About the Workshop

There is no such thing as simple text

Workshop description and aim

Finding knowledge and information on the web is becoming an almost trivial task, understanding its value, the credibility and perspective of its source, on the other hand, is becoming increasingly problematic. While the Internet and other digital modes of communication have made access to knowledge and information much easier, the burden on the users of the Web to filter and judge information has increased. Technology lags behind when it comes to providing users with proper tools to track information provenance, the perspective of the source, and the validity of information. It is a common saying that on the Web you can find any truth that is convenient, yet we have little knowledge about how beliefs actually spread and how this impacts our view on the world. Many questions with respect to the impact of being online will remain unanswered unless we come to grips with these social communication processes. When do online discussions in distributive communities lead to convergence or to polarization? How does social filtering work and does it change as result of online debates? What is the impact on identity and group membership when dealing with convergent or hostile beliefs?

Besides the fact that, nowadays, we can track many social activities on the web and analyze what people ‘talk about’, the language that people use is in itself a source of information about the attitudes and perspectives that people hold. There is no such thing as simple text. We often express our emotions, opinions and positions verbally in subtle ways that can sometimes be made explicit more easily by computer software than by human observers. Automatic mining of subjectivity and perspective relations from language could be exploited to learn more on the impact of communication on the Web. The availability of digital data at all these different levels opens up the unique opportunity to dynamically model information, knowledge and mediated communication as well as their social implications. This can be done from the lowest micro-level of symbolic data (such as texts), to interpretation and knowledge, up to the higher macro-levels of (1) social groups and identity, (2) social activity and dynamics, and (3) societal impact.

The proposed workshop aims to set the research agenda for an interdisciplinary approach to model information, knowledge and mediated communication from the micro-level up to the macro-level. Our vision is that such integrated models will allow us to: a) develop technologies that make the spreading of and perspective on knowledge explicit and transparent, b) measure the impact communication have on our world views and beliefs as shown by social activities and measurable choices, and c) increase people’s awareness of the tacit meanings that knowledge and information may carry. 

Can we build the Perspective Web or Web5.0?

The theme and guiding concept of this workshop is the question if we can build the Perspective Web or Web5.0. The Perspective Web (or Web5.0, the emotional interaction web) would be a projection on Web1.0, 2.0 and 3.0 content that makes explicit and transparent the content and perspective of the sources of the content according to the standards of the semantic web. It allows for eliciting the dynamics of debates, opinions and views on a global scale. While Web1,2,3 are based on links, Web4 will reorganise this and connect the knowledge through sources and the content, showing the range of perspectives.

Tim Berners Lee proposed a system of 5 stars to qualify Open Data: http://5stardata.info/en/. In this workshop, we would like to explore if we can come up with a similar star system to qualify the perspective of data. This project site is used to start the discussion on the perspective stars. You can contribute by send a discussion blog to the organisers. We will post the blog on this site  under the menu Perspective Stars. You can comment to blogs of other people or start your own thread to address any issue or aspect of the Perspective Web and the stars.

Furthermore, we will provide a data set and demos on debates on vaccination collected from the web. This data set can be used for demos and make the discussion concrete. Here you can find a slide presentation that explains how the Perspective Web could help to clarify the debate on vaccination: ConnectWorldKeyNote2016.

Workshop organisers:

Piek Vossen (piek.vossen@vu.nl)

Julia Noordegraaf (j.j.noordegraaf@uva.nl)

Ivar Vermeulen (i.e.vermeulen@vu.nl)

Lora Aroyo (lora.aroyo@vu.nl) 

Chris Welty (cawelty@gmail.com)

Antske Fokkens (antske.fokkens@vu.nl) 

Roser Morante (r.morantevallejo@vu.nl) 

Chantal van Son (c.m.van.son@vu.nl)

Isa Maks (isa.maks@vu.nl)

Other information

Public websitehttp://www.understandinglanguagebymachines.org/events/lkpp/

Lorentz Center websitehttps://www.lorentzcenter.nl/lc/web/2017/878/info.php3?wsid=878&venue=Snellius