D6.8.- Report on project communication and engagement ......PROTEUS Deliverable D6.8 687691 Page 2 of 37 Abstract This document provides details about the communication and engagement

PROTEUS Scalable online machine learning for predictive analytics and real-time

interactive visualization 687691

D6.8 Report on project communication and engagement activities – v2

Lead Author: TRI With contributions from: DFKI, AMIII, BU, TREE

Deliverable nature: Report (R) Dissemination level: (Confidentiality)

Public (PU)

Contractual delivery date:

November 30 2018

Actual delivery date: November 30 2018 Version: 1.2 Total number of pages: 35 Keywords: Communication, clustering, engagement

PROTEUS Deliverable D6.8

687691 Page 2 of 37

Abstract This document provides details about the communication and engagement activities and materials produced during the PROTEUS project, as part of the communication activities and engagement with relevant related initiatives. It is important to highlight that this deliverable is focusing on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). This document provides, initially, information about the general dissemination of the project, including the corporate image design, communication materials, and established communication channels and tools in use. And then a summary on clustering and engagement activities carried out in the course of the project.

Deliverable D6.8 PROTEUS

687691 Page 3 of 37

Executive summary This document provides details about the communication and engagement activities and materials produced during the PROTEUS project, as part of the communication activities and engagement with relevant related initiatives. It is important to highlight that this deliverable is focusing on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). This document provides information about the general dissemination of the project, including the corporate image design, communication materials and established communication channels and tools used to build a stakeholder community. In terms of the corporate image, a logo was designed already at the proposal stage, and this has been kept during the project, guiding the overall corporate image for PROTEUS. On this basis, a number of templates were also designed to be used in communication activities; in particular these were implemented for both presentations and documents, assuring that specific rules related to communication (e.g., standard formats, stable brand identity, etc.) have been fulfilled. Following from this, a template for acknowledgements and disclaimers was also produced in order to assure that the H2020 dissemination guidelines in this regard have been followed. About the communication materials, PROTEUS is currently running three different channels targeting different purposes. Firstly, for the general public and for general communication purposes, the project’s public website is available online (www.proteus-bigdata.com) and includes research results as well as blog posts. Secondly, for a more interactive communication and presence on the social media, a Twitter account is being used (https://twitter.com/proteus_bigdata). And finally, since the PROTEUS results are being released as Open Source, a GitHub organisation is hosting the software results of the project in several repositories, constantly updated with the evolution of the project (PROTEUS GitHub at: https://github.com/PROTEUS-H2020).

Channel Main purpose Target audience

Website Share general information about the project

Members of the public, academics, industry, other projects

Social media Two-way communication with interested stakeholders

Members of the public, academics, students

GitHub Sharing of software results Programmers and big data experts from industry and academia, students

In terms of clustering and engagement activities, a number of actions have been taken. Several workshops and conference have been attended with the aim of engaging with related initiatives and projects exploring relevant areas. In particular, FiCloud, Big Data Spain, the Big Data Value Association or the EuroPro workshops were venues where PROTEUS was presented, interacted with other initiatives, and engaged with the relevant community and stakeholders. In addition, other activities included public talks and presentations given by the project’s partners during the project meetings or in conjunction with close organisations (e.g. other academic bodies). The interaction with related initiatives has also been fostered during this period, especially with sister projects (those funded under the same call and topic, ICT-16-2015), but also other relevant initiatives and projects in the field of Big Data management. This document provides a summary for those activities. And finally, PROTEUS has also played in role in the co-organisation of meetups, in


687691 Page 4 of 37

particular the Flink meetup series in Madrid (Spain) where initiated by Treelogic at the beginning of PROTEUS, in cooperation with other organisations; the series are currently ongoing and there are plans for its continuation.


687691 Page 5 of 37

Document Information

IST Project Number

687691 Acronym PROTEUS

Full Title Scalable online machine learning for predictive analytics and real-time interactive visualization

Project URL www.proteus-bigdata.com EU Project Officer Martina EYDNER

Deliverable Number D6.8 Title Report on project communication and

engagement activities – v2 Work Package Number WP6 Title Communication, exploitation and

dissemination Date of Delivery Contractual M36 Actual M36 Status version 1.2 final ■ Nature report ■ demonstrator □ other □ Dissemination level public ■ restricted □

Authors (Partner) TRILATERAL

Responsible Author

Name Rachel Finn, Graham Hesketh, Giuseppe Maio

E-mail

[email protected]

Partner TRILATERAL Phone +44 (0)207 559 3550 Abstract (for dissemination)

This document provides details about the communication and engagement activities and materials produced during the PROTEUS project, as part of the communication activities and engagement with relevant related initiatives. It is important to highlight that this deliverable is focusing on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). This document provides, initially, information about the general dissemination of the project, including the corporate image design, communication materials, and established communication channels and tools in use. And then a summary on clustering and engagement activities carried out in the course of the project.

Keywords Communication, clustering, engagement Version Log Issue Date Rev. No. Author Change 23-March-2017 0.1 Marcos Sacristan

(TREE) Initial ToC

28-March-2017 0.2 Marcos Sacristan (TREE)

First draft contents

07-April-2017 0.3 Marcos Sacristán (TREE)

Adding abstract, summary.


