English composition reviewing control system and method of intelligent pen and intelligent pen

The English essay grading and control system using smart pens automates the grading of English essays using OCR and NLP technologies, solving the problem of low efficiency in existing systems and achieving efficient, real-time, multi-dimensional assessment and feedback, thereby improving students' writing skills.

CN118918601APending Publication Date: 2024-11-08GUANGDONG UNIV OF TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410943191.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing English essay grading control systems rely on manual intervention, resulting in low grading efficiency and a lack of automation and real-time feedback.

Method used

A smart pen-based English essay grading and control system was designed, including a scanning and recognition module, a grammar correction module, a comprehensive evaluation and feedback module, and a user interaction module. It uses OCR technology to convert paper essays into electronic text, and natural language processing technology to perform grammar correction and multi-dimensional evaluation, providing instant grading results and revision suggestions.

Benefits of technology

It improves the efficiency and objectivity of English essay grading, provides instant feedback, reduces human intervention, and offers detailed comprehensive evaluation reports to help students improve their writing in a timely manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918601A_ABST
    Figure CN118918601A_ABST
Patent Text Reader

Abstract

The invention discloses an English composition review control system and method of an intelligent pen and the intelligent pen. The system comprises a scanning recognition module, a grammar error correction module, a comprehensive evaluation and feedback module and a user interaction module. The scanning recognition module is connected with the user interaction module, and the grammar error correction module is connected with the comprehensive evaluation and feedback module and the user interaction module; the comprehensive evaluation and feedback module is connected with the user interaction module; the scanning recognition module is used for generating an electronic text according to the obtained English text image to be reviewed; the grammar error correction module is used for performing grammar error correction on the electronic text, generating grammar review data and transmitting the grammar review data to the comprehensive evaluation and feedback module; the comprehensive evaluation and feedback module is used for outputting a composition review result according to the grammar review data and transmitting the composition review result to the user interaction module; the user interaction module is used for displaying a composition review result; the technical problem that due to an existing English composition review control system, the English composition review efficiency is poor is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an English composition review control system and method of a smart pen and the smart pen. Background Art

[0002] English composition is a common language writing output exercise in English learning and an important applied exercise in English learning. Different from the real-time and casual nature of oral language, English composition requires students to use written language to describe the course of an event or narrate and prove a certain point of view.

[0003] With the development of natural language processing technology, it is possible to automatically review English essays. NLP (Natural Language Processing) technology can analyze and understand human language. By building grammar, syntax, semantics and other models, it can realize functions such as automatic error correction and content analysis of texts. Specific technologies include text segmentation, part-of-speech tagging, syntactic analysis, semantic understanding, etc. These technologies can help computers automatically identify errors in essays and provide corresponding modification suggestions.

[0004] Most existing English composition review control systems require users to input the starting level and the composition to be scored, and use different standards for reviewing according to the different levels of English learners, or teachers spend a lot of time reviewing the electronic version of English compositions scanned and recognized from paper compositions. The above processes are too dependent on manual intervention and have a low degree of automation, resulting in poor efficiency in reviewing English compositions. Summary of the invention

[0005] The present invention provides an English composition review control system and method of a smart pen and a smart pen, which are used to solve the technical problem that the review efficiency of English compositions is poor due to the existing English composition review control system.

[0006] The first aspect of the present invention provides an English composition review control system of a smart pen, the system comprising a scanning recognition module, a grammar correction module, a comprehensive evaluation and feedback module and a user interaction module;

[0007] The scanning recognition module is connected to the user interaction module, and the grammar error correction module is connected to the comprehensive evaluation and feedback module and the user interaction module respectively;

[0008] The comprehensive evaluation and feedback module is connected to the user interaction module;

[0009] The scanning and recognition module is used to generate an electronic text based on the acquired English text image to be reviewed and transmit it to the grammar correction module through the user interaction module;

[0010] The grammar correction module is used to perform grammar correction on the electronic text, generate grammar review data and transmit it to the comprehensive evaluation and feedback module;

[0011] The comprehensive evaluation and feedback module is used to output the composition review result according to the grammar review data and transmit it to the user interaction module;

[0012] The user interaction module is used to display the composition review results.

[0013] Furthermore, the grammar review data includes a target error correction text and grammar error correction information; the grammar error correction module is specifically used to:

[0014] Preprocessing the electronic text to output structured text;

[0015] Using a preset text error correction model to generate a target error correction text according to a plurality of text sentences in the structured text data;

[0016] Generate grammatical correction information based on the target correction text and the structured text;

[0017] The grammatical correction information and the target correction text are transmitted to the comprehensive evaluation and feedback module.

