May 19, 2011 at 15:08Imperfection of plain text as a medium of information
Speech, as we know it, is absolutely necessary, moreover, it is seen as the best means of exchanging information between people at this stage of development. Better than it can only be a telepathic exchange of thoughts, which we do not yet own (although some say that we no longer own). However, with the development of IT, the advent of computers and the Internet, we have acquired much greater opportunities for organizing our communication. But we continue to use the good old text (as I do now), or speech (skype). Nothing fundamentally new has appeared, and if it did, it is used in some separate organizations and isolated groups of people, but the broad masses for some reason do it the old fashioned way. This is the number one problem.
The second problem I want to tell about is the problem of preserving, systematizing and convenient access to information received on the Internet. I do not know how you personally, Khabrovsk residents, but I spend a lot of time on the Internet, hours every day. I find where intentionally, where by chance a huge amount of useful and potentially useful information for me. And if in the first case I take a number of steps to save this information, then in the second I often just too lazy to do this activity and I have to hope that the brain will keep this info-yu and so on. I have a program for hierarchical storage of information at my disposal (in fact, there are dozens of them, but I preferred one free) - CherryTree. In this program, I build trees, for example: software-> OS-> Linux-> useful_commands and in the sheet of this tree I save what interests me in the copy-paste. However, this is for text, just throw the usual links into the “bookmarks” of Firefox. Before moving to Linux, under Windows, I also used CyberArticle, a paid program for saving parts of web pages (with pictures and more).
This approach is inconvenient (searching for information among several programs is slower than in one), not flexible, not optimal, and not particularly effective, but I did not find the best. And it was precisely this impasse that prompted me to these fabrications presented above and below.
I thought, in fact, of the familiar methods for conveniently storing and organizing information, only tables and hierarchical structures can be noted. I note that by “information” I understand: text (1..N characters, with support for markup when displaying) or links to any files lying somewhere on the disk. It is important for me that statistical information is automatically saved: where and when this or that copy-paste / picture was taken, who is its author, in what context it was present initially and so on. Unfortunately, not one of the programs I have found fully functions in the required volume. From this moment, the image of a program that would fill this niche is already beginning to emerge: hierarchical structure, convenient preservation of parts of web pages, support for ordering and systematization (tree hierarchy + tags,
It was such thoughts that I had right up to the realization of the existence of the first problem (I recall that we just talked about the second). The first problem is the imperfection of the text as a means of exchanging information between people on the Internet. Namely: each of us, studying some kind of subject area, any special knowledge puts information in his head. I know little about the work of the brain, but from what I imagine about it: every new conscious object / fact the brain is trying to reconcile with what it already knows. It’s impossible to fit something completely new like this without a connection with the rest. Communication should be. Surely the brain breaks the object into the main components: shape, color (information from the eyes), the sound produced by the object (information from the ears), smell, texture, temperature, etc. received from the senses. Each of these signs fits into the corresponding area of the brain, between them there are connections that say that this is a single object, an additional link to a place in memory that makes it possible to understand exactly when we met also, each of the signs is attached to the previously existing signs of other objects, so that we can easily name all the objects of red color that we know, or cold in temperature, since having come to acc. brain area (“red color” or “cold to the touch”) from the links from there we quickly find everything we need.
Why not implement the same approach in a program for storing information? The simplest version of such a network is two-dimensional, in the plane, and with links and elements of the same type has already been implemented in programs such as the mind map . But such softwares are designed more for throwing simple diagrams for easy study of any area and simplification of memorization. I would like to adapt them for the full storage of many megabytes of various textual information, which, of course, imposes certain requirements on the implementation of such programs.
However, with this approach, I see another potential advantage. Imagine that one person owns a subject area, for example, is well versed in Linux / * BSD or mastered Adobe Premiere and knows how to mount videos (for simplicity, I touch only on computer-related topics). Now this person, if he wants to share knowledge, writes articles, lessons, examples, guides on the forum / blog. And it's good that he writes. But the problem is that writing a text requires a certain mental cost and effort from a person. The expression of their thoughts in text form is given to everyone differently and for some it is a long painful process.
In a word: continuous problems both for those who want to share information, and for those who want to perceive it. I was quite curious and at one time was interested in many topics, from capturing video from an analog camera and video editing to programming, so I’m familiar with all this firsthand.