687691 Page 6 of 37

03-May-2017 0.4 Marcos Sacristán (TREE)

Version ready for review

09-May-2017 0.5 Rachel Finn (TRI), Marcos Sacristán (TREE)

Refinement and improvement after review

15-May-2017 1.0 Marcos Sacristán (TREE)

Final release

20-Nov-2018 1.1 Rachel Finn, (TRI) Version ready for review 29-Nov-2018 1.2 Rachel Finn, Graham

Hesketh, Giuseppe Maio (TRI)

Final release


687691 Page 7 of 37

Table of Contents Executive summary ......................................................................................................................... 3 Document Information .................................................................................................................... 5 Table of Contents ............................................................................................................................ 7 List of figures and list of tables ....................................................................................................... 8 Abbreviations.................................................................................................................................. 9 1 Introduction ............................................................................................................................ 10 2 Communication activities, channels and tools ......................................................................... 11

2.1 Corporate image ............................................................................................................... 11 2.1.1 PROTEUS logo ......................................................................................................... 11 2.1.2 Corporate templates ................................................................................................... 12 2.1.3 Acknowledges & disclaimer ...................................................................................... 14

2.2 Communication materials ................................................................................................. 14 2.2.1 Promotional video...................................................................................................... 14 2.2.2 Corporate presentation ............................................................................................... 15 2.2.3 Promotional leaflet .................................................................................................... 16 2.2.4 Corporate poster ........................................................................................................ 19

2.3 Communication channels and tools .................................................................................. 20 2.3.1 Public website............................................................................................................ 20 2.3.2 Twitter account .......................................................................................................... 23 2.3.3 GitHub organisation................................................................................................... 24 2.3.4 Blogs ......................................................................................................................... 27

3 Clustering and engagement activities ...................................................................................... 28 3.1 Joint workshops and conferences...................................................................................... 28

3.1.1 Big R&I workshop at FiCloud 2016 ........................................................................... 28 3.1.2 Big Data Spain 2016 .................................................................................................. 28 3.1.3 BDVA Summit 2016 ................................................................................................. 29 3.1.4 EuroPro Workshop at the EDBT/ICDT 2017 ............................................................. 29 3.1.5 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases ........................................................................................................... 30 3.1.6 Smart Manufacturing workshop at the European Big Data Value Forum Summit 2018 30

3.2 Other activities ................................................................................................................. 31 3.2.1 Public talks and presentations .................................................................................... 31 3.2.2 Sponsored events ....................................................................................................... 31 3.2.3 Interaction with sister projects and related initiatives ................................................. 32 3.2.4 Co-organisation of meetups ....................................................................................... 34

4 Conclusions ............................................................................................................................ 36 References .................................................................................................................................... 37


687691 Page 8 of 37

List of figures and list of tables Figure 1: PROTEUS logo ............................................................................................................. 12 Figure 2: PROTEUS corporate document template (cover) ........................................................... 13 Figure 3: PROTEUS corporate presentation template (cover) ........................................................ 14 Figure 4: PROTEUS acknowledges and disclaimer ....................................................................... 14 Figure 5: PROTEUS video, snapshots ........................................................................................... 15 Figure 6: PROTEUS presentation, cover and index snapshots ....................................................... 16 Figure 7: PROTEUS leaflet, main page ......................................................................................... 18 Figure 8: PROTEUS poster ........................................................................................................... 19 Figure 9: PROTEUS website, overview ........................................................................................ 21 Figure 10: PROTEUS website, analytics (overview) ..................................................................... 22 Figure 11: PROTEUS website, analytics (most visited pages) ....................................................... 23 Figure 12: PROTEUS Twitter account, snapshot ........................................................................... 24 Figure 13: PROTEUS GitHub organisation, home page (screenshot) ............................................ 25 Figure 14: PROTEUS GitHub organisation, analytics ................................................................... 26 Table 1: Summary of interaction activities with PROTEUS’ sister projects (ICT-16-2015) ........... 32 Table 2: Summary of interaction activities with other projects or initiatives .................................. 34


687691 Page 9 of 37

Abbreviations

BDVA: Big Data Value Association

PEICH: the PROTEUS Elastic Cache SOLMA: the PROTEUS library for Scalable Online Machine Learning and Data Mining Algorithms


687691 Page 10 of 37

1 Introduction This document provides details about the communication and engagement activities and materials produced in the project during the PROTEUS project. It is produced under the course of the task T6.5.- Communication activities and engagement with relevant related initiatives. It is important to highlight the context of this deliverables D6.8 with regard to other deliverables in WP6. In particular, it is to be noted that this deliverable is focusing on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). This document is structured as follow:

• Section 2 provides information about the general dissemination of the project, including the corporate image design, communication materials, and established communication channels and tools in use.

• Section 3, focuses on the clustering and engagement activities carried out in the course of the project, providing a summary of the actions taken.

• Finally, Section 4 will close the document with some conclusions.


687691 Page 11 of 37

2 Communication activities, channels and tools In this first section the document focuses on some core communication elements for the project: firstly, the corporate image established for PROTEUS (including logo, branding details, corporate colours, etc.); secondly, several communication materials (assets) that have been produced as promotional resources; and finally, the main communication channels and tools that have been established and are being used for communication purposes of the project at the time of this report. PROTEUS aims to reach a wide range of stakeholder categories, from the specific scientific fields tackled by the project, to the end user community, and finally to players who will ultimately contribute to delivering results to the market. The consortium will target the following stakeholder categories:

• End users, including industrial players, in different application domains.

• Industry and technical experts, including large companies, SMEs.

• Academia, including universities, research institutes and think tanks involved in the fields of big data processing, streaming analytics, machine learning, distributed systems, and visualization techniques.

• Related research projects, including consortia working on projects related to big data, especially projects funded under the same call topic (ICT-16-2015) and the Big Data Value Association.

• The media, including local, national and international TV, radio and printed media.

