Welcome to our first Newsletter

I N S I D E T H I S

I S S U E :

Welcome 1

What is

MOVING?

Concept and

approach

EconBiz

RecSys tool

Online demo

for lecture

videos

Parcipaon

in TrecVid

Interview at

ZBW-

mediatalk

Parcipaon

in ESWC 2016

Recent and

upcoming

events

Who we are 5

Contact details 6

What is MOVING?

O C T O B E R 2 0 1 6

I S S U E 1

daily research tasks. We argue

that this type of informaon lit-

eracy is important for one to

become a data-savvy informaon

professional, and will have a de-

cisive impact on the innovave

capacity of the European society.

MOVING is a three-year "Re-

search and Innovaon Acon"

that started in April 2016 and is

implemented by a mulnaonal

consorum. MOVING’s goal is to

develop an innovave training

plaorm that will enable people

from all societal sectors (compa-

nies, universies, public admin-

istraon) to fundamentally im-

prove their informaon literacy

by training how to use, choose

and evaluate data/text mining

methods in connecon with their

We would like to welcome you to the rst issue of the MOVING Newsleer.

We aim to inform the community, our readers and supporters of what has been achieved and produced

during the course of the project. In this rst issue, in pages 1 and 2, we outline the concept and the ap-

proach taken in MOVING and how our plaorm tackles all parts of an e-learning system. Some rst re-

sults of the project, as well as informaon on selected disseminaon acvies that took place in the

rst six months of the project, are presented in pages 2, 3 and 4. Since this is our rst newsleer, we

conclude in pages 5 and 6 with a short presentaon of each MOVING partner and the project’s details

and contact informaon.

P A G E 2

“EconBiz

RecSys”,

a publicaon

recommender

system for

Economists

The

architectural

diagram of the

MOVING

plaorm and

its three key

areas

MOVING concept and approach

The MOVING project will train people to cope

with the large amount of Internet-based infor-

maon they are faced with as part of their daily

professional dues. The core challenge of our

current knowledge society is not the access to

informaon itself, but whether people have the

ability to manage this informaon in a profes-

sional way. Responding to this challenge, the

project will develop the open, innovaon train-

ing plaorm “MOVING”, which will serve as:

a) A working environment for the quality and

usability analysis of large text collecons and

free online contents with data mining methods,

equally open to people from science, public

administraon and business.

b) A training environment with informaon,

training and exchange oers in the broad eld

of digital informaon management.

c) A community of pracce where people can exchange ideas on problems, share soluons and experi-

ences with each other.

Tools - Demos - Results

EconBiz RecSys tool

guraons, i.e. HCF-IDF using

only tles of papers and HCF-IDF

using both tles and full-texts of

papers. Dierent from the tradi-

onal methods, HCF-IDF can pro-

vide compeve recommenda-

ons already using only tles.

You can access the tool at:

hp://amygdala.informak.uni-

kiel.de/Demo/TwierAccount

This is a tool developed by MOV-

ING partner ZBW that delivers

recommended scienc papers

in economics, based on what a

user tweeted about. It proles

papers as well as tweets using

our novel method HCF-IDF

(Hierarchical Concept Frequency

Inverse Document Frequency).

HCF-IDF extracts semanc con-

cepts from texts and applies

spreading acvaon based on a

hierarchical thesaurus, which is

freely available in many dierent

domains. Spreading acvaon

enables to extract relevant se-

manc concepts which are not

menoned in texts, and mi-

gates shortness and sparseness

of texts. The novel HCF-IDF

method demonstrated the best

performance in a larger user ex-

periment published at JCDL’16.

In the web-accessible demo of

the EconBiz RecSys tool, you may

compare the two dierent con-

P A G E 3 I S S U E 1

Online demo for lecture videos

Participation in the TRECVID benchmarking activity

Communication and Dissemination activities

ments this year were performed on a

set of Internet Archive videos totaling

about 600 hours of video duraon, and

using 30 dierent queries. Our fully au-

tomac runs performed very well in this

challenging task, compared to the runs

of the other parcipang instuons

from all over the world. Specically, our

best run was ranked 2nd-best, achieving

an inferred average precision of 0.051

(compared to 0.054 reached by the best

-performing parcipant in the fully-

automac category, and 0.040 reached

by the 3rd best-performing one). Inter-

esngly, our fully automac runs also

compared favorably to the manually-

assisted runs that were submied to

AVS: with an inferred average precision

of 0.051, our best fully automac run

also outperformed the runs of all but

one parcipant in the manually-assisted

run category. We also had very good

results in the event-based annotaon

task (MED), where we tested our ma-

chine learning techniques for video an-

notaon, using a variable number of

training samples. Our parcipaon in

the AVS and MED tasks this year was

jointly supported by MOVING and by

another H2020 EU project, InVID.

