English reading understanding ability analysis system based on deep learning

Abstract

The invention relates to the technical field of education evaluation, in particular to an English reading understanding ability analysis system based on deep learning, which comprises the following steps: an answer text analysis module, a reading mode analysis module, an attention ability evaluation module, an efficiency score calculation module, an understanding obstacle recognition module and a reading ability evaluation module. According to the method, the keyword in the answer text of the student is extracted and compared with the preset answer, so that the understanding degree of the student on the English reading understanding text is effectively identified, and the reading behavior of the student is monitored in real time, so that the attention of the student can be identified in a concentrated and dispersed time period; the attention level and the reading efficiency of the student are effectively evaluated, and in combination with the detection of reading disorder and the analysis of the ability level of the student, an educator is helped to quickly and accurately grasp the learning condition of the student and adjust the teaching strategy and content, so that the individualized requirements of the student are met, and the learning efficiency and the education quality are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of educational assessment technology, and in particular to an English reading comprehension ability analysis system based on deep learning. Background Art

[0002] The field of educational assessment technology involves the use of various methods and tools to measure students' or learners' knowledge, skills and abilities. It aims to use a variety of assessment methods to measure students' mastery of various subjects, monitor educational effectiveness and provide support for educational decision-making. It includes standardized tests, formative evaluations and diagnostic evaluations, combined with computer-assisted testing and online assessment platforms to provide instant feedback and personalized learning suggestions, improve the accuracy and efficiency of assessments, and is applied to the assessment of knowledge in various subjects.

[0003] Among them, the English reading comprehension ability analysis system is used to evaluate and analyze students' English reading comprehension ability, covering the content analysis of reading materials, tracking and analysis of user reading behavior, and evaluation of comprehension. It uses question answers to evaluate students' grasp and analysis of text content. By analyzing students' answers and interactions, including reading time, pause points and accuracy of answering questions, text analysis technology and user behavior analysis are used to assess students' reading comprehension level, supporting teachers and educators to understand students' learning needs and progress.

[0004] Traditional English reading comprehension ability analysis technology relies on standardized tests and formative assessments, lacks an in-depth understanding of students' individual differences and the ability to provide immediate feedback. The standardized test cycle is long and the feedback is slow, making it difficult to capture immediate changes in students' learning process. It is not flexible enough to deal with situations that require timely adjustment of learning strategies, and ignores students' micro-behaviors during the reading process, including attention distraction and reading pauses. It is difficult to understand students' reading strategies and reading difficulties. When dealing with students' reading difficulties, there is a lack of targeted diagnosis and support, making it difficult to achieve personalized teaching adjustments, resulting in students' specific needs not being effectively met, affecting the optimization of learning outcomes, including difficulty in identifying students' vocabulary and grammatical obstacles, making it difficult for teachers to provide necessary targeted tutoring, affecting the optimization of educational effects and the satisfaction of students' learning needs. Summary of the Invention

[0005] In order to solve the technical problem of the existing technology that lacks understanding of individual differences among students, the embodiment of the present invention provides an English reading comprehension ability analysis system based on deep learning. The technical solution is as follows:

[0006] On the one hand, a deep learning-based English reading comprehension ability analysis system is provided, which includes:

[0007] The answer text analysis module extracts keywords from the student answer text and combines it with pre-set answer information. It then evaluates the accuracy, completeness, and relevance of the student answer, analyzes the student's understanding of the reading text, and generates answer verification results.

[0008] The reading pattern analysis module identifies the target student's reading behavior pattern based on the answer verification result by monitoring the student's reading and answering behavior in real time, and obtains a reading behavior monitoring record;

[0009] The attention ability assessment module evaluates the persistence and concentration of the student's attention by identifying the time periods of distraction during the reading process based on the reading behavior monitoring records, and generates an attention analysis result;

[0010] The efficiency score calculation module analyzes the reading fluency and text processing ability of multiple students based on the attention analysis results and the students' reading behaviors, calculates the reading efficiency scores of the multiple students, and forms a reading efficiency evaluation result;

[0011] The comprehension barrier identification module detects the student's inefficient reading behavior based on the reading efficiency evaluation results, identifies the target student's comprehension barrier points, classifies the barrier points, and generates a reading barrier diagnosis record;

[0012] The reading ability assessment module evaluates the student's reading comprehension ability in real time based on the reading disability diagnosis record and the student's performance in reading tasks of various difficulty levels, and generates a reading comprehension ability score.

[0013] As a further solution of the present invention, the answer verification results are specifically keyword coverage analysis data, semantic consistency analysis results, and information integrity assessment information; the reading behavior monitoring records include reading start and end time points, stop point behavior marks, and annotation behavior monitoring records; the attention analysis results specifically refer to attention concentration time period marking information, distracted time period identification results, and persistence level; the reading efficiency assessment results specifically include reading time, annotation behavior frequency, and stop point duration; the dyslexia diagnosis records include grammatical difficulties, vocabulary difficulties, and phrase difficulties; and the reading comprehension ability score specifically refers to reading task difficulty, text comprehension level, and reading behavior performance.