• The public. PROTEUS has enabled different materials, channels and tools for different purposes, aiming to address these specific categories. These will be detailed in the coming parts of this section. In terms of activities, however, not all of the categories of stakeholders have been addressed in this document, considering the context of D6.8 with regard to other deliverables in WP6: this deliverable is focusing on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). For this reason, out of the categories previously referred, this document is focusing on engaging with the media, the general public, and other related research project, while other deliverables are addressing end users and industry (D6.4), as well as academia, universities, research communities (D6.6).

2.1 Corporate image

At the beginning of the project, a corporate image was designed to reflect the branding elements to be considered for any communication from the project. The following key elements were agreed in terms of corporate design and branding rules for dissemination.

2.1.1 PROTEUS logo

The PROTEUS logo had already been designed at the proposal stage, and this has been kept as the final logo for the project. It consists on a purple degraded-to-white capital “P”, with a purple triangle on the top left corner, and the project acronym in capital grey letters below.

The logo design can be better appreciated in the figure:


687691 Page 12 of 37

Figure 1: PROTEUS logo

2.1.2 Corporate templates

Following the styles, colors and font defined by the project logo, a whole corporate image was created. In particular and for practical reasons, several templates were produced for its usage in the project communications, including:

• Document template, for all project documents (e.g. particularly used for deliverables).

• Presentation template, for all project presentations.

The figures shows a snapshot of the cover for the presentation and document templates.


687691 Page 13 of 37

Figure 2: PROTEUS corporate document template (cover)


687691 Page 14 of 37

Figure 3: PROTEUS corporate presentation template (cover)

2.1.3 Acknowledges & disclaimer

For the avoidance of doubts, in order to fulfil the communication requirements of the Grant Agreement, default acknowledges and disclaimer were also established for all project communications. The figure shows an example of the acknowledges and disclaimer text, including the EU flag:

“This project is funded by the European Union (Horizon 2020, Ref: 687691)”

Figure 4: PROTEUS acknowledges and disclaimer

2.2 Communication materials

This section provides information and description about the communication materials that the project has produced to date, following the corporate image and providing the essential information required for the different communication activities carried out by the project (both digital on-line, and physical in face-to-face events, conferences and meetings). 2.2.1 Promotional video

A promotional video has been produced in the early stages of the project. The video lasts for 2 minutes, and provides the key information as an abstract of PROTEUS on a multimedia format. It has been conceived as a way to disseminate the basis of the project targeting the general public, the media, and the industry (end users) on multimedia channels, social media and the website, but it can also be used as a resource for any other purpose.


687691 Page 15 of 37

Just as examples, the figure shows a couple of snapshots taken from the video:

Figure 5: PROTEUS video, snapshots

For sharing and dissemination purposes, the video has been uploaded to YouTube, currently available at: https://www.youtube.com/watch?time_continue=3&v=9q48IT0EZxU. It was originally hosted under the Treelogic’s YouTube channel [1], where it had 276 views as of May 2017. After Treelogic left the project, the video now sits on the PROTEUS website and the more general YouTube address above where it has 44 views. This amounts to at least 320 views, given that the views between May 2017 and the video’s transfer off the TREE YouTube channel are uncounted. Nevertheless, 320 views demonstrates a reasonable impact for PROTEUS in the big data and manufacturing space. In addition, the video is also embedded in the home page of the project’s website (see section 2.3.1 for further information about the project’s public website). 2.2.2 Corporate presentation

A corporate presentation of the project was also produced in the early stages in order to provide the basic information in the form of presentation for any public event attended to promote the project and create a stable, recognisable brand identity. The template allows users to specify and personalise their presentation for specific targets and is presented in a format familiar to stakeholders from academia, research and industry. As an example, a couple of snapshot of the presentation’s cover and index can be found in the figure below:


687691 Page 16 of 37

Figure 6: PROTEUS presentation, cover and index snapshots

For the full presentation, this is publicly available at the project’s website at: https://www.proteus-bigdata.com/app/download/9235596170/20151014_PROTEUS_PPT.pdf

2.2.3 Promotional leaflet

A leaflet for PROTEUS was produced at the beginning of the project as part of the project’s communication materials. The aim is for the partners and the project as a whole to count on a factsheet


687691 Page 17 of 37

to be distributed in a printed format when attending events or conferences, targeting mostly industry and research communities. The PROTEUS leaflet is presented in a diptych form, with a cover page, a back page and the main contents inside, including the following sections:

• Cover page: including the project’s logo, name and acronym.

• Main contents:

o Abstract.

o Duration. o Contributions & impact.

o A short project description.

• Back page: including the project’s logo, a list of the partners, and acknowledgements. More detailed information on the leaflet was released as deliverable D6.10 PROTEUS factsheet leaflet. The leaflet itself can be found as a public document available at the PROTEUS website at: https://www.proteus-bigdata.com/app/download/9107917670/20150111_Proteus_Brochure.pdf. As an example and reference, the central part of the leaflet providing the main contents is shown in the figure:


687691 Page 18 of 37

Figure 7: PROTEUS leaflet, main page


687691 Page 19 of 37

2.2.4 Corporate poster

A corporate poster was also produced in the early stages to provide the basic information in the form of a poster for any public event attended to promote the project. The poster’s design is shown in the figure below:

Figure 8: PROTEUS poster


687691 Page 20 of 37

For the high quality pdf file, this is publicly available at the project’s website at: https://www.proteus-bigdata.com/app/download/9235596270/Proteus_poster.pdf

2.3 Communication channels and tools

While the previous sections provided information on the project’s communication material, this section provides information about the communication channels established by the PROTEUS project. In particular PROTEUS is using three online communication channels and tools for different purposes:

• A public website, targeting the general public and anyone generally interested in the project.