[0018] Furthermore, the comprehensive evaluation and feedback module includes a text summarization module, a grammar scoring module, a topic consistency scoring module, a paragraph coherence scoring module, and a comment generation module;

[0019] The grammar error correction module is connected to the grammar scoring module, the text summary module, the comment generation module, and the text coherence scoring module respectively;

[0020] The text summarization module is connected to the topic consistency scoring module;

[0021] The grammar scoring module, the topic consistency scoring module and the text coherence scoring module are all connected to the comment generation module, and the comment generation module is connected to the user interaction module;

[0022] The text summarization module is used to generate a target embedding representation based on the received target error correction text and transmit it to the topic consistency scoring module;

[0023] The topic consistency scoring module is used to generate a topic consistency score based on the received target embedding representation using a preset consistency scoring model and transmit the score to the comment generation module;

[0024] The chapter coherence scoring module is used to divide the received target error correction text, determine a plurality of text continuous segments, and then generate a chapter coherence score according to each of the text continuous segments and transmit it to the comment generation module;

[0025] The grammar scoring module is used to generate a grammar score based on the received grammar error correction information by using a preset grammar scoring function and transmit the grammar score to the comment generation module;

[0026] The comment generation module is used to generate a composition review result according to the received grammar score, the grammar error correction information, the topic consistency score and the paragraph coherence score by using a preset comment generation model.

[0027] Furthermore, the target embedding representation includes a summary embedding representation and a title embedding representation; the text summarization module is specifically used to:

[0028] Using a preset text summary model to generate a summary embedding representation based on the received target error correction text;

[0029] Encoding the text title of the target error correction text by a preset encoder, and outputting a title embedding representation;

[0030] The summary embedding representation and the title embedding representation are transmitted to the topic consistency scoring module.

[0031] Furthermore, the chapter coherence scoring module is specifically used to:

[0032] Dividing the received target error correction text to determine a plurality of continuous text segments;

[0033] Using a preset encoder to encode a plurality of consecutive sentences in each of the continuous text segments to generate a plurality of encoded consecutive sentences corresponding to each of the continuous text segments;

[0034] Using a preset coherence score model, outputting a coherence score corresponding to each of the continuous text segments according to a plurality of encoded continuous sentences corresponding to each of the continuous text segments;

[0035] The coherence scores are aggregated to output a chapter coherence score.

[0036] Furthermore, the system also includes a voice playback module;

[0037] The voice playing module is connected to the user interaction module;

[0038] The voice playing module is used for performing voice playing according to the received composition review result.

[0039] A second aspect of the present invention provides an English composition review method using a smart pen, comprising:

[0040] Acquire an English text image to be reviewed, and generate an electronic text according to the English text image to be reviewed;

[0041] Performing grammatical correction on the electronic text to generate grammatical review data;

[0042] Based on the grammar review data, the composition review results are output and displayed.

[0043] A third aspect of the present invention provides a smart pen, the smart pen comprising a pen head and a pen body connected in sequence;

[0044] The front end of the pen tip is provided with a scanning identifier;

[0045] The surface of the pen body is provided with a touch screen, and the lower side of the touch screen is provided with a power button;

[0046] A voice player is arranged on the back of the pen body;

[0047] A charging port and a volume button are disposed adjacent to the side of the pen body;

[0048] The pen body is built with a control mainboard, and the control mainboard is equipped with any one of the above-mentioned smart pen's English composition review control systems.

[0049] A fourth aspect of the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the English composition review method of the smart pen as described above.

[0050] A fifth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, the method for reviewing English compositions using a smart pen as described above is implemented.

[0051] It can be seen from the above technical solutions that the present invention has the following advantages:

[0052] The first aspect of the above technical solution of the present invention provides an English composition review control system of a smart pen, the system includes a scanning recognition module, a grammar correction module, a comprehensive evaluation and feedback module and a user interaction module; wherein the scanning recognition module is connected to the user interaction module, and the grammar correction module is respectively connected to the comprehensive evaluation and feedback module and the user interaction module; the comprehensive evaluation and feedback module is connected to the user interaction module; the scanning recognition module is used to generate an electronic text according to the acquired English text image to be reviewed and transmit it to the grammar correction module through the user interaction module; the grammar correction module is used to perform grammar correction on the electronic text, generate grammar review data and transmit it to the comprehensive evaluation and feedback module; the comprehensive evaluation and feedback module is used to output the composition review result according to the grammar review data and transmit it to the user interaction module; the user interaction module is used to display the composition review result; based on the above solution, the process of performing grammar correction on the electronic text by the grammar correction module, generating grammar review data, and outputting the composition review result according to the grammar review data by the comprehensive evaluation and feedback module does not require the user to input the starting level and the composition to be scored, reduces the dependence on manual intervention, and further improves the review efficiency of English compositions.

[0053] The second aspect of the above technical solution of the present invention provides an English composition review method of a smart pen, which first obtains an image of an English text to be reviewed, and generates an electronic text based on the image of the English text to be reviewed; then, grammatical errors are corrected on the electronic text to generate grammar review data; finally, based on the grammar review data, the composition review result is output and displayed; the above process does not require the user to input a starting level and a composition to be scored, thereby reducing dependence on manual intervention and further improving the review efficiency of English compositions.

