The disadvantages and advantages of the informee

This is sort-of part 2 of my post titled “data, information, knowledge, and intelligence“.

The following graphic is a simplified process-of-processes of informer to informee.

Firstly, this is a depiction of super-processes for knowledge transfer as I understand it. It is simplified because the network of networks of networks that makeup how we attain our knowledge is far more complex. And it is that complexity that I wish to understand in detail by analyzing the relationships of each of the open-loop systems that make up each of the above boxes: data, information, knowledge, and intelligence, both from the perspective of an informer, an informee, and from an artificial intelligence that is programmed to perform an array specific functions.

With respect to my last post on understanding information via multidimensional network analysis, data, information, knowledge and intelligence would make up one entire dimension. However, as clearly as it is depicted in the above diagram, each of these are also their own respective super-processes, and in my opinion, are of their own dimension because of the dramatic differences in how they independently and, at times, dependently influence each other. They do not always exist together and certainly never in the same order of process.

It’s in my opinion that because we are so dependent on the meaning and intent of the information that we consume, we are not aware of the vast complexity in the alternative: an entirely different network of information, with different structure, and intelligently designed to accomplish a more precise task.

For example, one will naturally consume a great deal of information in a mixture of personal volition and uncontrollable influence. They are not mutually exclusive when an information organism has the indefinite consequence of orchestrating a continually developing network of personalized information (aka: knowledge). What if you could increase the probability of controlling the information that you consume if you could automate and construct the networks of information that you process?

Arguably, this is what the Homo sapien has been doing throughout evolution. When we choose a major area of study at university, we are increasing the probability of extending a specific network of information in our brain. When we choose to more or less consume information from a specific news source, we are using operational information, information about how an information system operates, to influence the exposure rate of an expected network of information (of varying classification).

Secondly, looking at the above diagram, one might visually notice that the informer has the control over the larger part of an information-transfer system. However unfortunate for the control-bent informer, with the advent of the Internet and an increasing spectrum of information processing organizations such as Wikileaks, information consumers are able to more selectively increase the probability of building their knowledge according to their personal volition. Journalists and intelligence analysts, on the other hand, should have the objective of increasing an informee’s network of information. It is in the interest of both parties to strategically grow an informee’s network of information, whether it becomes solidified knowledge or not.

Each section from my above Venn diagram could be observed as a dimension to control analysis. Each section will have vastly different effects on the other sections depending on the situation. From my earlier post I stated: “Data directly affects information. Information does not directly affect data. The transformation of data to information is an open-loop system.” The above diagram is where that came from. Each box is an open-loop system because they require either the input, output, or both of another system in order to have an observable behavior. It should be obvious that that system (the network) that is data affects the system (the network) that is information. When data is gathered and organized, it is done so with the goal of providing specific information to one who wishes to demonstrate expected information. And using that same example, it becomes clear that without having previous information, in order to have expectations, information can only affect data after it is processed by the other systems that are knowledge and intelligence.

Before going further into the explanation of these super-processes, I need to touch on the importance of analyzing the actual processes that make up these super-processes. And it will not be without understanding the sub-processes, not to mention the relationships between a super-process, it’s processes, and it’s sub-processes; aka: a systemic analysis of multiple complex systems, all having expectedly different characteristics when observed in different situations.

A process that makes up the super-process that is the system of information includes a single information classification. Floridi’s “primary information” is of it’s own network of complexity depending on the dimension for which it is observed. Primary information exists objectively not only to people, but also to other classifications of information. When primary information has a relationship with other primary information or a different classification such as secondary information, it may behave differently to the informer and to the informee, both being distinct dimensions of complexity. This growth in complexity is expected to exponentially increase as we add other processes, other information classifications, thus increasing the importance to understand this system systemically in order to design artificially-intelligent systems.

In another network we have data classifications. The notions of “information classifications” also apply to data classifications. Primary data, secondary data, meta data, operational data and derivative data. Each of these processes partially make up the super-process and system that is “data”. The complexity just increased, and we haven’t even gotten into the various types of information, including Floridi’s “biological information” — information as reality, information for reality, and information about reality. Each of these dimensions will affect the processes that allow for knowledge transfer, especially when we design our information systems to purposefully affect the semantic information that we consume to grow, change, or destroy our information networks.

