:

STSM calls
We discussed the fact that STSM calls regarding to COST regulations are not related to single WGs but are done on the level of a whole action. 

WG1 has been carefully planified and an STSM call has already been prepared. Since for the first year, the need of WG1, WG2 and WG3 are fairly similar, it has been agreed to switfly adapt this call to accommodate other WGs needs. 
WG4 and WG5 start after 6 month and do not seem to have needs for this STSM call at present time. 

Nonetheless, if such needs occur before the STSM call is released, their needs will be included in the call. The call should be send out in early July.

Logo and visual identity competition
Corina prepared a call for interest for a competition aiming at creating logo and visual identity to be transmitted to all members. 
The call still requires some input from L&V and will be send out early enough to allow a MC vote regarding the winner in September meeting. 
The prize for the winner will probably be an invitation for the meeting in February/March.

Local Dissemination
EnetCollect proposal foresee in its dissemination part to push articles in local newspaper or communication means of the partners’ institution. 
Eurac has a local regularly-distributed  magazin called “Academia” which informs South Tyrolean regarding the research performed at Eurac. Sorbonne has also a newsletter and the french INRIA has also communication means we could try to target.

Difference between explicit and implicit crowdsourcing
During the report on WG1 activities, it has been disussed that the difference between explicit and implicit crowdsourcing is not clear.
Typical examples are Wikipedia article production and editing (explicit) and Recaptcha human input verification (implicit).
Many cases are in between. 

The question also is how to define a crowd, e.g  Does a small group of teachers count as a crowd?
Crowdsourcing by definition is an "open call" to everybody. Question: Do we want to consider everybody a potential teacher.
Due to this close connection and fuzzy boundaries between WG1 and WG2 they are expected to closely work together.