MetaTextbook creates textbook images customised to student interests
StartupRealities No 143
MetaTextbook is an artificial intelligence (AI)-led content producer which creates visual content for electronic textbooks customised to the interests and hobbies of the reader. For example, if a student likes football, most of the images in the physics or geography textbooks for this particular student will be related to football.
PSYCHOLOGY OF LEARNING
The customisation of imagery makes the content easier to understand and internalise thanks to two key factors:
1) The relevance and familiarity with the images creates a shortcut to existing knowledge in the student's mind;
2) The virtuous cycle of reassurance engendered by familiarity gives confidence to the student that the effort required to learn a lesson is minimal.
Aristotle described this process in Rhetoric: "To learn easily is naturally pleasant to all people."
PRINCIPLES OF IMAGE SELECTION
The company applies the principle of conceptual metaphors to select or create images on a given theme in hundreds of domains. For example, the concept of kinetic energy in physics is represented by a moving object, so for a football fan the customised textbook will display an image of a flying football; while for a student interested in animals the image will be of a trotting hippo or a galloping horse. Similarly, in a geography textbook, a lesson on the world's continents will be exemplified by an image of the Earth as a football with the black hexagons on the ball's surface shaped like the continents; or as a globe with the outlines of the typical animals that live on each continent.
USE OF AI IN IMAGE GENERATION
The company's database of existing images covers most basic concepts. Beyond that, images are generated using AI technology such as Generative Adversarial Networks (GAN) and Guided Generative Adversarial Networks (GGAN). GGAN is particularly useful in multiplying a metaphorical concept into images in multiple domains, for example in creating a sense of motion in the images of hundreds of different animals to represent kinetic energy in physics. AI-generated images also solve the problem of copyright, as self-developed images bring down to zero all external image acquisition costs.
CEO QUOTE
"Traditional teaching tries to take a student into the domain of the discipline. Our customised textbooks bring the discipline domain into the student's personal space, into their mental map of the world. This means less effort is required by the student to internalise and integrate what is already within their walled garden." - said MetaTextbook founder and CEO Methodius Texter.
MetaTextbook is an artificial intelligence (AI)-led content producer which creates visual content for electronic textbooks customised to the interests and hobbies of the reader. For example, if a student likes football, most of the images in the physics or geography textbooks for this particular student will be related to football.
PSYCHOLOGY OF LEARNING
The customisation of imagery makes the content easier to understand and internalise thanks to two key factors:
1) The relevance and familiarity with the images creates a shortcut to existing knowledge in the student's mind;
2) The virtuous cycle of reassurance engendered by familiarity gives confidence to the student that the effort required to learn a lesson is minimal.
Aristotle described this process in Rhetoric: "To learn easily is naturally pleasant to all people."
PRINCIPLES OF IMAGE SELECTION
The company applies the principle of conceptual metaphors to select or create images on a given theme in hundreds of domains. For example, the concept of kinetic energy in physics is represented by a moving object, so for a football fan the customised textbook will display an image of a flying football; while for a student interested in animals the image will be of a trotting hippo or a galloping horse. Similarly, in a geography textbook, a lesson on the world's continents will be exemplified by an image of the Earth as a football with the black hexagons on the ball's surface shaped like the continents; or as a globe with the outlines of the typical animals that live on each continent.
USE OF AI IN IMAGE GENERATION
The company's database of existing images covers most basic concepts. Beyond that, images are generated using AI technology such as Generative Adversarial Networks (GAN) and Guided Generative Adversarial Networks (GGAN). GGAN is particularly useful in multiplying a metaphorical concept into images in multiple domains, for example in creating a sense of motion in the images of hundreds of different animals to represent kinetic energy in physics. AI-generated images also solve the problem of copyright, as self-developed images bring down to zero all external image acquisition costs.
CEO QUOTE
"Traditional teaching tries to take a student into the domain of the discipline. Our customised textbooks bring the discipline domain into the student's personal space, into their mental map of the world. This means less effort is required by the student to internalise and integrate what is already within their walled garden." - said MetaTextbook founder and CEO Methodius Texter.
 |
| Bowling ball in motion (image source: Wikipedia) |
Comments
Post a Comment