10 Jul 2015

Share this

Video of the 2nd GIPO webinar on key findings from Sofia workshop

By: Kasia Jakimowicz

On the 1st July GIPO team discussed the key findings from live workshop that took place in Sofia during EuroDIG2015 conference.

As the results of the working sessions and webinar, the next steps for GIPO platform were outlined:

Fostering user engagement

Defining common Internet Governance taxonomy

Increasing sources of information

Testing and adding more semantic analysis service

Improving accuracy of filtering

Implementing best content sharing standards

 

The discussion followed the outline of the workshop and was divided into three subsections. 

1) Synergies between GIPO and other initiatives - How to ensure interoperability of mapping initiatives through common taxonomies

Taxonomy is considered as a basis to organize data set. As the taxonomy is bigger and more general, tagging comes as the way to slice date further and enables more detailed search.Slicing of the data set and its naming in a proper way is crucial, especially when the interaction between the sources is one of the goals. Though there is a need for a common dialigue and a universal way of describing datasets across initiatives and in this way creating a sharing environmentNevertheless, taxonomies should be dynamic and flexible to some extent to be able to incorporate new toics as well as to ensure some flexibility in interoperability with other initiatives.

During the working session the need to engage community in giving comments on how to structure the information has been identified. The joint recommendation was to organise the GIPO brainstorming session to start the process with all the mappers and observatories being involved
 

2) How should the social media data be analyzed and presented for informing policy debates 

The first challenge is to actually get access to social media data. It's not easy to get access to social media sources in their entirety. Facebook is particularly challenging, Twitter is easier but the public API is not complete. In many cases it is necessary to use third parties clearinghouses that have a cost.

Once we have access, the second issue is how to organise knowledge. The challenge here is to identify the sweet spot between top down taxonomy and bottom up user generated tags (and hashtags).

For GIPO, social media would not so much be useful for opinion mining purposes, but for identifying interesting sources. Social media could help identifying policy positions and relevant statements to widen the sources covered by GIPO. Advanced analytics could provide interesting insight about some measures and policy statements that are very popular on social media but that are totally absent from official sources, thereby identifying some potential hot and conflictual issues before they become untreatable. Another interesting application of social media is related to the identification of hot issues across time, helping to identify sudden viral emergence of attention over a specific topic / issues. This could help identifying important topics in real time and be a useful service to stakeholders.

 

3) Text analytics and semantic data extraction -  How to get the most from semantic analysis services

Two distinct problems for GIPO were identified in this area:

- Location of desired (“correct”) elements (detection).

Classification of the elements (by the type of entity they refer to e.g. person, organization, location, etc.) and linking them.

IT tools still does not fully reach human semantic processing (e.g. not well in areas like sentiment analysis). Semantic tools are undoubtedly useful, but need to be complemented with feedback from users. GIPO needs users engagement in redesigning algorithms (filters), confirming the taxonomy and curating contents (or at least voting them).

 

 

The calendar of future GIPO events
 

Comments

Add comment

Image CAPTCHA

More Blogs