HU226347B1 - Computer system and method for teaching handwriting and reading - Google Patents

Abstract

A computerised system (S) and method is described for teaching handwriting and reading in the form of copying exercises, the system (S) comprising : a) a curriculum module (11) comprising written and verbal copy exercises and motivation materials associated with the exercises, b) a demonstration module (12) presenting the motivation material for the exercise and/or providing information to describe the exercises on the basis of the exercises of the curriculum module (11), c) an execution module (13) operating on the basis of the exercises of the curriculum module (11), which execution module (13) has - means for starting the exercise described by the demonstration module (12), - means for demonstrating the written and/or verbal samples to be copied, in writing and/or on a verbal basis, - means for scheduling and controlling the copying of samples which means also detects the written and verbal copies made by the learner, - means for comparing the copies with the samples and furthermore determining and storing the extent of deviation between the copies and the samples, and - means for transferring to an assessment module the deviation rates determined between the copies and the samples, and d) assessment module (14) evaluating on the basis of the deviation rates the exercises performed and choosing the next exercise based on the assessment, and furthermore informing the learner on the assessment, preferably in accordance with the motivation material of the exercise.

Description

FIELD OF THE INVENTION The present invention relates to a computer training system and method for automated instruction in handwriting and reading. The essence of teaching automation is that the student performs copy exercises in a given syllabus by copying written patterns written on a pen screen and written on a headset or speaker, written on a pen screen, orally into a microphone. The system detects duplicates, compares them with patterns, and presents comparisons visually to the learner, both in writing and verbally, so that they can make appropriate corrections. Based on the success of the copying exercises performed, the system determines after each exercise whether there is a need for feedback, ie some modification of the sequence of exercises, repetition of an exercise under unchanged or modified conditions, or a return to an earlier exercise during the exercise being performed. to remedy any deficiency noted. Each exercise is presented verbally to the learner in verbal, still and motion pictures, and with feedback is provided with an oral explanation for repetition or retraction.

In automated education, students are not forced to learn, but motivated. The system introduces a visual motivational world in which stories take place and the learner becomes an active participant in these stories in a way that requires them to complete the story by completing specific reading and writing tasks. Pictures of the motivational world also serve as context for writing and reading assignments.

The field of application of the invention is education. Using the system, students are able to learn handwriting and reading independently without the aid of a teacher.

There are many ways to teach handwriting and computer reading.

U.S. Patent No. 5,730,602 deals with the teaching of the technical, mainly motor, elements of handwriting. The essence of the method is that the computer-connected digitization screen measures the parameters of the student's pen strokes and the characters formed from them (pen pressure, tilt, angle of turns, typing speed, etc.) and assesses student's writing, drawing conclusions and correcting errors. it provides the student with textual, verbal and visual instructions and advice. The teaching of the technical elements of writing can thus be embedded in a computer-based educational curriculum.

The method described in U.S. Patent No. 6,215,901 is also used to measure, record, and analyze the movements of a pen during each stroke of a character. Analyze pen movement parameters such as typing speed, pen pressure, pen stroke duration, pen stroke quality, character tilt, position, pen tilt, and then calculate key characteristics of the character you write, such as shape , inclination, greatness. The key features are then compared with those of the reference character, and then advice is given on how to improve the writing quality of the character. The method can be embedded in the handwriting education curriculum.

The two known solutions aim to accurately measure the pupil's letter formation and to explain the perceived deficiencies and their correction. However, it is questionable whether a 6- to 8-year-old child will be able to form letters better if the computer explains to him how hard he is to press the pen, to rotate the letters, to zoom in or out, to round or to sharpen the reference letter.

Palm Computing Ltd. of Revelation describes helping kézírástanuiást implemented computer program handful of http://homepage.powerup.com.au/ evcom _r / web / CWWorks.html website. The teacher writes punctuation marks on the screen as a pattern and the student is responsible for copying the pattern. For this solution, the educators' legitimate objection is that they can do little more than a paper sheet. Educators are reluctant to use a computer and are reluctant to deviate from the usual solutions.

Another common shortcoming of the solutions described is that they do not deal with writing and reading education, which is contrary to the view of modern pedagogy, another common shortcoming is that they cover only a certain section of writing and reading education, from the very beginning until the student is fully proficient in handwriting and reading. Another disadvantage is that it is difficult to integrate into the school education process, which is only due to teachers' reluctance to use computer solutions. Another reason is that the use of computer-based education solutions requires a substantial change in the curriculum and the appropriate training of teachers.

The object of the present invention is to provide a system and method for teaching handwriting and reading without the assistance of a teacher, preferably for children aged 6 to 8, from the beginning to the full acquisition of the skills.

It is also an object of the present invention to effectively utilize the advanced results of pedagogy, the functional teaching of writing and reading, in the context of automated education, whereby handwriting and reading instruction is to be carried out on the basis of live speech and mutually.

It is also an object of the present invention to provide an educational system that consists of a series of exercises included in an educational program but is flexible to the capabilities of the individual student.

It is also an object of the invention that students do not compulsively learn writing and reading, but playfully acquire it through the built-in motivational world.

The object has been achieved by a system and a process according to the main claims. Preferred embodiment of the invention2

EN 226 347 Β1 and their embodiments are disclosed in the dependent claims.

The invention by way of example! preferred embodiments and embodiments thereof are hereinafter described in the drawings, wherein

Figure 1 is a schematic diagram of a computer system according to the invention, a

Figure 2 is a flowchart for performing exercises, a

Figure 3 is a diagram illustrating a snapshot of performing a duplicate pattern exercise, a

Figure 4 is a snapshot illustrating a practice of using a graphical grid

Fig. 5 is a drawing illustrating light-point overwriting, a

6A-C. diagrams illustrating virtual grid operation, a

Figure 7 is an exercise example! task sheet, a

Figure 8 is a drawing illustrating a snapshot of a reading exercise, a

Figure 9 is a snapshot of a pronunciation exercise, a

Figure 10 is a snapshot of a continuous written copy with frames, a

Figure 11 is a snapshot of a continuous written copy without frames, a

Fig. 12 is a drawing illustrating a snapshot of a written copy of an oral pattern without frames, a

Figure 13 is a flowchart of the practice of copying a written pattern marked with frames, a

Figure 14 is a flowchart of the practice of duplicating framed double patterns, a

Figure 15 is a flowchart of an exercise in copying a written pattern marked with frames, a

Fig. 16 is a flowchart of a practice written using frame-by-frame copying of an oral pattern, and a

Figure 17 illustrates a specific embodiment of a system according to the invention. schematic drawing of a target device.

An example of automated teaching of handwriting and reading! system is built on a general purpose computer, but it becomes a specific structure through the implemented functions.

The methods and methods by which the functions of the training system of the present invention will be accomplished are briefly described below. Because these methods and procedures are related to writing and reading instruction, the training device will be described with specific methods, methods, and concepts for teaching writing and reading, but with the requirement that every detail of the methods and procedures shown in the as it stands today should be an intelligent computer solution.

The most important feature of the automated training described in connection with the present invention is that the training is carried out not by man but by machine. The automated writing and reading instruction is provided on the computer system S shown in FIG. The S system contains 10 computers and 15 student stations. The learning station 15 is equipped with a touch screen, pressure, capacitance or inductance change or other physical impact sensitive digitalization screen 17 for handwriting recognition, text information, still images and motion pictures. The S system stores samples of characters, words, sentences in digital form, and displays them on the 17 digitization screens during the exercise, in order for the student to copy the samples. The student uses a special pen 16 (pen, style) to write on a flat digitalized screen 17, which senses and converts the motion of the pen 16 into electrical signals. These signals are converted by the S system to a digital form that is similar to the stored written patterns.

The surface of the 17 digital picture screens is divided into an upper and a lower part, and the upper, demonstration area shows still and moving pictures that serve as motivation material for the exercises, demonstrates the exercises and evaluates the completion of the exercises, it senses and converts the student-written copies into electrical signals.

System S also includes a microphone / headphone 18 or similar unit that serves as an input or output device. The S system also stores verbal patterns corresponding to written patterns, which are reproduced by the learner or speaker for the learner. As the student copies loudly the S patterns given by the S system or reads the written patterns, the microphone senses and converts the speakers and words into electrical signals. These signals are converted by the S system into a digital form that is similar to the stored oral samples.

The student's written and oral copies of the exercise are compared by the S system to the written and oral samples provided for the exercise and presented visually to the learner so that the student can correct errors during or after the exercise. From the comparisons, System S makes an assessment to determine whether the curriculum for reading and reading practice needs to be modified, such as repeating an exercise or returning to an earlier exercise.

In addition to copying as a basic methodological element, another conceptual feature of machine learning is the creation and consistent application of context. Contextualizing means that each practice is preceded by an image or series of images that establishes an associative relationship between the written and spoken word patterns in the practice and their original meaning.

An important complement to machine learning is the paper-based worksheets 19, which contain the written part of some exercises, in the same format as they are on the 17 digitization screens

EN 226 347 Β1 appear. The student repeats the on-screen written exercises on the 19 worksheets to practice writing on paper.