Additional processes that make up the super-process that is “information” include the nature of misinformation and disinformation, relative to the classifications of each, and relative to the previously explained complexity of the five information classifications.

To continue explaining the process of informer to informee, “Knowledge affects intelligence. Intelligence affects knowledge. The transformation of knowledge to intelligence and intelligence to knowledge is an open-loop system.” Here we have two open-loop systems that both affect each other, creating a particularly interesting system in and of itself. I specified that my understanding of knowledge, relative to my interest with information, as being “memory dependent, networked information”. Knowledge exists in my brain because I am able to sense various kinds of information from my world, and it becomes networked with previously processed information, thus growing, changing, or destroying previously existing networks. And of course, all of this is possible because we have a place to store this information.

Intelligence, on the other hand, includes what one actively does with their consumed information. Not being a cognitive biologist (just a wannabe cognitive neuro-scientist), while I have a genuine interest in the systems that are knowledge and intelligence, I am only concerned with them insofar as their direct abilities to affect the behavior of information. Intelligence may include the biological processes that manages our consumed information, and also our retrospective activity that affects the further retrieval and design of information systems and its influence on collecting and organizing specific data.

Depending on where each of these four super-processes exist, whether in control of the informer or in control of the informee, their behavior will change in different information-processing situations. It is these behaviors that I wish to understand in order to design the high-level processes that will dictate how artificial intelligence will be programmed.

Advertisements

Exploring a Multidimensional Network of Information

There are so many different aspects of information that are so incredibly fascinating. I am hoping to start giving my thoughts some better structure by implementing network theory as my high-level information manager. Being so high level, I expect it to remain largely theoretical, all while the goal is to pull out a few ways to control and predict specific aspects of semantic information (meaningful metrics). The purpose is to generate alternative ways to understand the semantic information that we encounter online.

Goal #1: Apply network theory to multiple networks and multiple dimensions of information in an attempt to articulate the multidimensionality of information networks.

I knew that the idea of “multidimensionality” had to exist. Thank you Dr. Karine Nahon for tweeting about it.

“…a fully multidimensional network is one that includes multiple types of relations both among the same types of nodes and between different types of nodes. Thus, a fully multidimensional network has multiple types of connections among all possible types of entities.”

Continuing my lust for Dr. Luciano Floridi’s work on the philosophy of information, there are two obvious dimensions of information that I am particularly interested in: the five information classifications and the three aspects of biological information.

Within those two distinct yet completely intertwined dimensions, the nodes, I expect, will become quite complex. I also expect to apply aspects of complex adaptive theory onto, particularly, the aspects of biological information, primarily because complex adaptive systems are used to understand the relationships between social networks.

I expect to apply various aspects of systems science onto the five information classifications. Each of these classifications and my expected sub-classifications of information will makeup nodes; there will be sub-nodes where data that I collect from my research will be plugged in, with the objective of observing the behavior of the multidimensional network.

The objects that I plan on including into this network include specific pieces of semantic media and the people who consume them. I wish to perform analysis on different styles of semantic information. One style will be Wikipedia articles, and another focus will include political news articles. Each of the respective articles will be juxtaposed with the same source of articles, and later I will include additional sources to diversify the research. Eventually I plan on analyzing multiple types of semantic information sources which will greatly complicate the the multidimensional network.

In order to increase the complexity of this multidimensional network, people will also be part of my research. I will juxtapose two additional dimensions: one being the relationships between people and the semantic media that they consume. The second will be the documented retrospective analysis of how this multidimensional network categorizes the same semantic information.

In retrospect I want to further elaborate on relationships, taking care of my applied notion of complexity.

Dimensions:

1. information classifications: As independent yet open-loop systems each existing as a node; but as I mentioned, each of these nodes is likely to include multiple sub-nodes.

1.1. general types as they objectively exist and can be meaningfully applied to an information system.