Obviously, it is the good old exchange of information through text that becomes the bottleneck of modern information exchange.
Now imagine the situation using the program for the approach to organizing the information proposed above. A person who knows the subject area (let's call its author) already initially, for his beloved, organizes his data in the form of this ... network (the language asks for a "neural network", but this name is already takenclass of artificial intelligence algorithms ) and uses it independently happily ever after. But if there comes a time when he wants to share knowledge ( share your knowledge ... and it will come back to you more than once ), he does not give birth to a tormented website with guides, he simply opens access to his network with the “easy cursor movement” (or rather to its part) and anyone who wants to get this information (let's call it a subscriber) just uploads this branch to himself and integrates into his network with subsequent synchronization and updating of changing nodes and connections. Of course, all this is as simple and transparent as possible for the user, no more complicated than using a modern social network. The problems are solved:
Thus, both problems stated at the beginning of this treatise are solved. I already see the Internet in which such networks are common among ordinary people just as social networks are now popular. Good or bad, but mostly sites with guides on a particular topic will be a thing of the past, sites unique, but confusing, peculiar, which you won’t understand at once. The forums will leave (often administrators impose on the forum engines not only the functions of communication between people - for which forums were created - but also guides collections, incomprehensible collections for the beginner, spread out in sections, filled with topics with the headings "READ BEFORE POSTING!", Closed with the only messages: the negatives of admins, moderators and simply old-timers who spread rot for newcomers who did not understand the intricacies of historically stratified ones
Probably such sites will be externally freed from the heaps of manuals and instructions, they will disappear from the eyes, dissolve in networks . The site of the near future will be completely and completely filled only with the communication of people among themselves, an analogue of modern forums or blogs. At the top of each site, as we now recognize the RSS icon, a “network” sign will be visible. Each user, as of now, has a link to ICQ or social. the network will link to its network ...
From the point of view of implementation, the following features and moments are presented:
In general, these are just the main points that I described in order to try to evoke the same image that I had. Please introduce it, consider, evaluate the possibility of implementing such an idea and its prospects from your subjective point of view.
The second problem I want to tell about is the problem of preserving, systematizing and convenient access to information received on the Internet. I do not know how you personally, Khabrovsk residents, but I spend a lot of time on the Internet, hours every day. I find where intentionally, where by chance a huge amount of useful and potentially useful information for me. And if in the first case I take a number of steps to save this information, then in the second I often just too lazy to do this activity and I have to hope that the brain will keep this info-yu and so on. I have a program for hierarchical storage of information at my disposal (in fact, there are dozens of them, but I preferred one free) - CherryTree. In this program, I build trees, for example: software-> OS-> Linux-> useful_commands and in the sheet of this tree I save what interests me in the copy-paste. However, this is for text, just throw the usual links into the “bookmarks” of Firefox. Before moving to Linux, under Windows, I also used CyberArticle, a paid program for saving parts of web pages (with pictures and more).
This approach is inconvenient (searching for information among several programs is slower than in one), not flexible, not optimal, and not particularly effective, but I did not find the best. And it was precisely this impasse that prompted me to these fabrications presented above and below.
I thought, in fact, of the familiar methods for conveniently storing and organizing information, only tables and hierarchical structures can be noted. I note that by “information” I understand: text (1..N characters, with support for markup when displaying) or links to any files lying somewhere on the disk. It is important for me that statistical information is automatically saved: where and when this or that copy-paste / picture was taken, who is its author, in what context it was present initially and so on. Unfortunately, not one of the programs I have found fully functions in the required volume. From this moment, the image of a program that would fill this niche is already beginning to emerge: hierarchical structure, convenient preservation of parts of web pages, support for ordering and systematization (tree hierarchy + tags,
It was such thoughts that I had right up to the realization of the existence of the first problem (I recall that we just talked about the second). The first problem is the imperfection of the text as a means of exchanging information between people on the Internet. Namely: each of us, studying some kind of subject area, any special knowledge puts information in his head. I know little about the work of the brain, but from what I imagine about it: every new conscious object / fact the brain is trying to reconcile with what it already knows. It’s impossible to fit something completely new like this without a connection with the rest. Communication should be. Surely the brain breaks the object into the main components: shape, color (information from the eyes), the sound produced by the object (information from the ears), smell, texture, temperature, etc. received from the senses. Each of these signs fits into the corresponding area of the brain, between them there are connections that say that this is a single object, an additional link to a place in memory that makes it possible to understand exactly when we met also, each of the signs is attached to the previously existing signs of other objects, so that we can easily name all the objects of red color that we know, or cold in temperature, since having come to acc. brain area (“red color” or “cold to the touch”) from the links from there we quickly find everything we need.