Writing and reading are mirror symmetric and related activities. Writing is the process of recording a thought or speech, essentially a coding activity. Reading is the decoding of written text, which results in the text being reproduced into thought or, in the case of reading aloud, into speech.

Pedagogy holds that teaching writing and reading is effective if the two processes are built on one another and run in parallel. The most effective way to get to know the characters, to memorize them first, and to match the spoken words with your writing skills is to learn how to spell the letters manually. During reading, the characters learned in handwriting are captured by converting the characters into spoken voices and spelling the voices into words, allowing for further development of writing skills, which in turn can be built on advanced reading skills and integrated recognition of word pictures. This is how learning and writing are built on each other step by step, so the two processes cannot be separated but must be handled together. In the initial phase of machine learning, we extend this to the extent that most of the exercises do the writing and reading together, but they are separated later.

One of the meanings of functional education is to start writing and reading meaningful words and sentences as early as possible. The other meaning of the term is that reading and writing is derived from speech and must be based on the spoken form of words during education. The obvious way to implement functional education is to use the dual pattern, which means that the written and oral patterns are present at the same time. Within the constraints of school education, intensive use of the double pattern is not easy to achieve due to the lack of a constant close relationship between the teacher presenting the double pattern and the individual student. There is no such limitation in machine learning; the S system is dedicated to a single student at a given time, allowing for continuous and intensive use of the duplicate pattern. The S system only "deals" with the student, writes the sample words on the screen for him / her, drops the words for him / her, either as a sound or as many letters as he / she needs, and keeps the student echoing to say.

This circumstance allows us, from the outset, to lead the learner to the correspondence between the spoken and written forms of the simpler words. Beyond that, the computer's ability to dynamically edit text information to highlight parts of words on the screen, illuminate individual letters and letter relationships, allows you to examine the sound of the characters within a given word, writing words, pronunciation of words in whole, in letters or in frames. Machine learning thus allows the student to complete all exercises in the context of meaningful words and word sentences. To begin this process as soon as possible, the curriculum should be structured in such a way that students initially learn only a subset of the alphabet, and words and sentences formed from this subset become the subject of the exercise. Later, the initial subset needs to be supplemented with additional characters to enrich the vocabulary of writing and reading practices.

In machine learning, methodological elements called sample copying and feedback play a key role. A well-known form of learning for the child is to copy and feed back the pattern, as he or she continues to practice this throughout his or her development, starting with the acquisition of speech. During speech learning, the child copies the verbal pattern provided by the parent, and if the copying fails, there is feedback, repetition of the pattern by the parent, followed by an improved copy from the child. Repetitive copying gives better results and the child learns to speak slowly. In literacy teaching, a similar process takes place, with copy and feedback as the dominant element. The student performs many copying exercises until he / she begins to practice writing and reading independently. Feedback plays an important role in its development, which takes the form of checking the results of exercises and repetitions aimed at correcting errors and eliminating deficiencies.

In machine learning, all exercises are about pattern, copy, and feedback. Each reading and writing exercise is also a test of how well a student is able to copy a given pattern. At the end of the exercise, the S system evaluates the differences and identifies the result, and then, taking into account the tendency in the given practice and in the previous practice, decides whether it is possible to move on to the next practice in the curriculum, practice with unchanged or modified conditions, and which exercise in the curriculum you should go back to.

In written practice, the written form must be produced from the spoken pattern, while in reading, only the written pattern is available, the spoken form must be created. In copy-based education, it is necessary to create a state in which both patterns are present, as this is the only way to illustrate the process of writing (coding) and reading (decoding). A special feature of the double pattern is that it illustrates the inseparable relationship between writing and reading. The problem with using a dual pattern is the need for a medium that can convey both written and oral patterns to the learner. At school, the teacher has the task of presenting the double pattern, for example, by saying a word, then pronouncing the sounds that make up the word, and writing them in letter form on the board. Then the

EN 226 347 Β1 re-spell the word you wrote down and pronounce it again. In this case, the process of coding and decoding takes place in front of the learner's ear, and the oral and written patterns are present, and one can observe how one is created from the other. The student repeats the same exercise at the board, copying the word pattern written by the teacher letter-by-letter, while uttering it with the teacher in a sound tone, and finally saying the whole word. A slightly similar effect is created when children read words on the board in a choir. The teacher and the best students read, and the others join and copy them. However, while copying, they also monitor the written text, make the mapping between the written and the read (spoken) form, and after several such mappings, they will be able to read independently. The function of the double pattern (besides copying) is thus the cross-referencing of the written and the oral pattern and thereby the modeling of writing and reading.

It is important that during the initial phase of machine learning, the student makes a duplicate copy of the duplicate pattern, both in writing and verbally, as this is a practice of writing and better captures the correspondence between the spoken and written form of the word. While copying a character, it repeats verbally the sound spoken by the S system, and when it reaches the end of writing a word, it pronounces the entire word after the S system.

The essence of the invention is affected by the difference between the oral and written patterns from the learner's perspective. Because he is able to speak well, it is not a problem for the learner to repeat the verbal patterns, even for him / her to break up words into vowels. However, writing is a new thing for him. Not only is the letter formation difficult, but the appearance, meaning and connection of the written pattern itself with the oral pattern, the recognition of the letters, the matching of the letters and the vowels are all issues that are only gradually becoming conscious of it. This awareness process is accomplished through duplicate exercises. The initial, long, phase of education is about this, and writing is ongoing.

There will be a period of learning when it comes to switching from dual to independent reading and writing. At this stage, you should gradually eliminate the duplicate pattern and move to the purely written pattern or the purely verbal pattern, which is called a single pattern to confront the duplicate pattern. The virtual sample that completes the single sample into a duplicate sample is called the complement sample. The S system stores duplicate samples for each exercise, and the complementary sample is available for a single sample exercise. From a formal logical approach, self-writing and self-reading mean copying the complementary pattern, in which case the S system checks the result of the exercise by comparing it with the complementary pattern. The introduction of the complementary pattern is also important for the present invention in terms of vocabulary, as the term "copy and complement" is often used in the description, not the words "read and write".

The double pattern enables the use of known reading models in machine learning.

Synthetic reading means breaking words into syllables or letters and then putting them together later. The learner can read when he / she develops a phonetic awareness in addition to knowing the letters, ie he / she understands that speech is made up of so-called phonemes, and that spoken words are characters, and reading a word means that spoken phrases such as Ibi, Iala and Izl in the word 'house' must be spelled out.

In synthetic reading, learners combine continuous reading of words and then syllables to practice continuous reading of words during dual pattern exercises. This method alone cannot be applied to languages where the written and spoken forms of words often differ significantly from one another.

The Global Reading Method teaches you to read whole word pictures instead of silencing the letters that make up the word, and later extracts each letter from the word pictures. In dual pattern practice, the machine spells out the entire word and then spells it, presenting the written pattern. The S system illustrates images of each word to be read, and context-sensitive images provide a great help for global reading.

The combined reading model uses both methods of reading, with particular reference to the specific features of the language. In Hungarian education, the synthesizing model is predominant, in English the global model is dominant, although in the early stages of education English also uses the synthesizing method.

Machine reading instruction through dual patterning can be very effective as the S system is able to relentlessly push the student into duplicate patterns using all three models, while relieving the task-related tension with playful elements.

Utilizing the capabilities of a computer, the present invention introduces a methodological innovation in handwriting instruction that facilitates, on the one hand, the graphical operation of copying and, on the other hand, a direct graphical comparison of a written pattern and a student-written copy. These advantages are achieved, as shown in Fig. 3, by providing geometric reference points in the working area 30 of the digitizing screen 17 for copying written patterns 32. In the workspace 30, a sample line and copy line are displayed according to the school notebook used in that language, for example, for European languages, the line 31 includes two horizontal lines that define the font size. A rectangular frame 33 appears around each character of pattern 32 when the character is copied, and a blank frame 33 of the same size as pattern 33 appears below the pattern, just below pattern 32, so that the student of the same shape

EN 226 347 Β1 and make 34 copies like the 32 patterns. Framing characters is possible in all languages, for example, Chinese characters are particularly beneficial. Frame 33 is a constraint, but its use has more advantages than disadvantages. Frame 33 allows for more accurate font shaping when copying because it provides even more geometric reference points along with the notebook-like train 31.

As the student copies a character, the S system places a copy (continuous, dashed, dotted) of a color, tone, thickness, or type as shown in Figure 3 on the pattern to immediately show the difference between the 32 patterns and 34 copies. This will illustrate to the student, more accurately and more accurately than any instrument, whether the copy was done correctly and what should be corrected in the next copy attempt. This feedback within practice is called visual self-testing.

Figure 3 is a snapshot of an exercise, the actual pattern element is the letter "I" (its pattern is marked with a contrasting 33 'frame), and its written copy has begun in the blank frame below. Before the letter "I", the learner had already copied the word "lili" and the S system placed a pattern on the copied letters with a line opposite to the learner's education. In addition to the written copy, the oral sample is also given and echoed. The spoken pattern can be pronounced by the name of the letter "I" (as / el /) or by its pronunciation (as / I /). The student repeats the verbal pattern during written copying, matching the written and verbal forms of the letter "I". Before and after copying the word "lili", the S system pronounces the word and the student repeats it as an echo.