[0054] The third aspect of the above technical scheme of the present invention provides a smart pen, which includes a pen head and a pen body connected in sequence; a scanning identifier is arranged at the front end of the pen head; a touch display screen is arranged on the surface of the pen body, a voice player is arranged on the back of the pen body, and a power button is arranged on the lower side of the touch display screen; a charging port and a volume button are arranged adjacent to the side of the pen body; a control main board is built into the pen body, and the control main board is equipped with an English composition review control system of the smart pen; based on the above scheme, by pressing the power button, the English text image to be reviewed is scanned and recognized by the scanning identifier of the smart pen, and the review of the recognized electronic English text is completed based on the English composition review control system of the smart pen equipped with the control main board. The review process does not require the user to input the starting level and the composition to be scored, reduces the dependence on manual intervention, can improve the review efficiency of English compositions, and is easy to carry. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0056] Figure 1 A schematic diagram of the structure of an English composition review control system of a smart pen provided in the first embodiment of the present invention;

[0057] Figure 2 A schematic diagram of a prompt template designed for the LLama model provided in the first embodiment of the present invention;

[0058] Figure 3 A flowchart of the English composition review control system of the smart pen provided in the first embodiment of the present invention;

[0059] Figure 4 A flowchart of a method for reviewing English essays using a smart pen provided in Embodiment 2 of the present invention;

[0060] Figure 5 A schematic diagram of the structure of a smart pen provided in Embodiment 3 of the present invention;

[0061] The meanings of the symbols in the accompanying drawings are as follows:

[0062] 1. Scanning identifier; 2. Power button; 3. Touch screen; 4. Voice player; 5. Charging port; 6. Volume button. DETAILED DESCRIPTION

[0063] The embodiment of the present invention provides an English composition review control system and method of a smart pen and a smart pen, which are used to solve the technical problem that the review efficiency of English compositions is poor due to the existing English composition review control system.

[0064] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] See also Figure 1 , Figure 1 This is a schematic diagram of the structure of an English composition review control system of a smart pen provided in Embodiment 1 of the present invention.

[0066] The present invention provides an English composition review control system of a smart pen, which includes a scanning recognition module, a grammar error correction module, a comprehensive evaluation and feedback module and a user interaction module;

[0067] See also Figure 1 In the embodiment, the scanning recognition module is connected to the user interaction module, the grammar error correction module is connected to the comprehensive evaluation and feedback module and the user interaction module respectively; the comprehensive evaluation and feedback module is connected to the user interaction module;

[0068] The scanning and recognition module is used to generate an electronic text based on the acquired English text image to be reviewed and transmit it to the grammar correction module through the user interaction module; the grammar correction module is used to perform grammar correction on the electronic text, generate grammar review data and transmit it to the comprehensive evaluation and feedback module; the comprehensive evaluation and feedback module is used to output the composition review results based on the grammar review data and transmit it to the user interaction module; the user interaction module is used to display the composition review results.

[0069] The comprehensive evaluation and feedback module includes text summarization module, grammar scoring module, topic consistency scoring module, paragraph coherence scoring module and comment generation module.

[0070] It should be noted that the scanning recognition module uses OCR technology (Optical Character Recognition) to convert the scanned English text image to be reviewed into an electronic text; the grammar correction module pre-processes the electronic text and detects and corrects grammatical errors in the composition (pre-processed electronic text); the paragraph coherence scoring module can evaluate the structure and logical coherence of the corrected article (target corrected text); the theme consistency scoring module is used to check whether the content of the corrected article is consistent with the theme; the comment generation module can use the output results of the grammar scoring module, the theme consistency scoring module, and the paragraph coherence scoring module to generate a comprehensive evaluation report, that is, the composition review result, which includes the composition modification suggestions. Finally, the user interaction module provides a user interface for viewing the detailed review results, and the voice playback module broadcasts the review results and modification suggestions in real time.

[0071] As a further improvement, the grammar review data includes the target error correction text and grammar error correction information; the grammar error correction module is specifically used to:

[0072] Preprocess the electronic text and output structured text; use a preset text correction model to generate a target correction text based on multiple text sentences in the structured text data; generate grammatical correction information based on the target correction text and the structured text; transmit the grammatical correction information and the target correction text to the comprehensive evaluation and feedback module.

[0073] It should be noted that the grammar correction module first preprocesses the received electronic text and outputs a structured text. The preprocessing process includes denoising, tilt correction, word segmentation, part-of-speech tagging, syntactic analysis, etc., to generate structured text data (structured text); then, the natural language processing (NLP) technology is used to perform grammar checking and error correction on the structured text, mark grammatical errors, and provide modification suggestions; the module has a strong grammatical analysis capability and can accurately identify errors in complex syntactic structures; specifically, each sentence in the structured text is input into the preset text correction model in turn to obtain all the corrected words corresponding to each sentence, and these corrected words correspond to the corrected sentences, thereby obtaining the corrected chapter, that is, the target correction text; wherein, the preset text correction model, the preset text correction model is composed of a conditional probability model and a feature function, the feature function and the corresponding weight can capture more context information and sentence features, and improve the accuracy of correction, which is specifically:

[0074]

[0075] Among them, y n is the word after error correction; g(y|X,y <n ) is the probability of a word in a text sentence; M is the number of feature functions; α m is the weight corresponding to the characteristic function; Φ m (y,X,y <n ) is the eigenvalue corresponding to the characteristic function; Z(X,y <n ) is the normalization constant corresponding to the preset text error correction model, which can ensure the correctness of the probability model and prevent the calculation result from being too large or too small; X is a text sentence in the structured text, and the text sentence consists of multiple words; Φ m (·) is the characteristic function.