[0014] As a further solution of the present invention, the answer text analysis module includes:

[0015] The answer keyword recognition submodule extracts keywords from the student answer and the preset answer text based on the student answer text and combines it with the preset answer information to obtain a keyword list;

[0016] The text similarity analysis submodule compares the similarity of text keywords based on the keyword list, analyzes the keyword coverage, semantic consistency and information completeness of the student answers, evaluates the accuracy, completeness and relevance of the student answers, and obtains a similarity evaluation result;

[0017] The comprehension level analysis submodule evaluates the student's comprehension level of the reading text based on the similarity evaluation result and generates an answer verification result.

[0018] As a further embodiment of the present invention, the specific formula for evaluating the student's understanding of the reading text is:

[0019]

[0020] Among them, S represents the student's overall understanding score, C represents the keyword coverage score, S represents the semantic consistency score, I represents the information completeness score, w1 is the weight coefficient of the keyword coverage score, w2 is the weight coefficient of the semantic consistency score, and w3 is the weight coefficient of the information completeness score.

[0021] As a further solution of the present invention, the reading pattern analysis module includes:

[0022] The reading behavior monitoring submodule monitors the student's reading and answering behavior in real time based on the answer verification result, records the student's reading start and end time points, and obtains reading time measurement data;

[0023] The reading stop marking submodule detects and identifies the student's stop points during the reading process based on the reading time measurement data, records the stop duration and stop frequency, and obtains the stop behavior recognition result;

[0024] The behavior pattern recording submodule monitors the students' annotation behavior of the text based on the dwell behavior recognition result, identifies the target students' reading behavior pattern, and generates a reading behavior monitoring record.

[0025] As a further embodiment of the present invention, the attention ability assessment module includes:

[0026] The real-time behavior detection submodule detects the student's non-reading behaviors during the reading process based on the reading behavior monitoring records, including seat-leaving, screen switching, and irrelevant notes, records the frequency and duration of the non-reading behaviors, and generates a non-reading behavior detection record;

[0027] The attention distraction identification submodule identifies and marks the student's attention distraction time period during the reading process based on the non-reading behavior detection record, and generates distraction time period information;

[0028] The attention level assessment submodule evaluates the persistence and concentration of students' attention based on the dispersed time period information by analyzing the fluctuation of students' attention on reading tasks of various difficulty levels, and generates attention analysis results.

[0029] As a further solution of the present invention, the efficiency score calculation module includes:

[0030] The reading speed assessment submodule extracts the reading time and the frequency of interruptions during reading of multiple students based on the attention analysis results, calculates the reading speeds of the multiple students, and generates a reading time analysis result;

[0031] The reading quality analysis submodule uses the reading time analysis results to evaluate students' reading quality by analyzing the duration and frequency of text annotation behavior and reading pause points, combined with the students' understanding of the reading text, and generates a reading behavior evaluation result;

[0032] The behavior efficiency analysis submodule calculates reading efficiency scores for multiple students based on the reading behavior evaluation results, according to reading speed and reading quality, and generates a reading efficiency evaluation result.

[0033] As a further embodiment of the present invention, the specific formula for calculating the reading efficiency scores of multiple students is:

[0034]

[0035] Among them, E represents the student's reading efficiency score, w v is the weight coefficient of reading speed score, N represents the total number of words in the target reading comprehension text, T represents the total reading time, and w q is the weight coefficient of the reading quality score, C represents the number of correct answers, and Q represents the total number of questions.

[0036] As a further embodiment of the present invention, the comprehension impairment identification module includes:

[0037] The inefficient behavior identification submodule detects inefficient reading behaviors of students during reading, including excessively long pauses on paragraphs and frequent rereading of previous texts, based on the reading efficiency evaluation results, and generates inefficient behavior detection results.

[0038] The obstacle point marking submodule identifies key obstacle points that affect reading efficiency in the reading text based on the inefficient behavior detection result, and generates a reading obstacle point identification result;

[0039] The barrier point classification submodule analyzes and classifies multiple barrier points in the text based on the barrier point identification results, including grammatical difficulties, vocabulary difficulties and phrase difficulties, and generates a barrier point diagnosis record.

[0040] As a further embodiment of the present invention, the reading ability assessment module includes:

[0041] The behavioral performance analysis submodule analyzes and records the student's performance in various reading difficulty tasks based on the dyslexia diagnosis record, including text comprehension, attention level, reading efficiency, and dyslexia, to obtain a student performance dataset;

[0042] The comprehensive ability assessment submodule, based on the student performance data set, analyzes the student's knowledge level by evaluating the student's response to reading tasks of various difficulty levels and generates a reading ability analysis result;

[0043] The ability level assessment submodule assesses the student's English reading comprehension ability level in real time based on the reading ability analysis results and generates a reading comprehension ability score.

