An Adapting Information Potential (AIP – /ˈeɪ ˈaɪ pɪː/ n [C]) is a piece of information containing a semantic core that is presented in several contexts and scenarios. It is evaluated using a binary assessment of the feasibility of its application (“carried out” or “not carried out”, “can do” or “cannot do”). Here, “semantic core” can be understood to mean the designation of a retained piece of information for a specific use.
Principles of AIP construction
- AIP possession is categorically evaluated using binary oppositions (“present/absent”, “has/does not have”).
- The semantic core is presented in a number of multi-level scenarios.
- Analog or digital presentation of the information is chosen according to the sphere of use.
- Infinite number of multi-level scenarios in which the AIP is presented.
- The semantic core remains the same regardless of any clarifications and changes made to individual scenarios that present the AIP.
- Classification of the AIP at an agreed level of perception.
History of AIPs
The AIP concept was conceived by Aleksandr Yuryev, who came up with it on the basis of experience he had gained from 2012 onward while working on the integration of non-linear/adaptive content technology and non-linear testing of language skills, as found in language audits. It was also the result of Yuryev’s observation of learning within the Duoyulong (Chinese name: “多语龙”) online university of foreign languages and the difficulties identified in structuring information for the Intelligent Learning Platform, a personalized online teaching tool using artificial intelligence.
The theory of universal AIP units was proposed as a solution for structuring information. AIPs constitute a structural element of non-linear, adaptive learning materials, as well as a breakdown of different professions’ work activities, complex compound skills and abilities into a binary evaluation of ability.
In 2019, the AIP Institute was founded in Shenzhen, China, and an international research team was formed to research and implement AIP theory.
Prospects for practical use
AIP theory has the potential to create a single methodological space for vertical and horizontal evaluation and training in both education and work settings, especially where work and education are already naturally connected, i.e. in corporations, higher educational institutions, and high schools. In other words, it has the power to make school, university, and manufacturing company spaces porous and communicative, allowing these institutions to interact using a shared language of ideas and evaluations.
This results in a high-added-value solution that is particularly desirable in the context of the current industrial transition, providing a new quality of human capital and improvements to learning and working; an opportunity for everyone to see their objective integral profile with the wealth of opportunities this includes; improved goal setting and personal development planning; a seamless transition from education to work, or from one place of work to another; conscious personal development; and equally conscious management of human capital.
At the same time, it is important to create AIP content environments that are very substantial in size. In other words, materials from academic disciplines must be structured, professions mapped, expertise deepened, specialists trained, and scientific and practical research expanded.
AIP evaluation and learning algorithm
Aspects of the AIP evaluation and learning algorithm:
- Cognigraphic data – this is information about mastery of AIPs, including an evaluation of the learner’s personal cognitive abilities, intellect, and styles.
- Non-cognigraphic data – this is information about the learner that is NOT cognigraphic data or that has not yet been converted into cognigraphic data. This includes the learner’s activity, demographic data (age, gender, location, etc.), preferences, habits, political and/or religious views, etc.
- Learning profile – a model of the learner based on attributes characterizing features of the learner’s bodily functions and consciousness, comprising cognigraphic and non-cognigraphic data.
- Level – this is a characterization of the learner based on their cognigraphic and non-cognigraphic data, determining the selection of levels, styles of content, types of content and methods of delivery.
- Scenario – a unique presentation of the AIP’s semantic core, tailored to the user’s unique absorption of information and their perception.
- Blocks – the smallest unit for storing and displaying AIP information.
- Checks – validation of the user’s activity parameters within AIPs, their transition to other scenarios or levels, or their end-of-scenario test score.
- Connections – these define the order of transitions between blocks based on certain conditions.
Description
This approach to evaluation and learning allows specific AIPs or a package of AIPs to be identified in the behavior of a user by monitoring their learning progress and its effect on their behavior. New AIPs, registered based on their indicated presence or absence, change the overall content, shaping the dynamics of the entire learning and evaluation process.
Individual user profiles are populated with cognigraphic data during the learning process, while non-cognigraphic data are added at the very beginning and are updated when relevant events occur (surname change, religion change, new hobbies, etc.).
An array of cognigraphic data is continuously created as a result of the user’s behavior, and in many ways it subsequently defines this behavior, offering scenarios for one level of difficulty or another that match the individual’s learning style to a certain extent.
Educational scenarios are short, between 2 and 10 minutes long. As a general rule, this is enough time for AIPs to be presented and for the user’s ability to handle the cognitive load to be evaluated. The binary scores obtained activate subsequent scenarios with a higher or lower cognitive load than the previous scenarios.
The set of AIP tests allow a comprehensive, integral score to be obtained. Binary oppositions also influence whether the user progresses to the next AIP, repeats the same material, or is presented with it in different forms or at a different speed, etc.
The binary opposition of “present/absent”, “has/does not have” in the evaluation of learning progress has deep epistemological roots. Binarism as a structural concept in epistemology, the theory of knowledge, describes the structure of binary oppositions as being innate and inherent. Binary evaluation mechanisms are integral to cognitive and decision-making processes; they have a solid neurobiological foundation. The same binary opposition mechanism of “yes or no” methodologically postulates the practice of revising, cataloging, and assessing whether a set is complete, as well as evaluating the ability to solve a problem, formulate a question, and answer this question.
In terms of the user who is undertaking the learning and whose progress is constantly being monitored across a wide range of parameters, the openness and flexibility of the binary opposition allows a dynamic, highly accurate, intellectual representation of the person to be constructed at any moment. Both the separate sections and the integral indicators allow evaluations to be carried out that are far more precise than traditional degree certificates and résumés, which the education and job market are using less and less, instead seeking more relevant solutions.
Flexible classification system
- Package – set of AIPs according to the needs of a given classification system. Packages allow unlimited nesting;
- AIP type – division by type of processes described;
- AIP class – system of classification within AIP types with multi-level nesting.
One AIP unit correlates with others not only semantically, but also thematically, by difficulty level, sources, size, number of identified links with other units, and a whole range of other parameters. Accordingly, packaging and repackaging are ways of organizing AIPs for storage, as well as for direct use. The combination of similar characteristics allows AIPs to be grouped into packages/containers, aggregated, or split again, and enables packages to be broken down according to the specific requirements in question.
The current algorithm for AIP-based learning can be adapted to various information classification systems, including to the classification of learning objectives presented in Bloom’s Taxonomy, as well as to systems for evaluating cognitive abilities.