• A Twitter account, as an interactive social media channel for the more dynamic and daily activity.

• A GitHub organisation (including different repositories), to engage with the developers community and provide the Open Source contributions of the project.

The following sections provide further information about the purpose and use of each of these tools for PROTEUS communications. 2.3.1 Public website

This section provides a summary of the design and implementation of the PROTEUS project website. The actual website is available at http://www.proteus-bigdata.com/. The website was hosted by the project coordinator Treelogic, and still remains active. It includes the following sections:

• Home

o Project fiche

• Project description

o Hybrid processing engine

o Scalable online machine learning o Real-time interactive visual analytics

o Industrial validation

• Partners

• Results

o Software o Public deliverables

o Publications o Dissemination

• Blog (news and events)

• Contact. For further information, deliverable D6.9 “PROTEUS project website” previously released in month 3 of the project provided details about the project’s website. The figure provides an overview:


687691 Page 21 of 37

Figure 9: PROTEUS website, overview


687691 Page 22 of 37

The website was launched in December 2015 (already at month 1 of the project) and has been active since then. The following presents some figures for the website analytics through 31st October 2018.

Figure 10: PROTEUS website, analytics (overview)

These total figures (as of Oct 2018) also demonstrate a sustained engagement with stakeholders during the term of the project, with approximately the same number of visitors accessing the site and its pages in each half of the project.

RP1 RP2 Total

Visitors 3,648 3,192 6,840

Page views 8,172 6,407 14,579

The screenshot from the analytics (below) indicates that PROTEUS has created a small community of stakeholders who are engaging with our results in a sustained manner. Our software page alone has been visited by 76 different users who each viewed an average of 4 pages. Additional page views indicate that the major innovations within PROTEUS – Scalable online machine learning, real-time interactive visual analytics and hybrid processing engine – were popular among visitors. Our results pages have been visited by up to 70 users who viewed an average of 10 (sub-)pages or viewed the same page 10 times. This indicates more than a passing glance at our results and suggests genuine interest in our deliverables and outputs. In addition, our blog posts have been visited by 35 people who appear to be reading information there regularly, given the figure of 305 page views from among those 35 visitors.


687691 Page 23 of 37

Figure 11: PROTEUS website analytics (project results)

2.3.2 Twitter account

Since February 2017, PROTEUS is also managing a Twitter account under the username @proteus_bigdata (https://twitter.com/proteus_bigdata). This account has been conceived for sharing more dynamic and quickly changing information, as well as providing an interactive communication channel for engaging with relevant stakeholders


687691 Page 24 of 37

(industry, researchers, academia, other projects and initiatives, etc.). This communication channel uses hashtags to ensure that our messages are amplified beyond our followers and show up on newsfeeds created and followed by users (e.g., #bigdata, #manufacturing, #industry40). This allows us to gain new followers and augment our network of interested parties. The project has already gathered 72 followers in the last three months, and this is set to increase as the release of results from the project enable PROTEUS to share more meaningful information. Our social media presence has gained momentum during the course of the project. Our followers have more than doubled since RP1 and our “Likes” have increased seven-fold since the start of the project.

RP1 RP2 Total

Followers 57 87 144

Tweets 75 207 282

Likes 24 148 172

Figure 12: PROTEUS Twitter account, snapshot

2.3.3 GitHub organisation

PROTEUS is a project releasing most of the results as public, including both document reports and prototypes (software). For this reason, and in order to further communicate and engage with the developers’ community, the project is managing a GitHub organisation named PROTEUS-H2020 (https://github.com/PROTEUS-H2020).


687691 Page 25 of 37

Figure 13: PROTEUS GitHub organisation, home page (screenshot)


687691 Page 26 of 37

The organisation is composed by a set of repositories where the different components of PROTEUS are being developed, released, and shared with the community. The most relevant ones include:

• The PROTEUS Engine: an overhauled version of Apache Flink supporting hybrid computation on batch datasets and data streams.

• The PROTEUS Language: a declarative language library for Scalable Data Analysis on top of Apache Flink.

• PEICH: the PROTEUS Elastic Cache.

• The PROTEUS Incremental Analytics: a backend module that implements incremental version (~O(1) computational cost using approximations) of the most common analytics operations.

• SOLMA: the PROTEUS library for Scalable Online Machine Learning and Data Mining Algorithms adapted to the data analytics platform Apache Flink.

• PROTEIC.JS: An HTML5 and CSS3 charts library, tailored for batch and streaming data visualisations.

In addition to these, some other utilities are also being released in the GitHub as the project advances, including test examples, demo setup configurations on a containers-based architecture (i.e. Docker), streaming dataset simulators for experimentation and testing, among others. The figure below provides some analytics of the usage and activity within the PROTEUS GitHub organisation on the different repositories (access: May-9th-2017):

Figure 14: PROTEUS GitHub organisation, analytics


687691 Page 27 of 37

2.3.4 Blogs

In addition to the social media presence, the project also produced a series of blogs hosted on the PROTEUS website to enable stakeholders to stay up to date with our activities and our research outreach activities. These blogs were published on a dedicated website page and covered the following topics: • Workshop at the European Big Data Value Forum: Industrial Data Platforms for the

Manufacturing domain • PROTEUS presented at the LibreCon 2017 • Presentation at the Flink Forward Berlin 2017 • First review meeting • PROTEUS at the Digital Business World Congress (DES 2017) • PROTEUS at the EDBT/ICDT 2017 • Upcoming consortium meeting in Berlin • PROTEUS presented at the Big Data Spain 2016 • Public events & consortium meeting in Bournemouth • FiCloud 2016 Where possible, blog posts were accompanied by pictures and/or slides to make the content more accessible and engaging for different audiences. In addition, blog posts were also written by partners and hosted on their organisations websites. For example, Trilateral wrote two blog posts about PROTEUS: Artificial Intelligence makes manufacturing smart Research ethics for Industry 4.0

In addition, PROTEUS was also featured in Trilateral’s annual newsletter in 2016 and 2017 to >300 individual subscribers. A final piece is planned for the 2018 annual newsletter to be distributed in December.


