A plethora of data formats containing tales of daily life: bank statements, event information, p2p messages, wish lists, video, image, audio, notes, preferences, and statistics. In the current universe of data, each interface incurs it's own specific data format. Though open formats and microformats are encouraging reinterpretation of previously locked in data, they remain separate categories; both in access and in structure. By defining a new common structure based on human thought processes, this paper will provide a means for non-linear semantic data traversing, and multiple interface use.
Current tools for accessing diverse information formats are explored with the intent to recreate/rethink the default interface. A method is then introduced which amalgamates the prior methods through a common approach. The challenges of multi-authorship and privacy are discussed in relation to these methods and application. Finally, examples of visualization under this common structure are explained.
general to today A plethora of data formats encapsulate daily life: bank statements, event information, p2p messages, wish lists, video, image, audio, notes, preferences, statistics, etc...
specific solution of today In the current universe of data, each interface infers it's own specific data format. problem of current state Though open formats and microformats are encouraging reinterpretation of previously locked in data, they remain separate categories; both in access and in structure. solution proposal By defining a new common structure based on human thought, this paper will provide a means for non-linear semantic data traversing, and multiple application of interface use.
Contents examine current apis & attempts at combination Current tools for accessing diverse information formats are explored with the intent to recreate/rethink the standard interface.
Introduce truthstructures as a tool/method A method is then introduced which amalgamates the prior formats into a common structure. The challenges of multi-authorship and privacy are discussed in relation to this method and application.
Examples Finally, examples of amalgamation and visualization under this common structure are explained.
social shared information There is no debate about the benefit of sharing information. The real hang up in the discussion of knowledge sharing is how to do it most efficiently. Other concerns such as privacy and intellectual property affect the culture of sharing, but surely in time those concerns will disappear in the face of greater benefits. Knowledge is most valuable when applied to a problem or directed to the right person to analyze and draw insights. Much of the knowledge in the information world is redundant, usually unintentionally. The scientific profession has evolved into a network of working bodies, aware of each others work, and ideally working on complementary problems. In the new paradigm of networked thought, ideas cannot not constrain themselves to linear text documents. We must have a new method of thought sharing, one that takes into consideration the overlap of disciplines and concepts. We need a system of sharing where information is not stacked but rather melded. I should be able to open 100 text documents, and instead of having 100 pages to read, I now have 500 relations on one page, with shared concepts not repeated.
non-destructive recording In the arena of shared information, revisions become the tool for a self sustaining network. Wikipedia has shown us that many others in different places can maintain a vast directory of information. Central to the systems of networked knowledge is the ability to revise, edit, and create without the explicit consent of others. Along with such a system comes a history of edits, which within time becomes a very long list. What we need in the future is a format of non-destructive data. Any information should contains it's own lifetime. This life time would contain changes, updates, and offspring. Non-destructive data can easily be viewed from different time periods and perspectives. This type of data needs a central focus on connection between things rather than a document which gets longer, this document gets denser.
designers exchangeable interface Even now, one can see a plethora of diverse data formats. These formats encode information: bank statements, event information, p2p messages, wish lists, video, image, audio, notes, preferences, statistics, etc... Due to different type of applications, these formats have traditionally been defined by interface requirements. There are thousands of different format standards because there are thousands of interfaces. Due to the internet, we are seeing an integration of interfaces. We can now view RSS on a calendar<calendar.google.com>, real-estate on a satellite map<housingmaps.com>, bookmarks through a node web<http://www.hubmed.org/touchgraphs/deltags.php?start=visualization>, and audio through lines and colours<itunes.com>. Though interfaces are being adapted to data formats, data formats increasing diverge, making it more difficult to apply new interfaces to existing data.
access through standard format A streamline of data formats would democratize the interface construction process as well as build new opportunities for interface reusability. Truthstructures proposes a semantic markup with 2 basic components: Concepts and relations. Using these basic properties, we can describe the relation of information in an all encompassing system. Forks are detrimental to progress, they signify a change of focus and large amounts of commonality are lost for the sake of very minor differences of opinion. Conversation can only be productive when both parties are using the same terms. Defining terms has become a profession for many academics, since it must be done before proposing any new insights. Truthstructures would allow non-destructive data manipulation through relations, rather than the copying and changing methods of today's technology.
unending traversing of structure The common metaphor of computer operating systems has been files and folders. The problem with such a systems is it's reliance on hierarchy of information. The new method of information structures should be an interconnected collection of concepts and relations. This web of information should be able to be navigated infinitely, much like the world wide web. Though the web is a non-linear experience, it remains a non-linear collection of linear information. The benchmark of a non-linear information structure, is that meaning is created by the very act of moving around the network. Each jump from one concept to another through, a relation, is a statement. Meaning is created through a history of information paths.
visualize concepts & relations A non-linear structure of data should be an easy to visualize system. Designers of all kinds should be able to interpret, and reinterpret the structure of the data through analytical design. This technology should be second nature to the way we as humans thing. Concepts are connects, and it's is the value of their relations which is visualized through various design practices.
users collaborative data creation Once we establish a non-linear method of editing, it is easy to see the applications to collaborative development of such structures. Multiple authors can simultaneously be adding and editing the same structure of information, fully aware of the other's actions. We being to work on a unified mass of observation, rather than diversified linear formats.