Why not implement the same approach in a program for storing information? The simplest version of such a network is two-dimensional, in the plane, and with links and elements of the same type has already been implemented in programs such as the mind map . But such softwares are designed more for throwing simple diagrams for easy study of any area and simplification of memorization. I would like to adapt them for the full storage of many megabytes of various textual information, which, of course, imposes certain requirements on the implementation of such programs.
However, with this approach, I see another potential advantage. Imagine that one person owns a subject area, for example, is well versed in Linux / * BSD or mastered Adobe Premiere and knows how to mount videos (for simplicity, I touch only on computer-related topics). Now this person, if he wants to share knowledge, writes articles, lessons, examples, guides on the forum / blog. And it's good that he writes. But the problem is that writing a text requires a certain mental cost and effort from a person. The expression of their thoughts in text form is given to everyone differently and for some it is a long painful process.
In a word: continuous problems both for those who want to share information, and for those who want to perceive it. I was quite curious and at one time was interested in many topics, from capturing video from an analog camera and video editing to programming, so I’m familiar with all this firsthand.
Obviously, it is the good old exchange of information through text that becomes the bottleneck of modern information exchange.
Now imagine the situation using the program for the approach to organizing the information proposed above. A person who knows the subject area (let's call its author) already initially, for his beloved, organizes his data in the form of this ... network (the language asks for a "neural network", but this name is already takenclass of artificial intelligence algorithms ) and uses it independently happily ever after. But if there comes a time when he wants to share knowledge ( share your knowledge ... and it will come back to you more than once ), he does not give birth to a tormented website with guides, he simply opens access to his network with the “easy cursor movement” (or rather to its part) and anyone who wants to get this information (let's call it a subscriber) just uploads this branch to himself and integrates into his network with subsequent synchronization and updating of changing nodes and connections. Of course, all this is as simple and transparent as possible for the user, no more complicated than using a modern social network. The problems are solved:
- Most importantly: the subscriber does not gain access to a certain part of the information presented in text form, which is never 100% reflecting the original volume of the author’s knowledge: since the author cannot look at his own leadership through the eyes of a person unfamiliar with the subject area, he simply does not know that he accidentally did not finish something important and did not explain. You can give an example of painting the fence with a second layer of paint: as a dyer, it is difficult to understand whether you painted this very place again or not. The subscriber gets access to the author’s network and it’s almost the same for him as to get access to the author’s brain ! It generates requests to this network.and extracts slices of information in any form that interests him. If he needs to quickly do something, he asks for the shortest step-by-step guide, the minimum set of actions that must be performed to achieve the desired result. If he wants to understand in more detail, he asks for everything in all details and in a form convenient for his personal perception.
- The subscriber receives information presented in a form that depends only on the settings of his program. Color, font, background - he sets everything himself and as a result, any knowledge from any sources looks the same and familiar.
- Information is 100% relevant (with the exception of a few seconds to sync). All of these copyrighted ones “I'll be free on weekends and post information on the site” are a thing of the past.
- It is easy and intuitive for the subscriber to edit and supplement the information received (since the matter essentially happens on his own network ), and, in turn, becomes the author himself for others
- You can solve the problem of the language barrier! Indeed, translating individual network nodes (which are atomic words or phrases) is much easier than parsing a “full-fledged” text with a computer. Also, when translating, to improve its quality, you can analyze the connections between network elements.