687691 Page 28 of 37

3 Clustering and engagement activities This section provides information on the activities carried out by PROTEUS aiming to link, interact, cluster and engage from a broad perspective, including other projects or activities, developers and technical communities, and the general public.

3.1 Joint workshops and conferences

In this section we present a list of joint workshops and clustering events attended or organised during the project, focusing on networking and engagement with other projects and initiatives (beyond the purely scientific dissemination and publications). 3.1.1 Big R&I workshop at FiCloud 2016

The 3rd International Symposium on Big Data Research and Innovation (BigR&I-2016) was organised as part of the FiCloud 2016 conference [2] (The IEEE 4th International Conference on Future Internet of Things and Cloud). PROTEUS was participating in this workshop with a publication “An incremental approach for real-time Big Data visual analytics” [3], jointly co-authored by TREE and BU, Ignacio García, Rubén Casado (TREE) and (BU). The workshop session enabled PROTEUS to interact with other existing initiatives and research in the field. The following list provides an overview of the workshop presentations and discussion:

• “Taxonomy for Unsecure Big Data Processing in Security Operations Centers”. Natalia Miloslavskaya, Alexander Tolstoy and Sergey Zapechnikov.

• “Forecasting The Natural Gas Price Trend – Evaluation of a Sentiment Analysis”. Tina Grundmann, Carsten Felden and Marco Pospiech.

• “Rule- and Cost-Based Optimization of OLAP Workloads on Distributed RDBMS with Column-Oriented Storage Function”. Takamitau Shioi and Kenji Hatano.

• “Infrastructure-Aware Functional Testing of MapReduce programs”. Jesús Morán, Bibiano Rivas, Claudio De La Riva, Javier Tuya, Ismael Caballero and Manuel Serrano.

• “An incremental approach for real-time Big Data visual analytics” Ignacio García Fernández, Rubén Casado and Hamid Bouchachia.

3.1.2 Big Data Spain 2016

The Big Data Spain [4] international conference is a business-oriented conference taking place yearly in Madrid around Big Data technology. 2016 edition brought together more than 1,000 attendees, and a successful program with more than 50 talks and workshops. PROTEUS was presented in Big Data Spain 2016 twofold, presenting both a talk and a poster session for networking:

• TALK: “Advanced data science algorithms applied to scalable stream processing”, by David Piris and Ignacio García (TREE) [5].


687691 Page 29 of 37

• POSTER session: “A novel real-time processing and visualization big data solution focused on incremental processing”, by Ignacio García and Jorge Yagüe (TREE).

3.1.3 BDVA Summit 2016

The Big Data Value Association (BDVA) [6] is a fully self-financed non–for-profit organisation under Belgian law. The Big Data Value Association (BDVA) is the private counterpart to the EU Commission to implement the BDV PPP programme (Big Data Value PPP). BDVA has over 150 members all over Europe with a well-balanced composition of large and small and medium-sized industries as well as research organizations. Even if PROTEUS was funded before the BDV PPP had been established and is not officially under the PPP umbrella, the links between PROTEUS and the BDVA are clear, and the association represents a natural target group for several purposes, including a research network, but also dissemination, communication and exploration activities. PROTEUS was present in the BDVA Summit 2016 [7] (November 30th to December 2nd 2016, in Valencia - Spain) through project partners TRI and TREE who attended the event. In this occasion, the summit was used for networking and exploration of future cooperation and synergies, and even if PROTEUS was not presented, there are plans to exploit BDVA events in the future for a more active participation of PROTEUS.

3.1.4 EuroPro Workshop at the EDBT/ICDT 2017

The EuroPro workshop (1st International Workshop on Big Data Management in European Projects [8]) was organised as part of the EDBT/ICDT 2017 Joint Conference [9]. Together with other 7 EU-funded projects in the field of Big Data Management, PROTEUS participated to this clustering workshop with a presentation and further discussion searching for synergies, represented by Bonaventura del Monte (DFKI).

The list of participants and presentations follow:

• Christos Doulkeridis, Nikos Pelekis, Yannis Theodoridis, and George Vouros. “Big Data Management and Analytics for Mobility Forecasting in datAcron”.

• Philipp Grulich, Tilmann Rabl, Volker Markl, Csaba Sidló, and Andras Benczur. “STREAMLINE - Streamlined Analysis of Data at Rest and Data in Motion”.

• Bonaventura Del Monte, Jeyhun Karimov, Alireza Rezaei Mahdiraji, Tilmann Rabl, and Volker Markl. “PROTEUS: Scalable Online Machine Learning for Predictive Analytics and Real-Time Interactive Visualization”.

• Edwin Morley-Fletcher. “MHMD: My Health, My Data”.

• Sergi Nadal, Alberto Abelló, Oscar Romero, and Jovan Varga. “Big Data Management Challenges in SUPERSEDE”.