MOVING, via its

consorum mem-

ber CERTH, suc-

cessfully parci-

pated in the Ad-

hoc Video Search

(AVS) and event-

based annotaon

(MED) tasks of

TRECVID 2016. The AVS task aempts to

model the end-user video search use-

case, where the user is looking for seg-

ments of video containing persons, ob-

jects, acvies, locaons, etc. and com-

binaons of the former. The experi-

Interview about MOVING to ZBW-mediatalk

ZBW’s interview about MOVING

The interacve

online demo for

lecture videos

such as automacally detected shots,

scenes, and visual concepts. You can

access the demo at: hp://mulmedia2

.i.gr/moving-project/lecture-video-link

ing-demo/results.html (best viewed wi-

th Firefox).

MOVING partner CERTH released an

interacve online demo linking lecture

videos, using general purpose concepts

that were produced from textual analy-

sis of their transcripts, with non-lecture

videos, using their visual analysis results

low the link:

hps://www.zbw-mediatalk.eu/2016/0

7/science-2-0-research-project-moving-

big-data-analyses-for-non-computer-sci

ensts/

Prof. Dr. Ansgar Schrep illustrated the

project details for ZBW-mediatalk, a

blog about technologies, services and

innovaons for libraries and media

companies. For more details please fol-

CERTH is going to pre-

sent in ACM Mulme-

dia 2016 a new transfer

learning approach that

leads to beer concept-based video annotaon

than exisng state-of-the-art methods.

Recent and upcoming events

CERTH presented at the ICIP

2016 confernece a new ma-

chine learning method that

takes the input data uncer-

tainty into consideraon dur-

ing training. This method is applicable for various

video understanding problems such as video event

detecon and video aesthec assessment. You can

get access to the full paper text at: hps://

zenodo.org/record/159236#.V_eSGcm9Hao

ZBW gave a presenta-

on at the JCDL’ 2016

of a novel HCF-IDF pro-

ling method that will

be used in the MOVING plaorm. Titles of scien-

c publicaons are sucient to achieve compe-

ve recommendaon results when employing this

novel method. You can get more informaon

about this paper at: hps://zenodo.org/record/61

391#.V-D4XzW9Hao

TUD presented at

the Internaonal

Science 2.0 con-

ference the MOV-

ING project. You can get more informaon about

this presentaon at: hps://www.open-science-

conference.eu/wp-content/uploads/2016/05/09_

Herbst_Koehler_Scherp_-_Data_driven_online_re

search_Potenal_specicaons_in_relaon_to_us

er_needs.pdf

PBF presented at the ICMFII’

2016 conference an applica-

on of a roadmapping meth-

odology to establish a strate-

gic plan for an innovave

knowledge repository that

could provide dynamically

updated economic informaon, online courses

and other data. PBF also illustrated an example of

building an exploitaon strategy for the digital

plaorm of MOVING.

TUD parcipated in the internaonal European

doctoral summer school in Linz, Austria (hp://

www.edu-tech.eu/?pid=49) and in the rst Asian

summer school in Yogjakarta, Indonesia (hp://

icvet.uny.ac.id/node/350), where Prof. Dr. Thomas

Köhler presented the MOVING project.

The MOVING

consorum

parcipated in

a variety of

naonal

and

internaonal

conferences,

workshops

and summer

schools

Successful participation in ESWC 2016

You can get access to the full paper text at: hps://zenodo.org/record/61520#.V-D4uTW9Hao

Parcipaon in

the 13th

European

Semanc Web

Conference

P A G E 4

ject on the video analysis ser-

vices, a scienc paper recom-

mender and the MOVING

plaorm based on TU Dresden’s

eScience plaorm were present-

ed. For more details please fol-

low the link: hp://2016.eswc

-conferences.org/program/eu-

project-networking-session

The MOVING project parcipat-

ed with an own presentaon in

the EU Project Networking Ses-

sion at ESWC on Wednesday,

June 1st, 2016. In total 14 EU

projects parcipated in the ses-

sion. For MOVING several highly

relevant connecons not only to

other projects but also to further

potenal industry partners were

made. During the session, early

prototypes of the MOVING pro-

The MOVING

consorum

consists of

public and

private

partners ; in

total, 9

academic,

industrial, and

research

organizaons

from Greece,

Germany,

Austria,

Slovenia, UK,

and Poland

MOVING consortium - who we are

P A G E 5

Centre for Research & Technology Hellas (CERTH), is the only re-

search centre in Northern Greece and one of the largest in the country.