To further facilitate the copying of characters and words and increase the number of geometric reference points, the area of the sample frame and copy frame can be provided with a geometric grid 40 as shown in Figure 4, thereby allowing the learner to relate not only to the horizontal train 31 and frame 33 also for grid elements. The geometric grid 40 will then be removed by the S system when the results of the exercises detect that the learner is capable of copying the characters correctly without the grid 40. The 33 frames may be dropped when the student has actually learned handwriting, but reading is still being taught and the student is also doing handwriting exercises.

Figure 4 is a snapshot of an exercise, the current sample element is the letter "a" of the word "purple", and the written copy has begun in the 33 frames below. Meanwhile, the oral pattern of "a" may be given and copied as explained in Figure 3, and the written and oral form of the letter "a" will be matched.

Detecting the difference between sample and copy can be accomplished using a relatively simple pattern matching procedure. Such pattern alignments are known, for example, in Automatic Handwriting Recognition (AHR) procedures, which are nowadays common to all handheld machines with such features.

However, the use of the 33 frame not only provides the opportunity for better font shaping and visual self-checking, but also significantly simplifies the comparison and evaluation of the digital image of the written sample and its copy, voluntarily offering the possibility of 32 samples and 34 copies. 6B. The digital virtual grid 42 of FIG. 4A is measured to determine whether the lines of the copy 34 pass through the same grid elements as the lines of the sample 32 and evaluate the copy quality based on the coincidences.

The virtual grid 42, which is not the same as the geometric grid 40, can be used not only to measure the difference between sample 32 and copy 34, but also to automatically adjust the beautiful grid for a given student based on analysis of a certain amount of copy practice. requirement level by controlling the density of the virtual grid 42. This is because not all students are able to learn the script to the same standard. For the weaker, the requirement is not the perfect appearance of writing, but the recognition of characters. For them, they rarely have 42 virtual grids, which means a lower spelling requirement level.

The human eye and the human ear are extremely fine instruments that are particularly sensitive to the difference. We use this natural endowment as a methodological element in our teaching of copying literacy patterns. As we have seen above, by stacking the 32 samples and the 34 copies that have just been made, the learner is able to judge the deviation from the pattern and use this experience in the next copy attempt. In pronunciation exercises, the situation is similar; when spelling punctuation, word elements, and words, the learner receives immediate feedback on the difference, because the S system corrects it immediately if the pronunciation fails. In any case, verbal copying takes place in such a way that the learner echoes the verbal pattern given by the S system. In the learner's ear, there is an oral pattern provided by the S system and his or her own pronunciation attempt, making it easy for him or her to distinguish and correct.

After learning a subset of the alphabet, the "minimal transition" method provides a functional solution for moving forward, which means that we learn to add another character to a word learned by changing or adding one character to introduce another character to expand we want a subset of the learned characters (for example, figures 3-4, moving from "lili" to "lilac"). The minimal difference in written and phonetic between the old and the new word in the double pattern practice clearly illuminates to the learner the appearance, sound and meaning of the new character. With the gradual expansion of the initial subset of the alphabet, the number of words that can be depicted increases significantly, enabling the inclusion of meaningful sentences, then small stories, in education and visualization.

Dual pattern exercises, by their very nature, do not provide the necessary writing and reading skills

EN 226 347 Β1 skills, so there are also practices that will gradually lead to independent writing and reading. Some important approaches to the transition from dual pattern to independent writing-reading in machine learning are also based on difference building. One of these is the "missing link" type of exercise. Missing link means that when repeating a double pattern exercise, the written or verbal pattern is missing some element, such as a letter, a letter, a sound, and the learner's task is to fill in the gap based on the memory of the previous exercise. Through these exercises, the skills learned are recorded and the learner is forced to make greater efforts to match the written and oral patterns.

The real transition to independent writing and reading is the repetition of double pattern exercises with a single pattern. Take, for example, the transition to reading. In the previous dual sample practice, the S system read out the words and the sounds of the written sample, and the learner echoed the S reading while observing the written pattern. During the repetition of the exercise, only the written pattern remains, the student's task is to read the written pattern. In fact, you only read half of it, rather, just copy from memory the complementary pattern you noted during the duplicate pattern exercise. However, if he would have difficulty recalling words or characters in sequence, the S system will help him, for example, by whispering, or verbally correcting the learner, if he has misspelled a character or word. Single patterned exercises can also be made difficult, for example by arranging the words to be read in a different order than the double patterned exercises.

The method of gradual transition to self-writing is similar: after a double pattern exercise, dictation follows, that is, repeating the same practice without leaving the written pattern. Although the Written Pattern is not present visually at present, the learner still remembers it from the previous exercise, and a blank frame of the same size as the dictated word may help. If writing dictation is difficult, however, the S system will provide additional assistance by displaying the 33 frames of the word characters within the word frame or by flashing the word for one second at the copying position (written help).

If self-reading practice is difficult even with the help provided, you can return to the dual sample practice and try again for independent writing or reading. This is a drag and drop game that can be built into the curriculum, but can be extended through feedback between exercises depending on the individual student's performance.

Each of the methods described above is based on distinction. In general, this means that a slight deviation from one practice to the previous practice (and the effort enforced by the deviation) results in the knowledge mediated by the difference being easily absorbed into the mind, and the learner is a little step closer to independent reading literacy. . As a result of taking many such small steps, you will eventually reach the level of being able to read and write independently and confidently.

The natural division between speech and writing is provided by words. At the same time, framed copying provides the ability to decompose the writing-reading process into punctuation level, schedule timed copying of frame-defined characters, to such an extent that a character can be copied simultaneously in writing and verbally, mutually matching the written and verbal forms. and memorizing the character. This method is effective primarily in the initial phase of learning. Framing words or parts of them in reading and pronunciation exercises is also a useful solution because it can be used to learn different pronunciation situations.

Copy per character! the sequential sectioning of the text later, during the copying of words and dictation, prevents continuous writing, so at this learning stage it should be possible for the learner himself to dictate the copying speed by continuously copying the letters into the 33 frames.

In the advanced phase of writing learning, once the student has learned the letters, copied the written pattern well, and is able to write dictated text, the use of 33 frames can be gradually discontinued. At this stage, only the natural segmentation of the text by words remains, but all practices should be limited in time.

Since the student is usually a child between the ages of 6-8, an automatic system for reading and reading should be designed for them. The primary requirement is for the S system to be student friendly and not to arouse resentment in the child. Since most students have been watching TV since they were young, playing various electronic games, it is not a strange activity for them to watch a screen while manipulating the S system. In school education, motivation for students is rather poor compared to TV and electronic games. The colorful alphabet book and the worksheets must not compete with fairy tales, cartoons or electronic games. At school, it is largely up to the teacher and at home to the parents to somehow persuade the children to take active interest and learn. In machine learning, a multimedia world needs to be created around educational functions that can motivate the learner to participate actively in the learning of the subject. In this way, the education is embedded in a series of interrelated short stories, with the student actively participating in each story by solving a task. The tasks, of course, are literacy exercises. Tasks can be done within the story, or by interrupting the story (as is customary in TV commercials) as a prerequisite for continuing the story. The animated story7

Net1 The net itself can serve as the topic of the exercises, a frozen picture of the story can provide context. Context means that written and spoken patterns are assigned images and sequences of images that illuminate their meaning, just like a colorful alphabet book. In the solution outlined, the context is ideally combined with motivation, since the images of the motivating story themselves are linked by their meaning to the sample words and sentences. Exercises must be completed online, or off-line, by the student during the exercise as the short story progresses through the demonstration interface of the 17 digitalization screens. While performing the exercise “online” or “off-line”, a typical still image of the short story can be displayed in the demonstration area of the screen, which serves as a context for the words or sentences in the exercise.

A 6-8 year old is not yet able to receive textual information, only verbal instructions and visual explanations can be provided. The motivational world should be structured to guide the student step by step into the world of letters. This will tell you what to read and write, how to write on the screen, how to hold the pen, how to angle, how much to press, what to posture properly, how to hold your hands, and so on.

The story then begins with the first exercises to introduce the writing and reading in more detail through the relationship between the written and oral patterns and the duplication of the pattern.

Each new exercise is introduced by a brief explanatory section, in which the story teacher presents verbally, in stills, and in motion pictures what the practice consists of, and how to complete the task. The presentation should preferably be visual in order to bring the learner to life in every detail, and the purpose is for the learner to experience as if he or she were to do the exercise.

After the presentation of the exercise, the exercise begins. In the meantime, if the S system detects that the learner is unaware of the task, it begins again from the beginning and the exercise is presented again. The copies made during the exercise are compared by the S system to the samples and, on the basis of this, at the end of the exercise an evaluation is made to decide whether it is possible to continue the curriculum or apply some feedback to an earlier learning phase.