• Claudio A. Ardagna, Paolo Ceravolo, Ernesto Damiani, and Marcello Leida. “Scouting Big Data Campaigns using TOREADOR Labs”.

• Hajira Jabeen, Phil Archer, Simon Scerri, Aad Versteden, Ivan Ermilov, Giannis Mouchakis, Jens Lehmann, and Soeren Auer. “Big Data Europe”.


687691 Page 30 of 37

3.1.5 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases

The PROTEUS project also featured regularly in workshops co-organised by BU within the ECML/PPDD annual conference series. ECML/PPDD is a long-running (16-year), well respected forum for researchers and industry to discuss machine learning and data analytics. BU organised three workshops within this larger conference focusing on slightly different machine learning themes but with an overall theme of large-scale learning and streaming data.

• 2016: Large-scale Learning from Data Streams in Evolving Environments

• 2017: IoT Large Scale Learning from Data Streams

• 2018: IoT Large Scale Learning from Data Streams Through these workshops and their associated presentations, PROTEUS innovations were shared with more than 500 relevant scientific and innovation-focused stakeholders across the term of the project.

3.1.6 Smart Manufacturing workshop at the European Big Data Value Forum Summit 2018

PROTEUS, jointly with other initiatives and projects jointly organised workshop on Industrial Data Platforms for the Manufacturing domain during the European Big Data Value Forum (EBDVF), held in Versailles (Palais des Congrès), France from 21st to 23rd November 2017. The EBDVF results of a joint initiative between the Big Data Value Association (BDVA) and the European Data Forum (EDF), as a key European event for industry professionals, business developers, researchers, and policy makers to discuss the challenges and opportunities of the European data economy and data-driven innovation in Europe. The workshop was organised as part of the smart manufacturing Working Group within the BDVA and in collaboration with PROTEUS, TOREDOR, MATIS, ARROWHEAD and CREMA projects. The list of participants and presentations are as follows:

• Smart Manufacturing Industry – by Mr. Sergio Gusmeroli (Engineering, Italy) and Mr. Davide Dalle Carbonare (Engineering, Italy)

• Digitalization needs on Interoperability and requests for standardization. The way to do it– by Mr. Thomas Hahn (SIEMENS, Germany)

• Securing IoT and Big Data. Exploring a huge iceberg by some illustrative projects – by Dr. Aizea Lojo (IK4-Ikerlan, Spain, Researcher), and Dr. Iñaki Garitano (Mondragon Unibertsitatea, Spain, Researcher/Lecturer)

• Predictive maintenance for Aerospace Products Manufacturing – by Prof. Ernesto Damiani (CINI, Italy, TOREADOR Principal Investigator), and Dr. Mariangela Lazoi (Università del Salento, Italy, TOREADOR Aerospace Pilot coordinator)

• Steel Production Optimization – by Mr. Marcos Sacristán (Treelogic, Spain, PROTEUS project coordinator)

The event was attended by all the PROTEUS partners and numerous links between different projects were explored.


687691 Page 31 of 37

3.2 Other activities

This section summarises a list of other engagement, general communication, and clustering activities carried out by PROTEUS during the period. It includes public talks and presentations, interaction with related projects and initiatives, or meetups, among other activities.

3.2.1 Public talks and presentations

In addition to the workshops and conferences attended as listed in the previous section, PROTEUS has also organised smaller scale presentations and public talks linked to the project’s events (e.g. project’s meetings). The following lists some talks organised by PROTEUS:

• Public talk at the University of Oviedo during the kickoff meeting (January 22nd, 2016). “Introduction to Apache Flink”, by Asterios Katsifodimos (DFKI) and Rubén Casado (TREE).

• Public talk at the University of Oviedo by Ignacio García and Jorge Yagüe (TREE). “ProteicJS - innovando en la visualización de datos a través del software libre” [10] (in English: ProteicJS – innovating in data visualisation with Open Source software). October 14th, 2016.

• Public talk at the University of Bournemouth during a consortium meeting (October 20th 2016). A guest speaker was invited together with a presentation of the visualisation library developed in PROTEUS, named PROTEIC.JS:

o “The Cloudera Big Data ecosystem”, by Justin Hancock (CLOUDERA, guest speaker).

o “Introducing Proteic.js, a novel visualisation library for Big Data”, by David Piris and Ignacio García (TREE).

• Public talk entitled “Apache Flink - Big Data Stream Processing” at the XLDB 2017 Conference (10-12 October 2017)

• Public talk by Graham Hesketh (TRI) entitled “PROTEUS: Example of an EU project around digital twins” at the SIRUS Digital Twins workshop, University of Oslo (20th March 2018).

• Public talk by Giuseppe Maio (TRI) at the “Data4AI: technical challenges to make it possible” workshop organised at the European Big Data Value Forum jointly organised by the BDVA and BDVF (14 November 2018)

3.2.2 Sponsored events

PROTEUS also increased stakeholders’ awareness of the project by sponsoring a large, international conference on Artificial Intelligence in September 2018. The inaugural AI Congress included more than 1000 delegates from industry and research organisations. It was a primarily industry focused event with more than 100 industry speakers, including:


687691 Page 32 of 37

Our sponsorship included a half-page spread in the conference booklet naming PROTEUS, our partners, relevant keywords, acknowledgement of EC funding and a link to our website for further information. Section 2.3.1 indicates that there was a swell of interest in the PROTEUS website near September 2018, which may be partly a result of our sponsorship of this major event.