CERTH parcipates in this project through Informaon Technologies

Instute (ITI) with the Mulmedia Knowledge and Social Data Analycs lab (MKLab), with experi-

ence and scienc experse in collecon, indexing and mining of mulmedia and social network data

from heterogeneous Internet sources. hp://www.i.gr

EY (Ernst & Young) is one of the largest privately-held rms providing ac-

counng, tax, transacon and advisory services. EY has more than 700 oces

providing auding and accounng services in 140 countries with worldwide

145.000 professionals (auditors, tax advisory professionals, transacons spe-

cialists and advisers). hp://www.ey.com/Home

Technische Universitat Dresden (TUD) is a leading university in Ger-

many and one of 11 German universies that were idened as

“University of Excellence” in 2012. TUD parcipates in this project

through the Media Centre (MZ). MZ lead by Prof. Kohler covers specialists for the areas of infor-

maon and computer science, audio-visual media, didaccs, psychology and business engineering.

hps://tu-dresden.de

The Know-Center (KC) is Austria’s centre for data-driven business and

big data analycs. KC conducts applicaon-oriented research in coop-

eraon with other academic instuons and with companies. KC par-

cipates in this project through Ubiquitous Personal Compung

group which researches computer-supported working and workplace learning, and the Knowledge

Visualizaon group which develops visual analycs methods for exploraon and interacve analysis

of large repositories. hp://www.know-center.tugraz.at/

Instut Jozef Stefan (JSI) is the leading Slovenian research instuon

covering research in natural sciences, life sciences and engineering. JSI

parcipates in this project through the Centre for Knowledge Transfer

in Informaon Technologies which specializes in the areas of research results disseminaon and

eLearning. The centre is well known by portals: VideoLectures.NET, ScienceAtlas.ijs.si and IST-

World.Org. hps://www.ijs.si/ijsw/JSI

Leibniz Informaon Centre for Economics (ZBW) is a public research ins-

tute and has a high-tech informaon infrastructure, conducng research in

computer science and related areas and develops technologies for oper-

ang the IT infrastructure for its own library, which is the world’s largest

specialist library for economics. ZBW parcipates in this project through

The research group Science2.0 which focus its research on Big Data and Science 2.0.

hp://www.zbw.eu/

The University of Manchester (UMAN) is the largest single-site universi-

ty in the UK, with the biggest student community. UMAN parcipates in

this project through the Bio-Health Informacs Group which carries out

research on the development and applicaon of complex artefacts such as ontologies for the health

and life sciences. hp://www.manchester.ac.uk/

P A G E 6 I S S U E 1

MOVING consortium - who we are

GESIS – Leibniz-Instute for the Social Sciencess (GESIS) is the largest

German infrastructure instute for the social sciences. GESIS parcipates

in this project through the Knowledge Technologies for the Social Scienc-

es (WTS) which is focused on advancing and improving digital services for

the Social Sciences. WTS is carrying out research in the elds Web Science, Semanc Web, Linked

Open Data and Informaon Retrieval. hp://www.gesis.org/en/instute/

Fundacja Progress and Business (PBF) is an independent research instuon and a

leader in implemenng best research and consultancy pracces in Poland. PBF par-

cipates in this project through the Internaonal Centre for Decision Sciences and

Forecasng (ICDSF) which is specialized unit of the PBF that organizes and carries

out research on the foundaons of Arcial Intelligence and cognive sciences, decision support sys-

tems, mulcriteria decision analysis, informaon retrieval and data mining for forecasng and fore-

sight and discrete-event systems. hp://www.pbf.pl

Project Coordinator: Dr. Vasileios Mezaris

Address: Informaon Technologies Instute (ITI),

Centre for Research and Technology Hellas

(CERTH) 6th Km Charilaou - Thermi Road, 57001

Thermi - Thessaloniki, Greece

Fax: +302310 474128

Tel: +302311 257770

E-mail: bmezaris@i.gr

Full Title: “TraininG towards a society of data-

saVvy informaon prOfessionals to enable open

leadership INnovaon”

Project Idener: H2020 - 693092

Start Date: 1st April 2016

End Date: 31st March 2019

Duraon: 36 months

Contact Details

Project Details

This project has received funding from the European

Union’s Horizon 2020 research and innovaon pro-

gramme under grant agreement No 693092