At the end of a well-done exercise, there is a reward, in the form of verbal praise, animated jokes and / or music, and the continuation of a story in which new adventures and challenges await the learner. After poor practice, it is expedient to show encouragement and a relaxing element.

Education takes place in the classroom or out of school, in stages, in the form of lessons. Community machines will be used by another student at the end of the lesson. In this case, the start-up begins with the identification of the learner and continues with the part of the curriculum that was interrupted by the learner during the previous use.

The computer training system S shown in Figure 1 consists of the following modules.

- 11 curriculum modules including written and oral copying exercises, practice motivational visuals in the form of still and / or moving images, written and oral samples to be copied during exercises, and presentation and evaluation of exercises serving verbal information, stills and movies in the form of appropriate data, image, audio and video files. Exercises are stored in a training program execution order, just like a computer program with instructions.

- 12 demonstration modules that use the files of the curriculum module to visualize the motivational and explanatory imagery for the next exercise, verbally, in still and in motion, and present the exercise in a way that reveals to the learner every detail of the exercise as if he were doing the exercise himself.

- 13 Execution Modules that launch the presented exercise, display written patterns to be copied in the digital image work area, verbal patterns during the exercise, schedule and direct the copying of patterns, detect student's written and verbal replicas, compare replicas to patterns and determines the degree or fact of discrepancy on each sample element, places the pattern on the copy in a different color, repeats the pattern to correct incorrect verbal copies. The module also passes the discrepancies found between copies and samples to 14 evaluation modules.

- 14 Assessment Modules, which assess practice based on the extent and trend of pattern and copy differences and, taking into account the results of previous exercises, decide whether to continue the training program or repeat the exercise, or to return to an earlier exercise.

The 14 assessment modules verbally, in the form of still and moving images, provide the learner with additional information for repeating or retreating, praising or encouraging them verbally, in the form of moving and still images, possibly linked to the motivational material of the exercise.

The 14 assessment modules collect the results of the student's exercises in a database and identify the exercises on the worksheets, and store them in the form of messages. The messages should preferably be accessed by a password supervisor.

The process of performing each exercise in Figure 2 is summarized as follows.

The presentation module 12 includes, in the course of 20 steps, a short story in motion and still images,

EN 226 347 Β1 and verbally provide motivation and context, presenting the practice in pictures and verbally.

Execution module 13 displays (reproduces) handwritten and / or verbal patterns according to the practice described in step 21, controls and schedules the copying of the patterns.

Execution module 13 detects the student's written and / or verbal copy in action step 22, converts it to the appropriate digital format, compares it with a sample, finds and records the degree of discrepancy, places the sample in a different color on written copies, incorrect verbal copies repeats the pattern to correct it.

At the end of the exercise, the assessment module 14 evaluates the exercise results from the magnitude and trend of the differences identified by the executive module at the end of the exercise, and then, in the 24 steps, determines what the next exercise will be and of the following, and in an action step, proceed to the next exercise in the curriculum, in an action step, repeat the exercise under unchanged conditions, in an action step, repeat the exercise under modified (eased or aggravated) conditions, or in an action step go back to a previous lower level exercise.

The patterns are derived from the spoken language, with the smallest unit being the word. Duplicate patterns consist of the written and oral forms of words, the written pattern being complemented by the oral pattern and vice versa. The curriculum consists of exercises to be performed sequentially, the patterns in each exercise are always closely assigned to the exercises in the form of a duplicate sample and form a vocabulary of patterns in this structure. System S thus has an oral complement for each written sample and a written complement for each oral sample to perform the assessment of completing the exercise. Characters that make up words and their corresponding vowels can serve as a double pattern in practice. In addition, each word in a written pattern of an exercise may include several types of spoken patterns, spoken and spelled and spelled. There are two spoken versions of a character, such as the alphabet name of the character (to spell words) and the spoken word (for synthetic reading) of the word in question. Words can serve as examples of pronunciation exercises not only in their entirety, but also in framed parts of the written pattern (Id. Fig. 8).

Written patterns in exercises can consist of one or more pattern elements, pattern elements can be the same or different characters, the same or different words, characters and words mixed. The words and characters that make up the pattern are called the elements of the pattern, and the pattern element that you want to copy in practice is called the current pattern element. Of course, the same pattern can be used in several practices, depending on whether it is only written, only verbal or simultaneous written and oral copying.

The syllabus module 11 preferably contains the following types of written and oral sample elements in digital form:

- written patterns of characters in different fonts and sizes, a pattern highlighting some features, a pattern without some features;

- spoken patterns of characters, pronounced according to the names and sounds of the characters;

- written patterns of words such as complete words and incomplete words, highlighted parts of words;

- spoken patterns of words, such as words and parts thereof pronounced, spelled, pronounced and pronounced;

- written and oral patterns of sentences.

The syllabus module 11 includes exercises that give you double patterns between exercises, which are characterized by the combination of written and spoken patterns of characters and / or words, and the student's role in copying one or both patterns and correspondence and identification of written and oral sample elements.

Duplicate sample exercises can occur in the following versions:

- duplicate samples, written and spoken,

- duplicate sample, oral copy,

- duplicate sample, written copy.

The syllabus module 11 also includes exercises in the form of simple patterns in which only a written or verbal pattern is displayed, and the learner's task is to read (pronounce) or describe the oral (dictated) pattern.

Single pattern exercises can occur in the following variants:

- written sample, written copy,

- written pattern, reading,

- oral pattern, written copy, ie dictation writing,

- oral pattern, verbal copying (for example, to emphasize stress).

Depending on the types of sample elements, the following types of copy (write, read) operations may occur in practice:

- copying of written and spoken patterns of characters, such as copying characters, overwriting characters or some of their features, overwriting characters with moving highlight, completing incomplete characters, dictating characters, naming characters, pronouncing sounds of characters,

- copying of written and spoken patterns of words such as written copy of words, overwrite of words, overwrite of words with moving focus, overwrite of characters or features highlighted in bold word, addition of missing words

EN 226 347 Β1 characters or traits, writing words for dictation, writing words for dictation per letter, reading words, reading words syllabically, pronouncing words and parts of words, by sound and by letter,

- writing and reading sentences, such as writing copies of sentences using punctuation, writing sentences for dictation, using punctuation, verbatim copying sentences, reading sentences.

There are two types of exercises that can be used to schedule copying:

- continuous exercises (eg written copying, reading of words, sentences), preferably with a progress bar, indicating the time remaining to complete the exercise, and

- intermittent exercises, where the student has at his / her disposal a T-time for the written and / or verbal copying of each sample element (usually a character), after which the next sample element should be copied.

The manner of staging / scheduling the copy depends on the type of exercise, with the time value T and the depletion time of the consuming light column being set by the evaluation module 14 for each exercise, depending on the result of previous exercises.

The use of the dual pattern is important especially in the initial phase of learning, when the emphasis is not on continuous writing, but on the graphic design of characters and their identification or memorization with their sound or names. At this stage of the learning process we start with meaningful words, however, copying the words is done by dividing them into characters, and they are verbally copied, identified and memorized.

Sample writing using the frames 33 can be done as follows.

In handwriting exercises, the learner makes written copy on the surface of a flat, digital 16-inch digitalization image with 16 pens. The workspace 30 of the screen is provided with a regular line 31 in school notebooks, such as double-line 31 for cursive writing and for Chinese writing with Chinese lines. In the workspace, two rows are lined up, the top row is the row of patterns 32, here are the handwritten font 32 patterns, and the bottom row is the row of copies 34. It is the student's job to copy the characters that make up the sample 32 to the 34 copies, one by one, while also verbally copying.

Each character forming the pattern 32 is provided with a rectangular frame 33 which limits the size of the character. In the copy line, blank frames of exactly the same size are placed under the frames of the characters to be copied, and here the student must copy the 32 patterns. The vertical size and position of the frames 33 are adapted to the characters of the language, for example, in the case of the Latin alphabet, it allows the largest letter, "f," to be accommodated. In words, adjacent character frames are matched, and in italics, the matching line determines where the characters join. For a different language alphabet and writing (such as Chinese Kanji writing), the frame size and position may change, but the principle remains the same: the frame delimits the characters well and creates appropriate graphic conditions for more accurate copying of characters. However, the application of the frame 33 goes beyond the strictly graphic-copying task by providing the ability to control, schedule, and evaluate written and oral copying.

In written copying (and later will be seen in oral copying), the S system uses the 33 frames to specify the mode and rate for copying the characters, the T time that can be used to copy each character. When the T time has elapsed while copying a character, the next (current) character in the copy will appear in a color or line width contrasting 33 ', indicating that the student must begin copying the current character regardless of the blank 33 below it. whether you have finished copying the previous character. The characters 32 of the pattern 32 must be copied accurately in shape and size, and at a rate dictated by the S system, into the blank frames 33 in the copy line. A flowchart of the operation steps 60-70 of the stepwise written copy of a written pattern is shown in Figure 13.

As shown in Figure 3, the position of the characters is precisely positioned within the word within the frames, so that during copying the characters do not slide so that the length of the word changes, making it impossible to compare with the 32 patterns. When copying, one character can slide to the right or left to the edge of the frame 33, this minimal amount of misalignment is tolerated by the S system during the evaluation of the copy.