[0076] Furthermore, after the entire article (text) is corrected, the text before and after the correction is obtained, that is, H = (X1, X2, ... X N ),H'=(X1',X2',...,X' N ), where H is the text before correction, i.e., the structured text, H' is the text after correction, i.e., the target correction text, X1 is the first text sentence in the structured text, X1' is the first text sentence in the target correction text, and N is the total number of text sentences; H and H' are collectively referred to as the grammatical correction information E g , that is, grammatical error correction information E g Composed of H and H'.

[0077] As a further improvement, the comprehensive evaluation and feedback module includes a text summarization module, a grammar scoring module, a topic consistency scoring module, a paragraph coherence scoring module, and a comment generation module;

[0078] The grammar error correction module is connected to the grammar scoring module, the text summary module, the comment generation module, and the text coherence scoring module respectively; the text summary module is connected to the topic consistency scoring module; the grammar scoring module, the topic consistency scoring module, and the text coherence scoring module are all connected to the comment generation module, and the comment generation module is connected to the user interaction module;

[0079] It should be noted that the paragraph coherence scoring module can evaluate the logical coherence between the paragraph structure and sentences of the composition, and ensure the overall fluency of the article by analyzing the transition sentences and connecting words of the article; the theme consistency scoring module can check whether the content of the composition is consistent with the given theme, and evaluate the theme consistency of the article through keyword extraction and theme analysis technology; the comment generation module generates a detailed comprehensive evaluation report (composition review results) based on the grammatical correction results, paragraph coherence score and theme consistency score; the report content includes the scores of each dimension, specific error markings and modification suggestions.

[0080] The text summarization module is used to generate a target embedding representation based on the received target error correction text and transmit it to the topic consistency scoring module;

[0081] The topic consistency scoring module is used to generate a topic consistency score based on the received target embedding representation using a preset consistency scoring model and transmit it to the comment generation module;

[0082] It should be noted that the preset consistency scoring model is as follows:

[0083]

[0084] Among them, C t Score the theme consistency; F t (·) is the scoring function; is the interpolation between the summary embedding representation and the perturbation summary embedding representation, and the perturbation summary embedding representation is obtained by adding random noise to the summary embedding representation; is the interpolation between the title embedding representation and the perturbation title embedding representation, and the perturbation summary embedding representation is obtained by adding random noise to the summary embedding representation; is the normalization constant corresponding to the preset consistency scoring model.

[0085] The chapter coherence scoring module is used to divide the received target error correction text, determine multiple text continuous segments, and then generate a chapter coherence score based on each text continuous segment and transmit it to the comment generation module;

[0086] The grammar scoring module is used to generate a grammar score based on the received grammar error correction information by using a preset grammar scoring function and transmit the grammar score to the comment generation module;

[0087] It should be noted that the grammar scoring module calculates the grammar score according to the grammar correction information output by the grammar correction module; wherein the preset grammar scoring function is specifically:

[0088]

[0089] Among them, G s Score grammar; F g (·) is the grammatical scoring function, which is designed according to the set scoring rules; H is the structured text; H' is the target correction text; γ is the scaling factor; Ψ(·) is the semantic similarity function; Z g (H,H') is the normalization constant corresponding to the preset grammar scoring function.

[0090] The comment generation module is used to generate composition review results based on the received grammar scores, grammar correction information, topic consistency scores and paragraph coherence scores using a preset comment generation model.

[0091] It should be noted that the comment generation module generates the composition review results based on the grammar score, grammar correction information, topic consistency score and paragraph coherence score; among them, the preset comment generation model is specifically:

[0092] W=G n (G s ,E g ,C p ,C t );

[0093] Among them, W is the result of essay review; G n (·) is a preset comment generation model based on LLama (Large Language Model Family of AI); G s Score grammar; E g C is grammatical error correction information; p Score the coherence of the passage; C t Score the theme consistency.

[0094] Furthermore, the grammar score, grammatical correction information, topic consistency score and paragraph coherence score are input into the LLama model and then the final comment generated will combine these parts to form a comprehensive feedback, such as Figure 2 Shown is a prompt template designed for the LLama model.

[0095] As a further improvement, the target embedding representation includes a summary embedding representation and a title embedding representation; the text summarization module is specifically used to:

[0096] A preset text summary model is used to generate a summary embedding representation based on the received target correction text; the text title of the target correction text is encoded through a preset encoder to output the title embedding representation; the summary embedding representation and the title embedding representation are transmitted to the topic consistency scoring module.

[0097] It should be noted that before analyzing whether the theme of the composition is consistent and whether there is a phenomenon of going off topic and giving a theme consistency score, the corrected text, that is, the target corrected text, is first input into the text summary module, and the preset text summary model is used to generate a summary embedding representation based on the received target corrected text. The process can be expressed as:

[0098]

[0099] Among them, R H is the summary embedding representation; G sum (·) is the text summary function; Z H' is the normalization constant corresponding to the preset text summary model; H' λ It is the interpolation between the target error correction text and the perturbed target error correction text. The perturbed target error correction text is obtained by adding random noise to the target error correction text.

