Reflecting News in Digital Media

In contrast to traditional media such as newspapers or television, digital media like the World Wide Web are generally dynamic. Here, news is no longer bound to a specific form or context. This separation of content and presentation poses challenges and opportunities at the same time.
The formal presentation is not necessarily able to synthesize with the content in order to support its communication.

This work tries to answer the question, how a preliminary linguistic analysis can contribute to finding adequate presentational forms that relate to each news’ specific content.

The formal presentation plays a crucial role in effectively communicating a given content. To assist the reader's comprehension, form and content should be mutually dependent. However, in current media products, the form mostly serves the purpose of organizing and distributing content. This work explores possible forms of presenting news, that relate to their content and therefore improve the readers ability to select and process news and complex contexts.

If the content is no longer embedded in a defined visual composition and context, new design strategies are needed. The designer has to think in classes, not instances: not the singular visual form is important — instead one has to design systems and principles that can then generate possible solutions.

However, if form and content are supposed to relate to each other, a system has to be able to comprehend the content to some degree. Only then, a design system can adjust formal parameters accordingly. Here, a lingustic analysis — Natural Language Processing — can be of great help. The results of such analysis provide good insight into each news specific content and structure. Therefore, a design system can greatly benefit from such a lingustic analysis and the resulting set of rich meta data.

The work is not so much about providing a specific solution or product but more about finding ways to create and evaluate a design system that can then have positive influence in many situations and scenarios.

Therefore, an environment was created, that allows to work with actual news content and the corresponding rich meta data. This set up was developed using the Reuters Spotlight API and the Daylife API. A PHP application then processed the data and allowed the quick prototyping of innumerable variations and interactive prototypes. These were realised using standard web technologies such as XHTML, CSS and Javascript.

Some of the most promising approaches were then combined and made accessible through a working prototype that runs on a touch screen device.

↑ Various visualizations of news content: is it possible, to find a specific and reproduceable formal representation for a given news item? Can we provide visual cues, that improve our selection of relevant content? Some of the visual cues or visualizations depicted above, allow us to predict the content and structure of a news item: one can see if the article is long or short, complex and broad or rather narrow, etc.

↑ A preliminary linguistic analysis allows for interesting new approaches in providing contextual and context aware information. Here, various information on a given entity is pulled from the Daylife service. Also, the extracted entities serve as filter: when clicked, all items containing the selected entity are highlighted, other articles are dimmed. By modifying the selection, new relations can be discovered.

↑ Not only during selection, also during the processing of news content, these visual cues accomodate the readers need to quickly scan and skim texts and serve as anchors and entry points into the text.

↑ The news items are here organized chronologically — which makes particularly sense in the case of news from the wire services (in this case: Reuters, accessed through the Spotlight API) — and can be flipped through, even on the level of processing.

This work was created as diploma thesis in the interdisciplinary design programme, field of Interface Design, at Köln International School of Design, Cologne, Germany. More information (and a working prototype, that can be experienced in a web browser) will soon be available from news.svenellingen.com.