To set the copy rhythm, follow these steps. The timing, the time T for copying the current pattern element, is always indicated by the appearance of the opposite frame 33 'on the next (current) pattern element to be copied. At the beginning of the education, the time module T for copying the current sample element is set longer by the 14 assessment modules, and later, the student automatically sets shorter values as the student progresses. Conversely, if the S system detects that the learner is copying at a slower rate than the scheduled schedule (making incomplete or poor quality copies), it will automatically slow down the scheduling.

The use of frames 33 allows for a visual representation of the difference between pattern 32 and learner replicas 34 as you practice. When the T time for copying a given character expires, the pattern will appear on the pattern 34 with a contrasting line so that the student can compare the pattern with the copy and use the experience of copying the previous character to copy the next character. The effect of visual comparison may be enhanced by the S rating of the copy, for example, by providing a good copy frame10

EN 226 347 Β1 colors the area green, acceptable yellow, unacceptable red.

If the evaluation module 14 detects from the results of the exercises that written copying is difficult, the copying of characters can be facilitated by the geometric grid 40 shown in the workspace 30, shown in Figure 4, which further enhances the geometric definition of the character provided by frame 33. Copy 34 is good if the proper features of the character pass through the same grid elements as in the pattern within the frame 34. The geometric grid 40 can also be used in exercises with and without frames 33.

When writing on worksheets 19, also known as worksheets, as shown in Figure 7, the learner repeats the same exercises he or she has already done on the digitizing screen 17 to get used to writing on paper. Each exercise sheet matches a specific exercise, with exercises on the screen and worksheets being identified by the same number or symbol so that they can be matched. The worksheets are printed in the same manner as the workspace 30 of the digitizing screen 17 except that each character of the pattern 32 is pre-encapsulated within a frame 33 of the pattern 32. Exercises 13 lead the exercises in exactly the same way as a learner would write on the screen: frame the on-screen pattern to frame the current pattern element, say an oral pattern, and the learner should copy the written pattern, but not the screen. sheet. These practices are verified by the parent or teacher based on reports from the 14 assessment modules. The figure shows 19 worksheets on which the student has not started copying.

In the early stages of learning, as in school practice, the copy must be made on the 32 samples themselves (this overwrite) so that the student's hand is accustomed to the character form. Overwrite can be used later to practice the attachment of characters and the continuity of writing. These exercises do not, of course, require the use of a copy line, and the copy should be made at the rate at which the 33 frames appear in the pattern characters. The pen strokes that make up the characters can be practiced individually, for example, by simply overwriting one of the features highlighted in the outline of the pattern mark, or by adding an appropriate pen stroke to an incomplete character. In machine learning, we use an advanced version of the known school practice of overwriting exercises, where large-sized characters are prepared with dots and small arrows to guide the learner on how to correctly formulate the character, the orientation of the features, and the order in which they are written. In machine learning, the learner's hand is not guided by arrows to form the character, but is preferably a moving light 41 shown in Figure 5, moving at the same speed as the learner guides the 16 pen, but always slightly before the tip of the 16 pen. . The figure shows the overwrite of letter "A" in four phases, illustrating the order of overwriting each stroke.

The overwrite exercise can be followed by a standard copy exercise in which the student copies the pattern characters previously known and practiced in the overwrite into the frame of the copy line, with or without the help of a graphic grid 40.

The written copy should preferably be evaluated as follows. The automatic writing recognition methods are not used in processing the copies written in frames 33, since the use of pattern 32 eliminates the need for writing recognition. In our system, it is not a matter of recognizing an unknown unknown letter, but of the extent to which the copy 34 differs from the sample 32. However, the use of the 33 frames not only provides the opportunity for better font design and visual comparison of the sample and its copy, but also a simple and effective method for comparing and evaluating the digital image of the written 32 samples and 34 copies, grid application. The virtual grid 42 (which is not the same as the graphical grid 40) is graphically illustrated in FIGS. 6A-C. are illustrated illustrating the evaluation of the just-described copy of the letter "a" of the word "ali." By placing the same virtual grid 42 on the pattern and copy of letter "a", we examine whether the lines of the copy 34 pass through the same grid elements as the lines of the sample 32 and evaluate the copy quality based on the coincidences. During copying, characters can slide to the left or right to the edge of the frame (see slide 6A and 6B to the right for a copy of letter "a"). This slippage is tolerated by the S system to the extent of a virtual grid element such that 6C on the copy of the note. Figure 3 shows three measurements. For the first measurement, slide the virtual grid 42 of copy 34 to the left relative to the virtual grid 42 of sample 32, then slide it to the right before each measurement, finally accepting the measurement result for which it found the most overlap (shown in Figure 6C). this will be the first measurement). Of course, the virtual grid 42 of the replicas 34 can be shifted not only by one grid element, but also, if necessary, in a vertical direction.

In the computer-implemented realization of the measurement using the virtual grid 42, the virtual grid of the sample and the copy is preferably transformed into a binary matrix, each of whose elements may have a value of 0 or 1. The value 1 is given to the elements of the matrix that correspond to the active elements of the virtual grid, that is, those that pass through the sample (or copy) line. By combining the binary matrix of the sample and the copy in a logical AND relationship, we get how many elements they have in common, that is, how many active elements of the copy grid overlap with the active elements of the sample. By comparing the number of elements in the overlay to the number of active elements in the pattern (that is, 1 in the pattern matrix), we get the extent to which the copy of the character overlaps the pattern. The matrix

The same method is used for measurements by sliding the copy frame in the same way as above.

Preferably, the spelling requirement level can be set as follows. Virtual grid 42 can be used not only to measure the difference between copy 34 and sample 32, but also to automatically adjust the spelling requirement level for a given student by adjusting the density of virtual grid 42 based on the analysis of a certain amount of copy practice. This is because not all students are able to learn the script to the same standard. For the weaker, the requirement is not the perfect appearance of writing, but the recognition of characters. For them, the machine sets a rarer grid, meaning a lower level of handwriting requirement.

The density of the virtual grid can be independent of the line thickness of the sample 32 and the copy 34 by thinening the line thickness toward the theoretical centerline of the line and smoothing the resulting virtual line by interpolation.

In the next phase of writing learning, once you have acquired a certain amount of writing skills and character recognition, you no longer need one character per character! for batch copying or copy scheduling. After a while, the student begins to be annoyed by the fact that he is not able to imi-words continuously because the S system interrupts the practice by characters. This condition is sensed by the evaluation module 14 as a result of improved writing quality, no incomplete, missed punctuation, and improved auto-scheduling rhythm due to improved results. Detecting a certain spelling level and copy speed, the 14 rating modules per character! switching from rigid control of copy scheduling to exercises for a freer copy style by omitting the highlighted 33 'frame of the current character to be copied, allowing the learner to determine the copy speed within certain limits. In these exercises, the remaining time is indicated by a consuming light column 50, the time of which is determined by the evaluation module 14 based on the results of previous exercises. Processing of a copied character begins when the execution module 13 detects that the learner has begun copying the next character of pattern 32 by entering the pen area 16.

Copy by example! Figure 10 illustrates the process. Each character of the pattern is enclosed in a frame 33, and the same line of frame 33 is placed in the copy line to accommodate the characters to be copied. At the moment depicted, the learner starts copying the letter "v" of the word "have", continuously after the description of the letter. The time remaining until the end of the exercise is indicated by the depleted 50 light columns. An already made copy of the word "we" has been added to the 32 samples so that the student can judge the difference between the 32 samples and the 34 copies.

Dictation is a peculiar written copy. In dictation practice, there is only a verbal pattern, but as each word is uttered, 33 copies of the dictation word will be displayed in the appropriate number and size to accommodate the dictation word. When you say the next word, you must stop typing the previous word and start writing the new word. When dictating, the S system will have a (complementary) written sample for the dictated word, comparing the word written by the learner to the blank frames 33 using the illustrated evaluation method for deviating from the written copy pattern: after describing each dictated word, the moment before the next word is spoken, a written pattern of the dictated word appears and a copy is planted on it. Figure 16 shows a flowchart of the operation steps 90-101 of a "Written Copy with Frames" exercise. These dictation practices not only serve to practice writing, but can also be considered a reverse pronunciation practice, since the spoken word must be spelled correctly. The 33 frames given to the dictated word help the student a lot.

The choice of the handwriting font used to represent the sample words and characters can be arbitrary and depends only on the language alphabet and the requirements specified in the curriculum. In the examples we use the cursive font "VA Pe 2" (Copyright © Will Softwareentwicklung, D-35641 Schöffengrund), which can be used to teach handwriting in the Latin alphabet.

The use of 33 frames and the requirement for form-fitting and dimensionally graphic copying of characters allowed for a simple grid based on 42 virtual grids. However, in dual-pattern and subsequent single-pattern exercises, the learner reaches a level of literacy that allows them to start gradually phasing out the boundaries of the 33 frames, so that the learner eventually reaches paper-free writing. . The first step in unlocking is to leave the staggered copy schedule as described above, after which the character copy is no longer tied to the schedule time. 33 frameless writing provides even greater degrees of freedom, which at first means "written pattern-writing copy practice," and later describes the oral (dictated) pattern. The use of certain elements of automatic word recognition can no longer be avoided for the evaluation of the 33 frames.