[0100] Furthermore, the article title (text title) of the target error correction text T = (t1, t2, ..., t l ) is input to the pre-set encoder Encoder(·) to obtain the title embedding representation R T , where t1 is the first word in the text title, t2 is the second word in the text title, and t l is the lth word in the text title, l is the length of the article title sentence, l∈R, the process can be expressed as:

[0101]

[0102] Among them, R T is the title embedding representation; Encoder(·) is the preset encoder; T j is the jth sentence in the text title. Usually the title contains only a single sentence. Here we assume that the text title may contain multiple sentences; ω j Embed the weights of different parts of the title; is the normalization constant corresponding to the preset encoder.

[0103] As a further improvement, the chapter coherence scoring module is specifically used to:

[0104] The received target error correction text is divided to determine a plurality of text continuous segments; a preset encoder is used to encode a plurality of continuous sentences in each text continuous segment to generate a plurality of encoded continuous sentences corresponding to each text continuous segment; a preset coherence score model is used to output a coherence score corresponding to each text continuous segment according to the plurality of encoded continuous sentences corresponding to each text continuous segment; each coherence score is aggregated to output a chapter coherence score.

[0105] It should be noted that the overall structure of the composition and the coherence between paragraphs are evaluated and the chapter coherence score is calculated; specifically, the received target correction text is divided to determine multiple text continuous segments, that is, continuous sentences with a window size of 2k+1 (k>1) are extracted from the corrected text H' as a segment (text continuous segment), denoted as U i =<S i-k ,...,S i ,...,S i+k >, where k is half of the window size, U i is the i-th continuous text segment, S i-k is the ikth continuous sentence; then the preset encoder is used to encode each continuous sentence in each continuous text segment. The process can be expressed as:

[0106] z i =Encoder(S i );

[0107] Among them, z i is the i-th encoded continuous sentence; Encoder(·) is the preset encoder; S i is the i-th continuous sentence.

[0108] Furthermore, the coherence score of the continuous text segment is represented by the coherence(·) function, and the semantic representation of the continuous text segment is obtained by combining the semantic combination function comp (such as concatenation). The representation is mapped to a coherence label or score through a classification network f (such as a forward network), that is, a preset coherence score model is used to output the coherence score corresponding to each continuous text segment according to multiple encoded continuous sentences corresponding to each continuous text segment. The process can be expressed as:

[0109]

[0110] Among them, coherence (U i ) is the coherence score of the i-th text segment; σ(·) is the sigmoid activation function; β i is the weight; comp(·) is the semantic combination function; δ is the scaling factor; Z c (U i) is the normalization constant corresponding to the preset coherence score model, Z c (U) = ∑ U exp(log(1+δ|||z i+k -z i-k || 2 ));z i Encode the i-th consecutive sentence.

[0111] Furthermore, the coherence scores are aggregated to output the chapter coherence score; specifically, the chapter coherence score can be achieved by aggregating the coherence probabilities or scores of multiple paragraphs. After the above steps, the received target error correction text is divided, and after determining multiple continuous text segments, the coherence scores are aggregated to obtain the chapter coherence score. The process can be expressed as:

[0112]

[0113] Among them, P coherence (U i ) is the coherence probability value of the i-th text segment; sigmoid(·) is the sigmoid activation function; C p is the coherence score of the text; ε is the adjacent text segment, that is, U i and U i+1 Parameter for the smoothness of the transition between i is the i-th continuous text segment; Z p (U i ) is the aggregation normalization constant,

[0114] As a further improvement, the system also includes a voice playback module; the voice playback module is connected to the user interaction module; the voice playback module is used to play voice according to the received composition review results.

[0115] It should be noted that the user interaction module allows users to view detailed review results. The interactive interface design should be simple and intuitive, so that users can quickly find the information they need. The voice playback module broadcasts the review results of the essay in real time through voice to enhance the user experience. The voice playback content should be optimized to ensure the clarity and accuracy of the information conveyed.

[0116] For example, see Figure 3First, the scanned composition image is converted into an electronic text by using the OCR technology through the scanning recognition module. The recognized electronic text is preprocessed by the grammar correction module, including word segmentation, part-of-speech tagging, syntactic analysis, etc., to generate structured text data. Then, the natural language processing (NLP) technology is used to check the grammar and correct errors of the structured text data. The paragraph structure and logical coherence between sentences of the composition are evaluated by the chapter coherence scoring module. Then, the theme consistency scoring module is used to check whether the content of the composition is consistent with the given theme. Based on keyword extraction and theme analysis technology, the theme consistency of the article is evaluated. The comment generation module generates the composition review result according to the grammar score, grammar correction information, theme consistency score and chapter coherence score. Finally, the voice playback module is used to play the received composition review result, and the user interaction module is used to display the composition review result.