1.2. biologically inferred types of information as they relate to biological organisms with perception and memory.

1.3. see 2.3 below

2. people: As independent yet open-loop systems each existing as a node. Most of my expected data will come from the relationships between information classifications and how they affect people.

2.1 The informee (one who is being informed) will have applied aspects of information retrieval. I expect people to have varying degrees of knowledge about a specific topic as it relates to the piece of semantic media that they consume. I expect that information classifications, especially their relative and perceived value, will change according to their A) prior knowledge and B) expected knowledge.

2.2. The informer (one who is intelligently crafting information from data and/or information).

2.3. Information, misinformation, and disinformation: A distinct sub-project of my work will include the analysis of these three aspects of persuading people. Contrary to communication science, misinformation and/or disinformation are perceived as information until said information can be retrospectively compared to a different yet similar network of information, misinformation, and/or disinformation. I also expect the behavior of the five information classifications to greatly change according to the use of misinformation and disinformation. I further expect, that instead of information, misinformation, and disinformation existing as distinct concepts, that they actually exist on a multidimensional spectrum, much like the lines of species of animals is blurred because of their distinctive evolution.

3. data as it relates to information, particularly how data affects information but information’s inability to affect data (directly). I expect to find that information can only affect data once information is understood by an information consumer who is then able to intelligently interpret and organize data.

4. information entropy

Each of the above dimensions, also being composed of specific and different aspects of each dimension, will have distinct relationships with each of the other dimensions, including each of the nodes (or objects) and sub-nodes of each of the dimensions. As you may presume, this will quickly become very complex, and complex in different ways (sorry for being vague here–more expectations and more to explore!).

Each of these relationships will aim to answer many of my questions about information.

Experimenting with personal storage encryption

Lately I’ve been playing around with different ways of encrypting my PC for fun, but not in ways that require a lot of time to get going. I am pretty comfortable with my current setup. All passwords are at a minimum 32-bits and complicated.

1. Instal Windows 7 Enterprise or Ultimate on a PC with a TPM chip
2. Encrypt with Microsoft BitLocker
3. Set BIOS and boot passwords
4. Install TrueCrypt and create two xxGB partitions, mount them both
5. Install VirtualBox and create an Ubuntu VM. Store the virtual-image file in one of the TrueCrypt containers
6. Move / link all of your user account folders (desktop, downloads, pictures, videos, documents, etc) onto the other TrueCrypt container

Each time you boot in to Windows you’ll need to mount the TrueCrypt container with your account’s folders, then log out / log in for Windows to link them properly. For someone that frequently re-installs Windows, this setup makes it easy to keep my files backed up and encrypted. Oh, and seamless mode in VirtualBox is pretty cool.

Happy hacking

On the Nature of Disinformation

From Star Trek: Deep Space Nine

Julian– You know, I still have a lot of questions to ask you about your past.
Garak– I’ve given you all the answers I’m capable of.
Julian– You’ve given me answers alright. But they were all different. But what I want to know is out of all the stories you told me, which ones were true and which ones weren’t?
Garak– My dear doctor, they’re all true.
Julian– Even the lies?
Garak– Especially the lies.

This is an ongoing quest. I have quite an extensive library of books concerning information theory yet only one of them, Encyclopedia of Information Science and Technology, 2nd Ed., mentions the word “disinformation.” But it doesn’t even bother describing it.

Dr. Luciano Floridi talks about the philosophy of misinformation and disinformation. He (among others before him, I presume, not looking in his books at the moment) includes discussing the logic-based fact that disinformation is not actual information, (I oversimplify) because it is not informative.

While I agree with this point of view, there is another that is even more important for information organisms. Disinformation is consumed as if it is information. Any artificial intelligence (designed with the intention of assisting our information processing needs) must be designed to process information in a way that can distinguish misinformation or disinformation. Even it it can only apply probabilities with the mention of the fact that it may be mis-informative or dis-informative, we need to know!