The function of automatic handwriting recognition (AHR) is to recognize unknown text from unknown handwriting and reproduce it as printed text. This is a difficult pattern-matching task, because every letter in the handwriting has to be compiled and identified, and every word in the letters has to be inferred, resulting in one letter in doubt situations due to the redundancy inherent in the words.

AHR is based on a dictionary of characters and words with their lines, shapes, and other geometric features (such as line 12).

EN 226 347 Β1 many intersections, point locations, etc.). The digitizing screen 17 detects the movement of the pen 16 and records the x, y coordinates of the line described by the pen 16. This is followed by pre-processing of the x, y coordinates of the line (equalization, extrapolation, and prior space-time sampling), then normalization, that is, correction of lines by size, inclination, and curvature, and finally identifying the line based on the lines stored in the dictionary. As a result of the operations, the characteristics of the described lines are created, which by comparing the pattern characters in the dictionary, their relative position, and other characteristics, recognizes the characters and recognizes the words based on the words recognized in the dictionary.

The present invention is not about general purpose writing recognition, because the task is not to recognize an unknown text, but to examine the extent to which the copy matches the pattern in shape and size. This is also true for dictation, since the dictated spoken pattern includes a complementary written pattern, and the S system compares the characters to those described by the learner.

The practice of Borderless Copy without maintaining the machine-based evaluation of copies can be accomplished by applying certain elements of the AHR as follows.

The visual comparison of the copy and the sample can be done as follows. At a given level of education, the sample consists of written or verbal sentences or words. When the student copies (writes) words based on a written or spoken pattern, System S places a copy 34 on the 32 after the description of each word so that the student can visually judge how accurate the copy is and how itself to the requirements.

An example of a borderless written copy 33 is shown in Figure 11. The figure represents the moment when the student described the letter "p" of the word "pen". The copied words "we" and "have" were added to their own pattern to allow the student to judge the difference between the 32 samples and the 34 copies.

Evaluating the words you write is already a difficult task. The simple and effective virtual grid method cannot be directly applied in frame-free copying because, when copying, characters can be shortened or lengthened horizontally, and thus the word copied can be shortened or lengthened compared to the 32 patterns. Another problem is that, in the absence of frames 33, it is not easy to determine the connection points between characters, i.e., where a character ends and where the next begins. However, the task of the S system itself is to compare the sample words with their copy and evaluate the copy quality of each word, which is much simpler than general-purpose writing recognition. The point of benchmarking is that the S system separates the first, second, ..., etc. words of the sample word when examining the word copied by the learner. and compares the first, second, ... etc of the copy. writing of tickets. This is a typical pattern matching procedure, an integral part of AHR procedures. The ability to compare is created by preprocessing, which involves positioning the copied character on the pattern character (similar to the way slip tolerance is realized), seeking maximum coverage between the 32 patterns and the 34 copies. If necessary, the S system will zoom in or out on the copied characters to a minimum (tolerable) but in a form-consistent manner to approximate their size to that of the sample characters. The actual comparison can then be made using a method similar to that of the virtual grid 42. In doing so, the S system determines the degree to which the copy of a given character resembles the pattern character in form and size. This value is noted by the S system and, at the end of the exercise, gives an estimate of the total word. If the student misses something in practice, such as skipping or swapping a character in a word, or misspelling a word, the S system will indicate this prominently, for example by making the copy field blink or discolor.

You can switch from dictation to practicing borderless copy writing. This practice inherently lacks a written pattern, but requires the student to write each character in a font that matches the patterns given in the previous framed and frameless copy exercises. When performing the exercise shown in Figure 12, after writing each dictated word, its written pattern appears in the pattern line, with the S system placing the learner's word opposite it. The evaluation is the same as in frameless copy, except that the word being described is not compared to the pattern, but to the complement of the dictated word. If the S system detects that the learner has difficulty writing, he or she flashes the written form of the dictated word in the sample line for one second.

Figure 12 is a dictated version of the exercise shown in Figure 11. Practice is dictated by the pace of dictation in time: the learner must begin describing the dictated word, even if he has not completed the previous word. The figure captures the moment when the student writes the dictated word "pen" after having written the words "we" and "have". When a word has been described, the S system places the written complement of the dictated word in the sample line and places the student-made copy on it. A flowchart of the operational steps 82-89 of the exercise is shown in Figure 15.

Staged verbal copying using an echo technique can be done as follows. In framed exercises with frames, in addition to writing the sample in writing, the learner also copies the oral form of the sample to practice matching between written and spoken words, and between characters and sounds in words, which is a prerequisite for independent reading. An effective form of such exercises is the 'echo technique', in which the S system reads' the actual sample element marked by framing on the written pattern. The current pattern13

The element Β1 may be a word or part of it, a character or a separate character. It is the student's responsibility to echo all words, sounds and letters spoken by the S system as echoes. An additional function is added to the role of the frame by marking the current sample element to be copied verbatim (or verbatim). In practice, the following oral sample element input and copy-read variants can be implemented for the current sample element marked with a frame on the written sample:

- Duplicate pattern, duplicate copy. With contrasting framing, the S system also adds a verbal pattern to the written pattern elements highlighted in series, the student must copy both patterns in writing (copying or overwriting) and verbally copying the verbal pattern of the S system. This is typically an exercise that needs to be broken down by machine scheduling. If the student makes a mistake in verbal copying, the S system corrects it immediately (but counts the error in the practice evaluation), giving the student the opportunity to copy the verbal sample correctly. Figures 3 and 4 can serve as an illustration of the double pattern, duplicate practice, by mapping the graphical representation of the oral system of written pattern elements marked with a highlighted frame at highlighting. The flowchart of the practice steps 71-81 is shown in Figure 14.

- Duplicate pattern, oral copy. There is a significant difference in the speed of written and spoken copying, so in dual pattern exercises, once a certain level of writing security has been achieved, it is possible to dispense with duplicate writing so that the learner is primarily used to match and memorize written and spoken forms , focus on pronunciation. This exercise is a variation of the previous exercise where you do not need to copy in writing, only echoing the verbal pattern while observing the written pattern element marked with a frame. The frame jumps from one sample item to the next as the exercise progresses.

Figure 8 illustrates a "double pattern verbal copying" exercise. When the words are fully framed, the whole word is pronounced so that the learner echoes it. When partial framing, only part of the word or only one character is pronounced so that the student will repeat this pattern only. In the practice of presenting synthetic and global reading, the S system uttered the words or the highlighted parts before and / or after reading each character. In the illustrated exercise, the first and third words "teach" are framed in their entirety, and the second word "teach" is per character. It is the student's task to read one by one the sample element marked with a contrasting 33 'frame. In the situation shown in the figure, the opposite frame is exactly the letter "a" of the second word, and this is the letter the student should name. Before that he had pronounced the first word "teach" and spelled the first two letters of the second word.

Similarly, the syllable reading of words is introduced, in which case the syllable currently being read by the S system is marked with a frame, and the learner echoes the syllable.

Pronunciation practices are of increasing importance in languages where there is no clear correspondence between written and oral forms, such as in Hungarian or Russian. In pronunciation exercises, the pattern is made up of words or their highlighted parts, for example, in Figure 9, two accents are pronounced in tEAch and teaCH, respectively, when pronouncing elements of the English word 'teach'.

The first word "teach" in the practice snapshot is framed, pronounced, system S provided and echoed by the student. In the second word "teach", the pronoun "ea" marked with a frame has been pronounced / / /, first by the machine, then by the learner, as well as the pronunciation of the letters "e" and "letter". second "teach framed" eh "part, the lyrics phoneme has already been uttered, the S system is just dropping the name of the current sample element, the letter" c "(precisely framed), so that the student will echo this. To pronounce the letter h.

The process of frame-controlled reading is as follows. In echo-type double pattern exercises, the student was given reading patterns, learned to identify characters, and practiced examples of synthetic and global reading. After completing a certain amount of practice, you can begin to practice independent reading with staged reading exercises, where the current sample item is framed and there is a time to read each sample item. As in school, there are two types of reading practice: "written pattern reading" and "written pattern writing and copying". In both exercises, the learner reads the sample items marked with the frame in turn, and in the second exercise, copies the written pattern while reading. When you have successfully described and read the current sample element, the frame skips to the next sample element. Machine verification of reading accuracy is based on the complement of the written pattern. During reading learning, the S system assists the learner by reading a word if he or she senses that he or she is reading incorrectly or is unable to keep to the reading schedule defined by the frames. Help is in effect a partial return to echo-type exercises.

Figure 8 also serves as a graphic illustration for this type of exercise. The S system indicates to the student by framing the written equivalent of the word, letter combination, sound or letter being read. A similar method is used to read words syllabled, in which case the syllable to be read is marked with a frame.