[0117] As a comparison of technical effects, it can be combined with existing technologies for reference. Traditional English composition review requires a lot of manual input, and teachers need to spend a lot of time to correct sentences and paragraphs, which is not only inefficient, but also easily affected by subjective factors. The present invention introduces an intelligent review system, uses scanning components and review models, realizes automated review of compositions, significantly improves review efficiency, and reduces the workload of teachers. Since manual review is inevitably affected by reviewers' personal knowledge, experience, emotions and other factors, there are often large differences in review results. The present invention ensures the consistency and objectivity of composition review by constructing a grammar correction model, a chapter coherence scoring model, a theme consistency scoring model, etc., reduces human errors, and improves the reliability of review results. Traditional composition correction feedback is usually one-time, and students need to wait for a long time to receive the review results, and it is difficult to make timely improvements based on feedback. The smart pen of the present invention can provide real-time review feedback of compositions through voice playback components and user interaction components, helping students to understand their own mistakes and deficiencies in a timely manner, and make corresponding improvements and learning to improve writing skills. Traditional essay review focuses on marking grammar and spelling errors, while ignoring the comprehensive evaluation of paragraph structure, coherence and theme consistency. The present invention can not only perform detailed grammatical correction, but also comprehensively evaluate the paragraph coherence, theme consistency and other aspects of the essay, provide more comprehensive and in-depth review opinions, and help students improve their writing level from multiple aspects. In summary, the technical problems to be solved by the present invention include improving the efficiency and objectivity of English essay review, achieving real-time feedback, and enhancing the comprehensiveness and comprehensiveness of the review, thereby improving students' writing ability and learning experience.

[0118] Furthermore, the limitations of manual marking: Traditional English essay review relies on manual marking by teachers. Due to the different knowledge levels, experience and review standards of each teacher, the review results are subjective and inconsistent. In addition, manual review requires a lot of time and energy. When the amount of essays is large, it is difficult for teachers to complete the review work quickly and accurately. Students cannot get feedback in time, which affects the learning effect. Especially for large-scale exams or homework, the review cycle is long, and students cannot understand the problems in their writing at the first time, so they cannot improve and improve in time. Natural language processing (NLP) technology: The development of natural language processing technology has made it possible to automatically review essays. NLP technology can analyze and understand human language. By building models such as grammar, syntax, and semantics, it can realize functions such as automatic error correction and content analysis of texts. Specific technologies include text segmentation, part-of-speech tagging, syntactic analysis, semantic understanding, etc. These technologies can help computers automatically identify errors in essays and provide corresponding modification suggestions. Machine learning and artificial intelligence (AI) technology: Machine learning, especially deep learning technology, has made major breakthroughs in image recognition, speech recognition, natural language processing and other fields. By training on a large-scale corpus, machine learning models can automatically learn the rules and standards for essay review, thereby achieving efficient and accurate automatic review. For example, neural network models can be used to build the core algorithms of essay review systems, including grammatical error correction, paragraph coherence analysis, and topic consistency assessment.

[0119] Furthermore, smart pens and scanning technology: With the development of intelligent hardware, various types of smart pens have emerged. These intelligent devices can realize functions such as handwriting input and real-time digitization. The smart pen of the present invention integrates a scanning component, which can quickly read paper compositions and convert them into electronic texts, providing data support for subsequent automatic review. The development of scanning technology and optical character recognition (OCR) technology has made the digitization of paper texts more accurate and efficient, providing basic data for intelligent review systems. An excellent composition must not only be grammatically correct, but also perform well in terms of chapter structure, content coherence, and theme consistency. Traditional reviews often focus on grammar and spelling, ignoring the overall quality of the composition. The present invention constructs multiple evaluation models, including a grammar correction model, a chapter coherence scoring model, and a theme consistency scoring model, which can comprehensively evaluate the composition from multiple dimensions, generate appropriate comments, give a comprehensive evaluation report, and provide more comprehensive and detailed feedback. Immediacy and interactivity: Traditional review feedback is delayed, and it is difficult for students to make improvements based on feedback in a timely manner. The smart pen of the present invention integrates a voice playback component and a user interaction component, and can provide students with review results and modification suggestions in real time, enhancing the immediacy and interactivity of learning. Students can use the smart pen to understand the problems in their own compositions in real time, and modify and practice according to the suggestions provided by the system, thereby achieving more efficient learning and improvement.

[0120] In summary, this invention is based on cutting-edge technologies such as natural language processing, machine learning, and intelligent hardware, and aims to solve the problems of low efficiency, strong subjectivity, and delayed feedback in traditional English composition review. Through an automated and intelligent review system, a multi-dimensional comprehensive evaluation and real-time feedback of the composition can be achieved, helping students improve their writing skills and reducing the workload of teachers.

[0121] Furthermore, although existing review technologies have certain advantages in flexibility and practicality, they still have some disadvantages and limitations: existing technologies require users to input the starting grade and the essay to be scored, which increases the complexity of use and the possibility of manual intervention; existing methods mainly focus on the calculation and evaluation of scores, lack instant feedback functions, and cannot provide instant improvement suggestions during the review process; existing technologies may not be detailed enough in grammar and coherence evaluation, and cannot provide specific error positioning and detailed error correction suggestions; although existing technologies cover the evaluation of vocabulary, grammar, topicality and fluency, they lack a comprehensive, multi-dimensional evaluation system to comprehensively analyze the quality of essays; the user interaction interface and operation process of existing methods may not be intuitive and convenient enough, affecting user experience and usage efficiency.