3.2.3 Interaction with sister projects and related initiatives

PROTEUS has devoted special attention to interacting and clustering with the sister projects, those funded under the same call as PROTEUS (ICT-16-2015) researching on synergistic and complementary fields. The table presents a summary on the interaction and clustering activities performed with other projects and initiatives during the first half of the project:

Table 1: Summary of interaction activities with PROTEUS’ sister projects (ICT-16-2015)

Project acronym Project name Website Activities


687691 Page 33 of 37

VaVeL Variety, Veracity, VaLue: Handling the Multiplicity of Urban Sensors Cloud-LSVA Cloud Large Scale Video Analysis

http://www.vavel-project.eu/

• Email discussions.

SETA An open, sustainable, ubiquitous data and service ecosystem for efficient, effective, safe, resilient mobility in metropolitan areas

http://setamobility.eu/


HOBBIT Holistic Benchmarking of Big Linked Data

https://project-hobbit.eu/

• Email discussions. • Bilateral e-Meeting

(March 10th 2017) • Physical meeting (14th

March 2018) STREAMLINE Streamline https://streamline.

sics.se/ • Email discussions. • Joint workshop (EuroPro,

March 21st 2017). • Bilateral f2f meeting

(Berlin, 26th April 2017) • Publications

datACRON Big Data Analytics for Time Critical Mobility Forecasting

http://www.datacron-project.eu/

• Email discussions. • Joint workshop (EuroPro,

March 21st 2017) SUMMA Scalable Understanding of

Multilingual Media http://summa-project.eu/

• Social media (twitter) contact.

TOREADOR TrustwOrthy model-awaRE Analytics Data platfORm

http://www.toreador-project.eu/

• Email discussions. • Joint workshop (EuroPro,

March 21st 2017) • Joint workshop (EBDVF,

November 13th 2017) SEE.4C SpatiotEmporal ForEcasting:

Coopetition to meet Current Cross-modal Challenges

http://see4c.eu/ • Email discussions.

Cloud-LSVA Cloud Large Scale Video Analysis

http://cloud-lsva.eu/


We collaborated with STREAMLINE consortium in several publications. DFKI's tasks in the two projects were different and executed by different people. As a result, we were able to transfer knowledge and experience between the projects and project members in several meetings. Also, there was no double booking of personnel in the two projects.

• Karimov, J., Rabl, T., Katsifodimos, A., Samarev, R., Heiskanen, H., & Markl, V. (2018, April). Benchmarking distributed stream data processing systems. In 2018 IEEE 34th International Conference on Data Engineering (ICDE) (pp. 1507-1518). IEEE.

• Kunft, A., Katsifodimos, A., Schelter, S., Rabl, T., & Markl, V. (2017). Blockjoin: efficient matrix partitioning through joins. Proceedings of the VLDB Endowment, 10(13), 2061-2072.


687691 Page 34 of 37

• Carbone, P., Traub, J., Katsifodimos, A., Haridi, S., & Markl, V. (2016, October). Cutty: Aggregate sharing for user-defined windows. In Proceedings of the 25th ACM International on Conference on Information and Knowledge Management (pp. 1201-1210). ACM.

• Alexandrov, A., Salzmann, A., Krastev, G., Katsifodimos, A., & Markl, V. (2016, June). Emma in action: Declarative dataflows for scalable data analysis. In Proceedings of the 2016 International Conference on Management of Data (pp. 2073-2076). ACM.

• Alexandrov, A., Kunft, A., Katsifodimos, A., Schüler, F., Thamsen, L., Kao, O., ... & Markl, V. (2015, May). Implicit parallelism through deep language embedding. In Proceedings of the 2015 ACM SIGMOD International Conference on Management of Data (pp. 47-61). ACM.

• Karimov, J., Rabl, T., & Markl, V. (2018, August). PolyBench: The First Benchmark for Polystores. In Technology Conference on Performance Evaluation and Benchmarking (pp. 24-41). Springer, Cham.

Besides sister projects, PROTEUS have interacted with additional projects or initiatives in the field of Big Data research during the period. The table provides a summary of these activities:

Table 2: Summary of interaction activities with other projects or initiatives Project acronym Project name Website Activities

BDVA Big Data Value Association http://www.bdva.eu/

• TRI and TREE attended the BDVA Summit (November 30th 2016, Valencia)

My Health My Data

A new paradigm in healthcare data privacy and security

http://www.myhealthmydata.eu/

• Joint workshop (EuroPro, March 21st 2017, Venice - Italy)

Big Data Europe

Empowering communities with data technologies

https://www.big-data-europe.eu/


SUPERSEDE SUpporting evolution and adaptation of PERsonalized Software by Exploiting contextual Data and End-user feedback

https://www.supersede.eu/


DataBench investigate existing Big Data benchmarking tools and projects to link economic and technical benchmarks

https://www.databench.eu/the-project/

• eMeeting (May 15th 2018)

3.2.4 Co-organisation of meetups

PROTEUS has been involved as a co-organiser of the Apache Flink Meetup series in Madrid (Spain) [11] since the first edition, and has participated in subsequent FlinkForward conferences organised as a follow-up to the Meetups. TREE and DFKI participated in the following Flink Community events throughout the course of the project.


687691 Page 35 of 37

• 1st Flink Meetup in Madrid, 25th February 2016, Madrid (Spain). “Introduction to Apache Flink” [12]. Presentations by Rubén Casado (TREE) and Dr. Fabian Hueske, PCM Apache Flink and co-founder of DataArtisans.

• 2nd Flink Meetup in Madrid, 20th May 2016, Madrid (Spain). “Workshop Apache Flink” [13], a more practical meetup with presentations by Rubén Casado and David Piris (TREE).