Pronunciation exercises continue between reading exercises, in which case words or word elements marked with a frame are no longer read out (pronounced) by the learner, after memorizing the pronunciation in the previous "double pattern verbal copy practice" . The single patterned pronunciation

In practice, the pattern is made up of words or parts highlighted by framing, Id.

A flowchart of the exercises for the “Copy verbatim written pattern” exercises is shown in Figure 16.

An example of taking an oral sample and processing an oral copy! course is as follows. Computer-assisted teaching of handwriting and reading raises the issue of automatic speech recognition (ASR). ASR can be defined as the computer-based transcription of spoken text into readable text in real-time mode. In very simplified terms, ASR works as follows. A microphone converts speech into analog electrical signals, which are converted to digital form by the ASR program, then pre-processed, filtered and compressed by stored information. The essence of ASR consists of a dictionary in which punctuation and words, in addition to their written form, are "in the form of speech models, composed of phonemes, an acoustic model linking words and speech models of spoken text to be recognized, and a linguistic model arranges the recognized words into meaningful sentences. During acoustic processing, ASR decomposes speech into phonemes, matching them with dictionary phonemes, and then sorts the phonemes in sequence to search the dictionary for the word corresponding to that phoneme sequence.

ASR is a complex pattern recognition program. It is important to note that the present invention does not require the use of an ASR system for general purposes, since our teaching method is not to recognize an unknown spoken text, but to distinguish between written and spoken patterns and student copies.

Oral samples are taken on a practice-by-practice basis as sample items in the variations described above. Oral samples are recorded by a (childlike) announcer into a microphone and thus converted into analogue signals by a digital transformation such that they meet three requirements: (1) each verbal sample element is assigned to its own written sample element, (2) the recorded verbal element the sample element is physically reproducible by digital-to-analog conversion during the exercise, (3) the oral samples are stored in a digitized format that allows for comparison with the oral copy of the learner's words spoken to the microphone and similar to the sample recording. Said digital transformations are standard procedures and include ASR elements used in speech perception and digital conversion, such as analog-to-digital conversion, prefiltering, phoneme decomposition, and the like.

Reproduction of oral samples is the reverse of sample sampling: the S-system reproduces the digitized sample through a headphone or loudspeaker after a digital-to-analog conversion.

During processing and evaluation of the student's spoken (read) oral copies, the task is to determine if there is any discrepancy between the given oral sample and the student's copy. Note that in plain text reading exercises, although the oral pattern is not spoken, the S system has a (complementary) verbal pattern for the written pattern to be read, thereby comparing the word read by the learner.

The verbal copy is processed as soon as the learner has spoken (read) the given sample item. Processing consists of two steps. In the first step, the S system detects the learner's oral copy and, after a preprocessing digital transformation, converts it into the same format as storing the oral samples. In the second step, the digitized copy is compared with the digitized sample, using one of the standard ASR procedures, the digitalized sound sample comparison method. In doing so, the S system identifies the first phoneme on the digitized copy and examines whether it is identical to the first phoneme of the sample. If so, it identifies the second phoneme of the copy and compares it with the second phoneme of the pattern. In principle, this test consists of as many steps as there are phonemes in the sample. If one of the phonemes in the copy does not exactly match the corresponding phoneme in the pattern, System S stops further examining the copy and corrects the learner, ie repeats the verbal pattern so that the learner can repeat it. However, the error is taken into account when evaluating the practice, as it is the basis for deciding whether to continue with the program according to the program or to repeat the practice or to return to a previous practice.

Here's how to evaluate your exercise and choose your next exercise. We have seen above how the S system evaluates written and verbal copying of sample elements. Each exercise consists of multiple sample elements, the execution module evaluates the copying of each sample element. At the end of the exercise, the evaluation module evaluates the entire exercise from the evaluation of the copy of the sample elements and decides on the following exercise:

- next practice according to the curriculum,

- the repetition of the exercise which has just been carried out under the same or modified conditions,

- a return to an earlier exercise.

Making such a decision is a difficult task, because it is necessary to take into account the results of current and previous exercises as well as the tendency in the results. The 14 assessment modules are in fact an expert system because you will have to decide between the following exercises and the conditions under which they will be performed:

- repeat a double pattern exercise with a single pattern,

- repeat a single pattern exercise with a double pattern,

HU 226 347 Β1

- switch from staged copy to continuous writing,

- abandon the use of 33 frames in continuous writing,

- repeat a dictation practice so that each dictation word flashes for one second at the copy position,

- inserting (or removing) a graphical grid 40 having a density corresponding to the size and line width of the pattern characters on the working area 30 of the digitizing screen 17,

- slow down, speed up or not change the rhythm of the copy by adjusting the T-time or the falling speed of the 50 light columns to the next exercise,

- adjust the handwriting requirement level for a given student by adjusting the density of the 42 virtual grids.

Preferred ways of displaying and highlighting the 32 patterns and the current pattern element are as follows. To superimpose the characters of pattern 32 and copy 34 when superimposed, it is best to display them with different colored lines. The design elements (word or part of it) can be selected, graphically defined, instead of framed, with a rectangle or word close to the frame, which in the shade or color of its area differs from the shade or base color of the workspace. The characters you want to overwrite, as well as some of their features, can also be marked with lines of a different shade, pattern, or color than other characters or features.

The S-system computer may be, for example, a tablet PC or a palm computer (PDA). From a functional point of view, the key is to write comfortably on the digitizing screen 17. A good image quality tablet PC can be a good solution for home and school use, but is more economical for bulk use. Many well-known companies make tablet PCs, a good example of which is the Fujitsu-Siemens Stylistic model, which is designed so that the 15.1-inch screen can be placed horizontally on a comfortable surface. In school education, it is advisable to build a classroom automated classroom, where student terminals consisting of a digitalization screen, speaker-headphones and microphone are networked with a hero computer. Wacom's Cintiq 15x model, for example, can be used as a digitalization screen for student stations (among many other products). The network can be complemented by a computer-controlled projector, which provides a very good quality of motivational material. It is also possible to set up a network where there is a wireless connection between the hero computer and the student stations and not a wired connection. For example, the Microsoft Mira monitor (featuring a wireless, pen-writable and controllable monitor) featured in Pen Computing's June 2002 issue (45) and has since been named "Windows Powered Smart Display, indicates that the display is running on a network version of the Windows CE operating system.

The advantage of using a handheld system is that it is a cheap solution for use at school and at home. All handheld machines with automatic handwriting recognition are suitable for this purpose. These include models from HP, Philips, Casio and others. The disadvantage of the small screen is that it cannot be used to draw longer written patterns, so you have to choose an accessibility solution. One possibility is to insert one or two newer pairs of pattern lines under the pair of pattern lines to represent the longer pattern. Alternatively, the machine represents as many sample items in the sample row as it can fit in the row, and new sample elements appearing at the end of the row during the exercise supersede the already copied elements at the beginning of the row.

A disadvantage of the size of the handheld is that it is not advisable to resolve the screen demo area workspace by splitting the screen area. The presentation of the motivational material and the exercises should be done in the whole screen area, and when the exercise is done, the whole screen space can be used for 30 workspaces, except for a corner where the typical picture of the motivational material needed for the exercise context can be reduced. A better solution would be a foldable handheld machine with two screens, one for the demonstration area and the other for the workspace.

When using a handheld machine, for writing convenience, the machine can be lowered into a flat recess in the table or placed in a wide frame of appropriate size so that the screen and frame surface are flush.

By way of a specific embodiment of a system according to the invention! 17 is a schematic drawing of a target device. The target device can be manufactured at a low cost using technology developed for mass production of mobile phones and handheld calculators, including the implementation of functions through firmware circuit solutions. A color LCD 102 screen is located on the top surface of the target housing to display an explanation of the motivational world and the exercises. A slender LCD 103 screen can be pulled out of the target housing for displaying exercises and for written copy. The target device further comprises a speaker 104 and a microphone 106, as well as audio inputs / outputs 105 for receiving a headset that can be connected to the target device. The target device can be commissioned by means of 107 on / off switches.

The system for automating handwriting and reading education is well supported by PalmOS and Windows CE operating systems on tablet PCs and PDAs, which control tightly coupled hardware and include user interfaces that help16

EN 226 347 Β1 software developer can easily integrate the use of pen writing and the built-in standard software for writing recognition in the application program.

In addition to the preferred embodiments and embodiments shown, other similar solutions are possible within the spirit and scope of the present invention.