[0122] To address the above issues, the smart pen hardware can be used to automatically scan and recognize essay texts, reduce user input dependence, simplify the operation process, and improve review efficiency. The voice playback component and display screen provide instant review results and improvement suggestions to help users understand and correct errors in a timely manner during the review process. Advanced natural language processing technology is used to conduct detailed grammar checks and coherence assessments on essays, and specific error annotations and detailed correction suggestions are provided. A comprehensive evaluation model is established, covering multiple dimensions such as grammar, paragraph coherence, and theme consistency, and a detailed comprehensive evaluation report is generated to help users fully understand the quality of the essay. An intuitive user interface is provided, and the display screen and buttons of the smart pen make it convenient for users to view review results and detailed feedback, thereby improving user experience and usage efficiency.

[0123] Therefore, the present invention realizes the automatic and intelligent review of English composition. Its core technologies include efficient scanning recognition, advanced natural language processing algorithms, multi-dimensional composition evaluation models and instant feedback mechanisms. The system not only improves the review efficiency and accuracy, but also provides comprehensive feedback information, helping students to effectively improve their writing skills and reducing the workload of teachers.

[0124] In an embodiment of the present invention, the present invention provides an English composition review control system of a smart pen, the system includes a scanning recognition module, a grammar correction module, a comprehensive evaluation and feedback module and a user interaction module; wherein the scanning recognition module is connected to the user interaction module, and the grammar correction module is respectively connected to the comprehensive evaluation and feedback module and the user interaction module; the comprehensive evaluation and feedback module is connected to the user interaction module; the scanning recognition module is used to generate an electronic text according to the acquired English text image to be reviewed and transmit it to the grammar correction module through the user interaction module; the grammar correction module is used to perform grammar correction on the electronic text, generate grammar review data and transmit it to the comprehensive evaluation and feedback module; the comprehensive evaluation and feedback module is used to output the composition review result according to the grammar review data and transmit it to the user interaction module; the user interaction module is used to display the composition review result; based on the above scheme, the process of performing grammar correction on the electronic text by the grammar correction module, generating grammar review data, and outputting the composition review result according to the grammar review data by the comprehensive evaluation and feedback module does not require the user to input the starting level and the composition to be scored, reduces the dependence on manual intervention, and further improves the review efficiency of English compositions.

[0125] See also Figure 4 , Figure 4 This is a flow chart of the steps of an English composition review method using a smart pen provided in Embodiment 2 of the present invention.

[0126] The present invention provides a smart pen English composition review method, comprising:

[0127] 401. Obtain an image of an English text to be reviewed, and generate an electronic text according to the image of the English text to be reviewed.

[0128] In this embodiment, an image of an English text to be reviewed is obtained, and an electronic text is generated according to the image of the English text to be reviewed.

[0129] 402. Perform grammatical correction on the electronic text and generate grammatical review data.

[0130] In this embodiment, grammar correction is performed on the electronic text to generate grammar review data.

[0131] 403. Output and display the composition review results based on the grammar review data.

[0132] In this embodiment, the composition review results are output and displayed based on the grammar review data.

[0133] In an embodiment of the present invention, the present invention provides an English composition review method of a smart pen, which first obtains an English text image to be reviewed, and generates an electronic text based on the English text image to be reviewed; then, grammatical errors are corrected on the electronic text to generate grammar review data; finally, based on the grammar review data, the composition review result is output and displayed; the above process does not require the user to input the starting level and the composition to be scored, reduces the dependence on manual intervention, and further improves the review efficiency of English compositions.

[0134] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of a smart pen provided in Embodiment 3 of the present invention.

[0135] The present invention provides a smart pen, which includes a pen head and a pen body connected in sequence;

[0136] A scanning identifier 1 is provided at the front end of the pen tip;

[0137] A touch screen 3 is provided on the surface of the pen body, and a power button 2 is provided on the lower side of the touch screen 3;

[0138] A voice player 4 is provided on the back of the pen body;

[0139] A charging port 5 and a volume button 6 are provided adjacent to the side of the pen body;

[0140] The pen body is built with a control mainboard, and the control mainboard is equipped with the English composition review control system of the smart pen of the first embodiment.

[0141] It should be noted that the scanning identifier 1 is an integrated high-resolution camera or scanner, which can use OCR technology to convert paper text into electronic text with high precision and high speed to ensure the accuracy of recognition and user experience; the voice player 4 is a high-quality speaker, which can clearly broadcast the review results and feedback information to ensure that the information is conveyed accurately and easily understood; the touch display screen 3 and the power button 2 are convenient for users to operate and view the review results. The touch display screen 3 has high resolution and good visibility, and the power button 2 improves the convenience of user operation.

[0142] In an embodiment of the present invention, the present invention provides a smart pen, which includes a pen head and a pen body connected in sequence; a scanning identifier is arranged at the front end of the pen head; a touch display screen is arranged on the surface of the pen body, a voice player is arranged on the back of the pen body, and a power button is arranged on the lower side of the touch display screen; a charging port and a volume button are arranged adjacent to the side of the pen body; a control main board is built into the pen body, and the control main board is equipped with an English composition review control system of the smart pen; based on the above scheme, by pressing the power button, the English text image to be reviewed is scanned and recognized by the scanning identifier of the smart pen, and the review of the recognized electronic English text is completed based on the English composition review control system of the smart pen equipped with the control main board. The review process does not require the user to input the starting level and the composition to be scored, reduces the dependence on manual intervention, can improve the review efficiency of English compositions, and is easy to carry.