overlaying/comparable data Truly one of the great benefits of a shared format is it's ability to be compared. Imagine being able to overlay your thoughts upon another person's truthstructure. You would immediately be aware(on a qualitative level) of conflicts, shared ideas, and new ideas which you have not considered. This qualitative information(due to the large scale) could be further investigated by delving into the deeper meanings of the connections of question.
decentralized ownership of data Bit-torrent has shown us the engineering value of information being shared across multiple platforms. Beyond just a quicker and persistent storage method, shared records of the same events or idea, maintains integrity in the nature of the data. Records are very difficult to remove when they are distributed among a number of viewers.
popular acceptance of truth system Just as in society, truths are subject to scientific scrutiny, and accepted as valid by through a rating system. Ideas on the web are popularized by linking to them; they spread if people feel strongly enough about them. A system of truth recording should take this into account and be able to distinguish widely held beliefs from fringe opinion. It should be able to project quickly growing ideas as well as trends which quickly drop in popular consciousness.
portable format A key to the holistic system of recording truths, is the ability to carry that information from one context to another. A personal information cloud should follow the person from the grocery store, to the bank, to work, and to sleep.
semantic non-linear language Extensible Markup Language has been a success largely due to the fact it is a semantically meaningful format which is also readable by software programs. The value of semantic encoding is that it allows for understanding of new formats by developers, as well as an emergence of functionality through creation of automated neural networks. The web is moving to a more semantic structure through XHTML and XML, but it still remains a collection of linear documents. We need a nonlinear, non-redundant method of encoding, which is semantically useful, but also distributed among perspectives.
semantic relations Relations between concepts are verbs. They are actions which have occurred at a specific time between two concepts. These are semantic because the relation between things is labeled as an action, they are useful to humans because the verb/noun paradigm has existed for a long time.
shared focus(talking about the same thing) Many problems, especially in the political sphere, have come about because of purposeful distortion of issues. Framing is a popular way to place concepts in a specific perspective which is advantages to the advocates agenda. Many times, words are the source of confusion because they carry meanings along with them. A good example is the difference between a state and it's people. Both can be referred to as a country, the USA for example. If I say the USA is corrupt it could be interpreted in many ways. A concept is nothing without the relations that connect it to it's conceptual sphere. By defining relations before discussing a concept we can avoid double meanings and framing of issues.
storing of actual observations rather than interpretations The driving force behind truthstructures is it's ability to record observation and not interpretations. When information is recorded digitally, it is usually interpreted and therefore stored in an interpreted format. For example, if a piece of content is rated by 5 people as 1,3,5,4, and 6, it is stored as 3.8, an average between numbers. An observational based record would store each rating as a time-stamped occurrence. The difference between interpretative recording and observational recording is in the storage of information. Interpretive recording makes assumptions about the information, it slims it down and then records it's interpretations. Observational recording takes raw time-based observations and allows interpretation to occur at the interface level. The advantage of Interpretational recording is it's small size, the advantage of observational recording is it's ability to be reapplied to many different contexts, and new connections to be made.
Spread Personal Applications Thomas Vander Wal http://www.personalinfocloud.com/ has already begun to investigate ways to merge the growing cloud of information that surrounds our lifestyle. Truthstructures is a perfect match for the personal information cloud, it's format is very complementary to the varied range of information in our clouds. It is clear that the average person will not adapt truthstructures as the new organizational environment. A very important principle of truthstructures is it's need to be available through many contexts and interfaces. Applications can feed into truthstructures and display the relations and concepts in many different ways.
Calendars Calendars are by nature a recording of event based information. These events are relations between a person or persons, and a specific place. The relations in these calendars lend themselves to being stored as event based information, but also place information. Some information is better entered into calendars, others into time-sheets, others into maps. All of this information should draw from the same base but be interpreted by different interfaces. Communications There is a rise in communication products designed for specific purposes. For an event I send an invite, or an image I share an image page, for an article I click on the article link, for a product I send a recommendation. Up until recently, the prominent method of communication has been email. All of these different message types have pushed themselves into a text based format with external links. These messages that are being sent are really just representations of occurrences, or truths. I recommend something to someone, I created an event and add a person to the guest list, these are statements of truth. I should be able to accept or reject truths. Instead of exchanging disconnected messages, we begin to work on a holistic data cloud. The communication comes from the creation of truth.
Surveys A large need in today's economy is one of consumer opinion. Survey applications are traditionally done as a form of questions. The answers are then stored in a database. Once the information is collected, it is analyzed and then kept to the research group as a valuable private asset. This preference information is repeated over and over, every time I fill out a survey I am answering questions which are redundant due to my previous surveys. The main problem with surveys is that the information is stored by the researcher and not by the participant. Researchers should be able to propose questions, but the information should be stored on the user's end. This would allow researchers to request access to consumer information. It would cut down on the workload of research, the researcher's would know what questions have already been asked, and it would put the power in the hands of the person of question.
Structure -concepts(id,type,title,description) -relations(from, to,time,duration) -truths(believer concept,belief concept)
Examples -undecided -truthbuilder -hive
Traditional systems take actual events and record them as abstract(person likes chocolate, chocolate=1, 1 star)
This system Records actual events and the interface interprets them into abstract(person likes chocolate, person-likes->chocolate, 1 star)