Story Abstracts
Today, news, articles and encyclopedia entries are authored using plain text. These texts might link to other articles and can be indexed in order to be found in searches, but semantically they are stand-alone. Computers do not understand the meaning of the texts, they cannot automatically link one article to related articles. People, events and entities that make up the stories cannot automatically be connected to related information. Computers cannot automatically analyze the information the texts contain, they cannot automatically put events into wider context.
The more complex a topic is, the harder it is to follow, especially if information are served as small bites in form of news. Authors need to build relations and give context within texts for readers not familiar with the topics, additional work that is often not justifiable.
Authors writing stories often need to understand complex situations, consider many little details, remember many entities, relations and sequences. Today, there is no tool available that would help them to design the draft for a story in the same way as for example an engineer can create a drawing using CAD.
Every story or article contains information that can be abstracted. The abstract is the raw information the story is made of: Facts about persons and entities. The relationships between them and the events they are involved in. The temporal sequence of these events. And so on. Every person, entity and event can be abstracted and put into a multi-dimensional structure that forms a story abstract.
Using information modelling, Havel allows authors to freely create an abstract of a story, article, event or situation. Like any other information modelled in Havel, the story abstract is an expression. This expression can contain qualifiable and quantifiable information, structures representing entities inside the story, relations between these entities, significant events including their temporal sequence, and so on. The expression is the information the story contains that – to a certain degree – can be understood by a computer and therefore be automatically analyzed, interpreted and merged with other information. The expressions are modelled using an IME or NME.
All the parts in the story abstract are semantically linked into a much wider infoverse. And vice-versa, the story becomes automatically part of that infoverse. If, for example, a story is referring to a person of public interest, all information about that person previously part of the infoverse are available to the author and reader. This mechanism goes much further than just linking an article, other articles (or general information sources) become automatically part of the new article – all sources merge to form a wider picture.
The text itself (or multiple versions of the text, e.g. a full story and an abbreviated version, different language versions) is joined with the story abstract. The presentation to the reader is a mix between text and visualization of the structure as graphs, charts and tables. The visualization can follow the reader’s progress through the story, becoming more complex as the reader progresses.
At the beginning, these story abstracts will be relatively rough, focusing on the essence of the story. Over time, as Havel itself will progress and will allow for more subtle expressions, closing in with human language abilities, the abstracts can also become more detailed and subtle. With later Havel generations it is conceivable that articles are authored by modelling the abstracts and having them automatically translated into multiple human languages. At some point it will be possible to translate the abstracts dynamically not only into any human language but also depending on the reader itself, considering his background, profession, known knowledge, etc. Computers will be able to tell us a story in a similar way as humans can, they will know what we know and dynamically adapt.
This post is part of a comprehensive introduction into SRDM, graph languages and Havel.
Continue reading: Privacy and Data Sovereignty
Comments