[0143] An embodiment of the present invention further provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory; when the computer program is executed by the processor, the processor executes the steps of the English composition review method of the smart pen in any of the above embodiments.

[0144] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of the method for reviewing English composition using a smart pen as in any of the above embodiments are implemented.

[0145] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0146] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0147] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An English composition review control system of a smart pen, characterized in that: The system includes a scanning recognition module, a grammar error correction module, a comprehensive evaluation and feedback module and a user interaction module; The scanning recognition module is connected to the user interaction module, and the grammar error correction module is connected to the comprehensive evaluation and feedback module and the user interaction module respectively; The comprehensive evaluation and feedback module is connected to the user interaction module; The scanning and recognition module is used to generate an electronic text based on the acquired English text image to be reviewed and transmit it to the grammar correction module through the user interaction module; The grammar correction module is used to perform grammar correction on the electronic text, generate grammar review data and transmit it to the comprehensive evaluation and feedback module; The comprehensive evaluation and feedback module is used to output the composition review result according to the grammar review data and transmit it to the user interaction module; The user interaction module is used to display the composition review results.

2. The English composition review control system of the smart pen according to claim 1, characterized in that: The grammar review data includes a target error correction text and grammar error correction information; the grammar error correction module is specifically used to: Preprocessing the electronic text to output structured text; Using a preset text error correction model to generate a target error correction text according to a plurality of text sentences in the structured text data; Generate grammatical correction information based on the target correction text and the structured text; The grammatical error correction information and the target error correction text are transmitted to the comprehensive evaluation and feedback module.

3. The English composition review control system of the smart pen according to claim 2, characterized in that: The comprehensive evaluation and feedback module includes a text summary module, a grammar scoring module, a topic consistency scoring module, a paragraph coherence scoring module and a comment generation module; The grammar error correction module is connected to the grammar scoring module, the text summary module, the comment generation module, and the text coherence scoring module respectively; The text summarization module is connected to the topic consistency scoring module; The grammar scoring module, the topic consistency scoring module and the text coherence scoring module are all connected to the comment generation module, and the comment generation module is connected to the user interaction module; The text summarization module is used to generate a target embedding representation based on the received target error correction text and transmit it to the topic consistency scoring module; The topic consistency scoring module is used to generate a topic consistency score based on the received target embedding representation using a preset consistency scoring model and transmit the score to the comment generation module; The chapter coherence scoring module is used to divide the received target error correction text, determine a plurality of text continuous segments, and then generate a chapter coherence score according to each of the text continuous segments and transmit it to the comment generation module; The grammar scoring module is used to generate a grammar score based on the received grammar error correction information by using a preset grammar scoring function and transmit the grammar score to the comment generation module; The comment generation module is used to generate a composition review result according to the received grammar score, the grammar error correction information, the topic consistency score and the paragraph coherence score by using a preset comment generation model.

4. The English composition review control system of the smart pen according to claim 3, characterized in that: The target embedding representation includes a summary embedding representation and a title embedding representation; the text summarization module is specifically used to: Using a preset text summary model to generate a summary embedding representation based on the received target error correction text; Encoding the text title of the target error correction text by a preset encoder, and outputting a title embedding representation; The summary embedding representation and the title embedding representation are transmitted to the topic consistency scoring module.

5. The English composition review control system of the smart pen according to claim 3, characterized in that: The chapter coherence scoring module is specifically used for: Dividing the received target error correction text to determine a plurality of continuous text segments; Using a preset encoder to encode a plurality of consecutive sentences in each of the continuous text segments to generate a plurality of encoded consecutive sentences corresponding to each of the continuous text segments; Using a preset coherence score model, outputting a coherence score corresponding to each of the continuous text segments according to a plurality of encoded continuous sentences corresponding to each of the continuous text segments; The coherence scores are aggregated to output a chapter coherence score.

6. The English composition review control system of the smart pen according to claim 1, characterized in that: The system also includes a voice playback module; The voice playing module is connected to the user interaction module; The voice playing module is used for performing voice playing according to the received composition review result.

7. A method for reviewing English composition using a smart pen, characterized in that: include: Acquire an English text image to be reviewed, and generate an electronic text according to the English text image to be reviewed; Performing grammatical correction on the electronic text to generate grammatical review data; Based on the grammar review data, the composition review results are output and displayed.

8. A smart pen, characterized in that: The smart pen comprises a pen head and a pen body connected in sequence; The front end of the pen tip is provided with a scanning identifier; The surface of the pen body is provided with a touch screen, and the lower side of the touch screen is provided with a power button; A voice player is arranged on the back of the pen body; A charging port and a volume button are disposed adjacent to the side of the pen body; The pen body has a built-in control mainboard, and the control mainboard is equipped with the English composition review control system of the smart pen according to any one of claims 1-6.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the English composition review method of the smart pen as claimed in any one of claim 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the English composition review method of the smart pen according to any one of claim 7 is implemented.

Citation Information

Cited By

  • English composition quality comprehensive evaluation system based on large language model

    CN119962532A