Claims (22)

PATENT CLAIMS 1. A computer training system for teaching handwriting and reading in the form of copy exercises, which includes elements for displaying the exercises for the student and for sensing and evaluating the exercises performed by the student, characterized by: (a) a syllabus module (11) containing written and oral copy exercises and motivational material for those exercises; and (b) a module for demonstrating motivational material and / or information for the exercise (12) based on the exercises in the curriculum module (11), c) an executive module (13) based on the exercises of the curriculum module (11), which provides - a means for initiating the practice described by the presentation module (12), - a written and / or oral presentation of the written and / or oral samples (32) to be copied, - means for scheduling and controlling the copying of samples (32) and for sensing written and oral copies (34) by the student, - means for comparing the copies (34) with the samples (32) and for detecting and recording the difference between the copies (34) and the samples (32), and - a means for transmitting to the module for assessing the difference between the copies (34) and the samples (32), and (d) an assessment module on the basis of the differences, which assesses the practice completed and, depending on the assessment, identifies the next practice and, preferably according to the motivation material of the practice, communicates the assessment to the learner (14). A system according to claim 1, characterized in that it is implemented on a computer device and a learning station (15) connected to the computer device, the learning station (15) comprising: (a) a digitization screen (17) containing exercises, motivational material, demonstration information on the exercises and / or demonstration information on the completion of the exercises, and written copies of the exercises (32) and copies of the learner (34) sensor working area (30), (b) a learner with a sample of oral exercises (32), a speaker or headphones reproducing voice messages for the exercise motivation, presentation and evaluation; and (c) a microphone to detect verbal copies of the learner's speech and answers to questions about the education system (18). System according to claim 1, characterized in that the curriculum module (11) comprises one or more of the following elements in a digitized form: (a) written patterns of characters in different fonts and sizes, patterns highlighting some features, patterns without some features, (b) spoken patterns of characters, pronounced according to the names and sounds of the characters, (c) written patterns of words such as complete words and incomplete words, highlighted parts of words, (d) spoken patterns of words such as words and parts of speech, pronounced, spelled, pronounced, and syllable, (e) written and oral patterns of sentences. The system according to claim 3, characterized in that the training module (11) comprises in practice one or more of the following copy tasks: (a) copying written and spoken patterns of characters (32), such as copying characters, overwriting characters or some of their features, overwriting characters with moving highlight (41), completing incomplete characters, dictating characters, naming characters, pronouncing sounds of characters, (b) copying written and spoken patterns (32) of words, such as writing words, transcribing words, overwriting words with moving point of light (41), overwriting characters or strokes highlighted in words, adding missing characters or traits to missing words, writing words for dictation, writing words for dictation per letter, reading words, reading words syllabically, pronouncing words and parts of words, by sound and by letter, c) Writing and reading sentences, such as writing copies of sentences using punctuation, writing sentences for dictation, using punctuation, verbatim copying sentences, reading sentences. The system according to claim 4, characterized in that the training module (11) comprises, between exercises, dual pattern exercises, single pattern exercises and / or echo-type reading and pronunciation exercises. System according to claim 5, characterized in that, based on the result of the practice (s) carried out, the assessment module (14) is set by the learner to set the time for the practice or parts thereof for the next practice. System according to claim 6, characterized in that the learner's copy (34) is converted to the digitized format used by the syllabus module (11), the copy (34) is placed on the sample (32), and the copy (34) ) includes an executing module (13) for comparing the sample (32) and for detecting and recording the difference between them. System according to claim 7, characterized in that the writing module (32) comprises an executing module (13) which encloses all characters of the written pattern in a rectangular frame (33) and places blank frames (33) of the same size in the replica line. HU 226 347 Β1 9. The system of claim 8, further comprising an executing module (13) for executing the exercise of the exercise in sequential copying according to the set schedule time and defining a frame of the next pattern after copying. The system of claim 9, further comprising an executing module (13) for comparing the pattern (32) and its copy (34) and for making a difference between them by means of a digital virtual grid (13). contain a) means for positioning a virtual grid (42) of density on the sample (32) and a copy (34) of the sample (32) and detecting the passage of the copy (34) over the grid elements of the sample (32), and b) means for sliding the frame (33) of the copy (34) relative to the frame (33) of the sample (32) per grid element and initiating the test according to (a) in most coincidental positions. System according to claim 10, characterized in that the evaluation module (14) comprises a graphical function (a) of the density (line) of the digitizing image (17), depending on the result of the practice performed, adapted to the size and line width of the sample characters. means for inserting or removing a grid (40); (b) means for adjusting the rhythm of copying for the next exercise; and / or (c) adjusting the level of writing requirements for the particular student by adjusting the density of the virtual grid (42) accordingly. A system according to claim 2, characterized in that it is portable and comprises a flat housing, a first display (102) disposed on its upper surface for performing the functions of a demonstration area, and a second housing for extending or folding out of the housing. screen (103). 13. A computer-based instructional method for teaching handwriting and reading in the form of copy exercises, characterized in that the method of claim 1-11. A system as claimed in any one of claims 1 to 6, and in the process, the practice of copying a written pattern (32) is as follows: (a) displaying two lines in the workspace (30) of the digitizing screen (17), displaying the written pattern (32) in the upper sample line, and instructing the student to write copies (34) exactly below the samples (32) in the lower copy line. (b) expecting the remaining time to copy the sample (32) on the screen (17) and preferably terminating the exercise after it has been consumed; (c) detecting the copy (34) entered by the learner and preferably by a line different from the sample; (d) converting the copy (34) into a digitized format identical to the samples (32) stored in the syllabus module (11); and (e) comparing the copy (34) with the sample (32) by a known pattern matching method, and the degree of discrepancy between them is determined and stored. 14. The method of claim 13, wherein, when copying an oral sample, (a) the student dictates the dictated words sequentially, (b) after describing each dictated word, displays in the sample its written complement, followed by the opposite (c) comparing each word described with the written complement of the dictated word by a sample-fitting procedure. Method according to claim 14, characterized in that, during continuous copying (b) of the written pattern (32), all characters of the written pattern (32) are enclosed in a rectangular frame (33) and blank lines (33) of the same size ) for the copies (34), (c) specifying the character frame (33) in the sample words so that it closely matches the character frame (33) in front of the word, (c) where the vertical dimension of the frame (33) and its location is closely matched to the largest vertical alphabet of the alphabet, (d) detecting each student's pencil copy (34) and placing a duplicate line against the sample sign, (e) converting the copy into a digitized format identical to the patterns and (f) comparing the copy (34) with the pattern mark by the pattern matching method and measuring the degree of deviation. and stored. 16. The method of claim 15, wherein, in an oral dictation pattern, (a) describing the dictation pattern elements in the copy queue with the student, (b) displaying as many frames (33) of appropriate size in the queue of copies as the description of the dictation pattern; (c) detecting a written copy of the dictated sample element (34), displaying the complement of the dictated sample element in the sample line and placing the copy (34) on the opposite line, (d) the copy (34) stored in the curriculum module converting the complement of the dictated sample element to the copy (34) per character, evaluating the copy (34) and storing the extent of the difference. The method of claim 16, wherein comparing the pattern (32) and copy (34) of the pattern and making a difference between them is performed by the following steps: a) placing a virtual grid (42) of density on the frame (33) of the sample (32) and its replica (34) and detecting the passage of the replica (34) through the grid elements corresponding to the sample (32), and HU 226 347 Β1 b) sliding the frame (33) of the copy (34) relative to the frame (32) of the sample per lattice element, and starting the test in (a) at most coincidental positions. The method of claim 17, wherein comparing the virtual lattices (42) using binary matrices is performed as follows: (a) forming a virtual grid (42) of the sample (32) into a binary matrix of the sample (32) such that the elements of the matrix are assigned a value of 0 or 1, the elements of value 1 corresponding to the elements of the sample Line 32), (b) forming a virtual grid of the copy (34) (42) forming a binary matrix of the copy (34) such that the elements of the matrix are 0 and 1, the elements of value 1 being the elements of the grid crosses the line of copy (34), (c) combining the matrix of sample (32) and copy (34) in a logical AND relationship to calculate how many elements of the same matrix have the same value of 1, i.e., how many active elements of the copy grid overlaps the active elements of the pattern grid, and then comparing the number of elements in the overlay to the number of elements of the sample matrix 1, to obtain the extent to which the copy of the character (34) overlaps the sample (32), and (d) The resulting ratio is stored. 19. The method of claim 18, wherein in the overwriting practice, the copies are made on the sample itself (32) and the student-written copies (34) are detected on the sample itself. 20. The method of claim 19, wherein the oral copying of the duplicate sample is carried out in the following steps: (a) tagging the current copy sample next to the copy by framing (33) and providing, via headphones or loudspeaker, a complementary verbal sample sample of the current record sample echoed by the learner; converting the phonemes of the oral sample element into a pattern matching with the phonemes of the copy, preferably by disconnecting the sample and the copy while serializing the phonemes; and re-enter the oral sample, providing the learner with a one-time opportunity for improved copying, and storing the error for summary evaluation of the exercise. 21. The method of claim 20, wherein, if the verbal sample element is not copied within the timeframe set by the scoring module (14), the verbal sample element is re-provided through the headphones or speaker, providing the learner with a one-time correction, to echo the word sample and echo the error for a summary evaluation of the exercise. 22. The method of claim 21, wherein the practice module evaluates and outputs from the differences in the written and oral copying and the tendency of the differences in written and oral copying, and stores the result in a database, and, depending on the results of the previous exercises, the following exercise is selected from: - following practice according to the training program, - repeating the exercise carried out under unchanged conditions, - repeating the exercise carried out under modified conditions, - return to preventive practice.