Esta es la versión HTML de un fichero adjunto a una solicitud de acceso a la información 'Documents and statistics on the EMM Open Source Intelligence Suite'.


Ref. Ares(2016)5296138 - 14/09/2016
Commission of the European Communities 
Date: 15.06.2016 
Joint Research Centre - Ispra Establishment 
Mission No:  DF-16-1221405 
 
   
       
 
 
Mission Report 
 
 
 
 
                    
Name 
Division/Action/Project 
 
  
GlobeSec/EMM 
 
 

Object of the mission 
 
 
The objective of the mission was to deliver an invited talk at the Deggendorf Forum for Digital 
Information Analysis
 (DFDDA; http://www.dfdda.de) and to attend the yearly conference. 
 
 
 
Place, date, organizations visited and contacted persons 
(Please continue on a separate sheet, if necessary) 
Deggendorf Forum zur Digitalen Datenanalyse 
Garching, Germany, 9-10 June 2016 
Institute for Advanced Study, Technical University Munich. 
 
 
 
Other Participants 
 
Distribution List 
Of the establishment 
Other services of the commission
Obligatory(*) + 

 
 
 
  (*), JRC 
 (*), JRC 
 (*), JRC 
xxxxxxxxxxxxxxxxxxxx@xxx.xx.xxxxxx.xx  (*), JRC 
GlobeSec Unit, JRC xxxxxxxxxxxxx@xxx.xx 
 
 
 
Summary 
 
The objective of the mission was to deliver an invited talk at the Deggendorf Forum for Digital 
Information Analysis
 (DFDDA; http://www.dfdda.de) and to attend the yearly conference, which took 
place in Garching, near Munich, Germany. DFDDA mostly targets auditors and it includes new 
technologies that may become important for their work. For the program of the event, see 
http://www.dfdda.de/dfdda2016/dfdda20160.html.  
 
The one-hour JRC presentation was one of seven presentations at the forum. It had the title ’Supporting 
the analyst by turning multilingual text into structured data’. In the presentation, I gave an overview of 
EMM and OSINT Tools, highlighting the benefits of large-scale data analysis, showing various ways 
to visualise data, and making users aware of the dangers of fully-automatic analysis. I aimed to address 
the specific needs of auditors and I tried to show what an EMM-like analysis can do for this type of 
users and what it cannot do. Finally, I presented the multilingual linguistic resources created as by-
products of EMM and downloadable from the JRC’s Science Hub pages, as well as from the EU Open 
Data Portal, with a special focus on JRC-Names and the JRC Eurovoc Indexer JEX. 
 
The talk was received extremely well. There were many questions regarding functionality, availability 
of the software and the possible usage of the linguistic resources. 
 
The DFDDA event was useful to get a good insight into the needs of auditors. The presentation I found 
most useful for our own purposes was the talk by 
 from Passau 
University (Germany), who talked about combining formal and distributional semantics to represent 

words and texts, showing the benefit for information retrieval and question answering. While in formal 
semantics, meaning aspects of concepts are intellectually identified and described (e.g. a ‘dog’ has a 
name, a colour, a race, a size, etc.), the individual values for each of these meaning aspects can be 
found in large amounts of texts, using distributional methods. The commonalities and the differences 
between the concepts ‘apple’ and ‘pear’ can thus be identified, for instance, by searching which 
adjectives and which verbs (subject of, object of) are found with the equivalent words. 
 also 
presented word clouds around the country name ‘Iraq’ (associated words) and how they changed over 
time. This latter kind of observation could also be useful as an analysis aspect to be displayed in EMM. 
 
 
, 15 June 2016