Thus, both problems stated at the beginning of this treatise are solved. I already see the Internet in which such networks are common among ordinary people just as social networks are now popular. Good or bad, but mostly sites with guides on a particular topic will be a thing of the past, sites unique, but confusing, peculiar, which you won’t understand at once. The forums will leave (often administrators impose on the forum engines not only the functions of communication between people - for which forums were created - but also guides collections, incomprehensible collections for the beginner, spread out in sections, filled with topics with the headings "READ BEFORE POSTING!", Closed with the only messages: the negatives of admins, moderators and simply old-timers who spread rot for newcomers who did not understand the intricacies of historically stratified ones
Probably such sites will be externally freed from the heaps of manuals and instructions, they will disappear from the eyes, dissolve in networks . The site of the near future will be completely and completely filled only with the communication of people among themselves, an analogue of modern forums or blogs. At the top of each site, as we now recognize the RSS icon, a “network” sign will be visible. Each user, as of now, has a link to ICQ or social. the network will link to its network ...
From the point of view of implementation, the following features and moments are presented:
- Problem: node synchronization when connecting two networks (or parts of them). Ideally, it is assumed that the “computer” node of one person will correspond to the “computer” node of another person and simply merge with it into a single whole. However, this word can be written in another language, it can be spelled incorrectly, spelled slang or abbreviation.
It is probably worthwhile to introduce a certain similarity coefficient, expressed as a percentage, and merge the nodes when a certain percentage is reached (70% -80%?). However, this will not save you from completely different spelling words.
Output 1: recognize the real meaning of the word. But this is a long-standing problem and it deserves a separate article.
Output 2:to do something similar to how it / in my philistine view / is implemented in the human brain. Namely, for each word, for each node, enter a list of qualifying nodes with which it will be associated. When a person hears the word “computer”, the definitions “metal case” “luminous screen”, “gray color”, “fast work”, “darling, infection”, in general, how someone rushes through his head. Ideally, each concept will correspond to the entire possible series of associations (which will be requested from a single database, more on that below), users who own part of them will be recognized by the program (as a subset of many associations) and their nodes will be successfully integrated.
It will not work to place them all in one plane, so you have to enter some ... layers or echelons. The “visual” layer, the “audio” layer, the “kinetic”, “olfactory” layer ..., generally corresponding to the basic human senses. However, firstly, the problem of describing synonyms of one and the same concept by different people is not solved here, and secondly, forcing a person to enter all this is too cruel. People will not use such a program. The advantage of this program, for which it will be loved, should be the convenience and very quick addition and editing of nodes. So search and add refinement nodesstill be on the conscience of the computer. Probably at first, until full-fledged text recognition tools are invented, it will be necessary to build a network of servers to which users' clients will connect and request words that are allowed to enter. Something like a modern Googley "maybe you mean the computer s serial ports, and not Computer?" - Obviously, the program should provide different access to information. For example, to be multidimensional, if we consider some subject area (part of the network) from the point of view of a multidimensional cube (extracting information from it is respectively a section). From the point of view of the network, there can be different types of nodes or nodes between them (brotherhood, submission
and humiliation, generation, ...). - What and how much to share. Quotes, “copy-paste” and entire literary works, which so far do not seem to be fully disassembled into components, are supposed to be stored in full text. Maybe they, like pictures or binary files, should be stored separately in the file system, and only links to them should be stored in the database. Everything else that can be broken down, various knowledge that is compact enough to be analyzed by a computer system at a finite time, needs to be broken up and stored.
- It is not clear what requirements will be imposed on computer resources for this program to work. Even if there is enough processor and RAM, the base in theory can grow on the hard drive by many tens and hundreds of megabytes, which will make it difficult for users to back up. Speaking of backups. I immediately take it as an axiom that the question of storing information in the database of each specific user is very important and this should be considered separately. People are experiencing big problems even simply losing access to their native ICQ or email with tons of contacts dialed over the years, and what if, as a result of a hard disk failure, they at once lose the entire database in which they store all their received “through the computer »Knowledge, its history, logs of correspondence that are dear to people as a memory, and everything else valuable? ..
However, a radical solution seems to be the initial construction of this program on the basis of clouds (and this will be their dawn, finally their start will be removed from the slip!). Moreover, this solution will radically preserve the memory of hard drives, if we exclude multiple repetitions of the same nodes and combine the networks of all users into one large network , remembering for each node and for each connection to whom they belong. - The interface should be visual and intuitive (graphical, three-dimensional, changing as the query parameters change), but at the same time, if necessary, provide all the wealth of settings for saving, editing and searching for information.
In general, these are just the main points that I described in order to try to evoke the same image that I had. Please introduce it, consider, evaluate the possibility of implementing such an idea and its prospects from your subjective point of view.