• 3rd Flink Meetup Madrid, 15th December 2016, Madrid (Spain). “Apache Flink en acción” (Apache Flink in action) [14], with presentations by Rubén Casado (TREE) and Zhenhao Li (Accenture Netherlands).

• Presentation at FlinkForward 2017 Conference: Members of DFKI team gave a presentation with title “I²: Interactive Real-Time Visualization for Streaming Data with Apache Flink and Apache Zeppelin” [1].

• Presentation at FlinkForward 2018 Conference: Members of DFKI team gave a presentation with title “Efficient Window Aggregation with Stream Slicing” [2].

There are plan to continue the meetups series in the future, and members of the PROTEUS consortium will remain involved in this community, pushing PROTEUS innovations and additional innovations enabled by PROTEUS results, to the community.


687691 Page 36 of 37

4 Conclusions This document has provided details about the communication and engagement activities and materials produced in the project during the PROTEUS project, as part of the communication activities and engagement with relevant related initiatives. It is important to highlight that this deliverable has focused on the general communication activities (e.g. general public) as well as on engagement and clustering activities with other related projects and initiatives, while other deliverables are more focused on the scientific dissemination activities (i.e. see deliverable D6.6) and the technology transfer and community building activities (i.e. see deliverable D6.4). PROTEUS has been active in communication activities, and ran three different channels targeting different purposes. Firstly, for the general public and for general communication purposes, a project’s public website is available online (www.proteus-bigdata.com), continuously being populated and updated with new information and contents (new, public results, events, etc.). Secondly, for a more interactive communication and presence on the social media, a Twitter account is being used (https://twitter.com/proteus_bigdata). And finally, since the PROTEUS results are being released as Open Source, a GitHub organisation is hosting the software results of the project in several repositories, constantly updated with the evolution of the project (PROTEUS GitHub at: https://github.com/PROTEUS-H2020). In terms of clustering and engagement activities, the project has initiated and participated in a number of actions. Several workshops and conference have been attended with the aim of engaging with related initiatives and projects exploring relevant areas. In particular, FiCloud, Big Data Spain, the Big Data Value Association or the EuroPro workshops were venues where PROTEUS was presented, interacted with other initiatives, and engaged with the relevant community and stakeholders. Finally, other activities included public talks and presentations given by the project’s partners during the project meetings or in conjunction with close organisations (e.g. other academic bodies). The interaction with related initiatives has also been fostered during this period, especially with sister projects (those funded under the same call and topic, ICT-16-2015), but also other relevant initiatives and projects in the field of Big Data management. And in addition, PROTEUS has also played in role in the co-organisation of meetups, in particular the Flink meetup series in Madrid (Spain) initiated by Treelogic at the beginning of PROTEUS, in cooperation with other organisations; and the series are currently ongoing. As a final conclusion of this document, the consortium feels positive about the results and impacts achieved during the project, in terms of general communication, clustering and engagement with other relevant initiatives. Several tools and channels were established, and the both the quantity and quality of engagement actions undertaken are considered satisfactory.


687691 Page 37 of 37

References [1] Treelogic Youtube channel:

https://www.youtube.com/channel/UC7YXVurU4sBwKaRTNhj124Ab [2] FiCloud, The IEEE 4th International Conference on Future Internet of Things and Cloud:

http://www.ficloud.org/2016/workshops.php [3] I. García, R. Casado and A. Bouchachia, "An Incremental Approach for Real-Time Big Data

Visual Analytics," 2016 IEEE 4th International Conference on Future Internet of Things and Cloud Workshops (FiCloudW), Vienna, 2016, pp. 177-182. doi: 10.1109/W-FiCloud.2016.46, available at: http://ieeexplore.ieee.org/abstract/document/7592720/?reload=true

[4] Big Data Spain: www.bigdataspain.org [5] “Advanced data science algorithms applied to scalable stream processing”, talk by David Piris

and Ignacio García, Big Data Spain 2016: https://www.bigdataspain.org/program/fri-advanced-data-science-algorithms-applied-scalable-stream-processing.html

[6] Big Data Value Association (BDVA): http://www.bdva.eu/ [7] BDVA Summit 2016 in Valencia: http://www.bdva.eu/?q=valencia-summit [8] EuroPro workshop, 1st International Workshop on Big Data Management in European Projects:

http://edbticdt2017.unive.it/?workshops#europro

[9] EDBT/ICDT 2017 Joint Conference: http://edbticdt2017.unive.it [10] Conference: “ProteicJS - innovando en la visualización de datos a través del software libre”:

http://ingenieriainformatica.uniovi.es/visor/-/asset_publisher/pM6Z/content/conferencia-proteicjs?redirect=%252Finicio

[11] Apache Flink Madrid, meetups series: https://www.meetup.com/es-ES/Meetup-de-Apache-Flink-en-Madrid/

[12] “Introducción a Apache Flink. Networking”, Flink Meetup Madrid, February 25th 2016: https://www.meetup.com/es-ES/Meetup-de-Apache-Flink-en-Madrid/events/228608831/

[13] “Workshop Apache Flink”, Flink Meetup Madrid, May 20th 2016: https://www.meetup.com/es-ES/Meetup-de-Apache-Flink-en-Madrid/events/230565426/

[14] “Apache Flink en acción”, Flink Meetup Madrid, December 15th 2016: https://www.meetup.com/es-ES/Meetup-de-Apache-Flink-en-Madrid/events/235573467/

Documents

D6.8.- Report on project communication and engagement ......PROTEUS Deliverable D6.8 687691 Page 2 of 37 Abstract This document provides details about the communication and engagement