Information Entropy

Concerning the second law of thermodynamics:

Dr. Luciano Floridi explains in chapter 3 of his book, Information: A Very Short Introduction:

Entropy is a measure of the amount of “mixedupness” in processes and systems bearing energy or information. It can be seen as an indicator of reversibility: if there is no change of entropy then the process is reversible. A highly structured, perfectly organized message contains a lower degree of entropy or randomness, […] and hence it causes a smaller data deficit, which can be close to zero […]. By contrast, the higher the potential randomness of the symbols in the alphabet, the more bits of information can be produced by the device.

Similarly, from chapter 8 of Dr. Yunus Çengel’s book, Introduction to Thermodynamics and Heat Transfer, (as it relates to information entropy):

Processes can occur in a certain direction only, not in any direction. A process must proceed in the direction that complies with the increase of entropy principle […]. A process that violates this principle is impossible. [….] The performance of [information] systems is degraded by the pretense of irreversibilities, and entropy generation is a measure of the magnitudes of the irreversibilities present during that process. The greater the extent of irreversibilities, the greater the entropy generation. Therefore, entropy generation can be used as quantitative measure of irreversibilities associated with a process.

In retrospect, it appears prudent to quantify information entropy in order to accomplish the maximization of information dissemination, a matter of importance to intelligence officers and journalists alike, a notion similarly expressed by to Dr. Floridi in chapter 5:

Thermodynamics and information theory are often allies sharing one goal: the most efficient use of their resources, energy, and information. [….] The ‘green’ challenge is to use information more and more intelligently in order to reduce that energy input to ever lower targets, while keeping or increasing the output.

To what degree does an informer create information entropy, and to what degree does an informer minimize or maximize said entropy when an informee has synthesized the informants information?

Can information entropy be measured in a singular piece of semantic media, such as a news article, and can information entropy be juxtaposed with other, related pieces of semantic media?

Philosophy of Information, Systematic vs. Systemic Analysis

How would a systematic analysis of philosophy of information differ from a systemic analysis?

Presuming a systems theory approach, one might look at different classifications of information and their dependent relationships with one another in order to understand the nature, and possibly the behavior of various information classifications.

In order to perform a systemic analysis on the nature of information, one might look at the system as a whole, possibly inclusively-juxtaposing it with other systems, in order to understand the behavior of the system as a whole and how said system relatively yet absolutely dictates the behavior of its subsystems.

Looking at Dr. Floridi’s identified five classifications of data/information as described in chapter 2 of his book, Information: A very short Introduction:

Primary – The principal information
Secondary – The absence of primary information
Meta – The indications about the nature of non-meta information
Operational – The “[information] regarding the operations of the whole [information] system and the system’s performance”
Derivative – “The information that can be extracted from some information whenever the latter are used as indirect sources in search of patterns, clues, or inferential evidence about other things than those directly addressed by the information themselves, for example for comparative and quantitative analysis.”

Each of these five classifications of information could presumably be described to be of their own respective system. The interrelationships of these systems should have great effect to a singular piece of semantic media, such as a news article, which is observable by simply reading written language (organized communication). In order to understand the effects of these respective systems on the greater whole (e.g.: a news article), and how the whole affects its respective subsystems, a systemic analysis should be performed only after providing quantitative evidence that these five classifications of information can be controlled and predicted. Only after proceeding through a systems science methodology, applying meaningful metrics and predicting the effects of semantic information, can we then begin to understand the dynamics between semantic information and biological information.

In chapter 6 of Dr. Floridi’s book, he describes biological information as including:

Information as reality, e.g. patterns, fingerprints, tree rings;
Information for reality, e.g. commands, algorithms, recipes;
Information about reality, e.i. with an epistemic value, e.g. train tables, maps, entries in an encyclopedia.

Something may count as information in more than one sense, depending on the context.

These possibly open-loop systems (in some contexts) and closed-loop systems (in other contexts) presumably affect the value of semantic information. In order to appreciate an analysis using a complex adaptive system approach, it may be possible to quantify certain aspects of information synthesis when building knowledge because of the unique ways that we encounter semantic information according to these three biological information classifications. Such an undertaking should be treated holistically, i.e. systemically.