[0044] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0045] By extracting keywords from students' answer texts and comparing them with preset answers, the system can effectively identify students' understanding of English reading comprehension texts. By monitoring students' reading behavior in real time, it can identify the time periods when students' attention is concentrated and distracted, effectively evaluate students' attention levels and reading efficiency, and combine the detection of students' reading disabilities and analysis of their ability levels to help educators quickly and accurately grasp students' learning status, adjust teaching strategies and content to adapt to students' personalized needs, and effectively improve learning efficiency and education quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0047] Figure 1 is a system flow chart of the present invention;

[0048] Figure 2 Schematic diagram of the system framework of the present invention;

[0049] Figure 3 This is a flow chart of the answer text analysis module of the present invention;

[0050] Figure 4 This is a flow chart of the reading mode analysis module of the present invention;

[0051] Figure 5 This is a flow chart of the attention ability assessment module of the present invention;

[0052] Figure 6 This is a flow chart of the efficiency score calculation module of the present invention;

[0053] Figure 7 This is a flow chart of the comprehension barrier identification module of the present invention;

[0054] Figure 8 This is a flow chart of the reading ability assessment module of the present invention. DETAILED DESCRIPTION

[0055] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0056] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0057] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0058] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0059] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0060] The embodiment of the present invention provides an English reading comprehension ability analysis system based on deep learning, please refer to Figures 1 to 2 The present invention provides a technical solution, an English reading comprehension ability analysis system based on deep learning, comprising:

[0061] The answer text analysis module extracts keywords from the student answer text and combines it with pre-set answer information. It then evaluates the accuracy, completeness, and relevance of the student answer, analyzes the student's understanding of the reading text, and generates answer verification results.

[0062] The reading pattern analysis module verifies the results of the answers and monitors students’ reading and answering behaviors in real time to identify the target students’ reading behavior patterns and generate reading behavior monitoring records.

[0063] The attention ability assessment module evaluates students' attention persistence and concentration based on reading behavior monitoring records by identifying periods of distraction during reading and generating attention analysis results.

[0064] The efficiency score calculation module analyzes the reading fluency and text processing ability of multiple students based on the attention analysis results and their reading behaviors, calculates the reading efficiency scores of multiple students, and forms a reading efficiency evaluation result;

[0065] The comprehension barrier identification module detects students' inefficient reading behaviors through reading efficiency assessment results, identifies target students' comprehension barriers, classifies the barriers, and generates dyslexia diagnosis records.

[0066] The reading ability assessment module is based on the dyslexia diagnosis records and the students' performance in reading tasks of various difficulty levels. It evaluates students' reading comprehension ability in real time and generates a reading comprehension ability score.

[0067] The answer verification results specifically include keyword coverage analysis data, semantic consistency analysis results, and information completeness assessment information. The reading behavior monitoring records include reading start and end time points, stop point behavior marks, and annotation behavior monitoring records. The attention analysis results specifically refer to the attention concentration time period marking information, distracted time period identification results, and persistence level. The reading efficiency assessment results specifically include reading time, annotation behavior frequency, and stop point length. The dyslexia diagnosis records include grammatical difficulties, vocabulary difficulties, and phrase difficulties. The reading comprehension ability score specifically refers to the difficulty of the reading task, the degree of text comprehension, and the reading behavior performance.

[0068] See also Figure 2 and Figure 3 , the answer text analysis module includes:

[0069] The answer keyword recognition submodule extracts keywords from the student answer and the preset answer text based on the student answer text and combines it with the preset answer information to obtain a keyword list;

[0070] Keywords are extracted through text processing technology. The process includes applying natural language processing technology to first perform language analysis on the student answer text, including part-of-speech tagging and syntactic analysis, identifying key information, and using preset answer information as a reference to extract its keywords using text processing technology to ensure the processing consistency of the two text sources. Through text mining algorithms such as word frequency-inverse document frequency, the keywords in the text are analyzed and a keyword list is generated. The process involves data extraction, including data cleaning and normalization processing, ensuring the quality and availability of keyword data, and extracting the most representative words from a large amount of text, laying a data foundation for subsequent similarity analysis.

[0071] The text similarity analysis submodule compares the similarity of text keywords based on the keyword list, analyzes the keyword coverage, semantic consistency and information completeness of students' answers, evaluates the accuracy, completeness and relevance of students' answers, and obtains similarity evaluation results;

[0072] Use the cosine similarity algorithm to perform text comparison, build a keyword vector space model, convert the keywords of student answers and preset answers into vector form, calculate the cosine value between the two vectors, determine the similarity of keywords in the two documents, quantitatively analyze the keyword coverage, that is, the degree of matching between the keywords in the student answers and the keywords in the preset answers, evaluate semantic consistency, check the contextual use of the same or similar keywords, and information completeness, evaluate whether the student answers fully respond to the information points in the preset answers, help teachers understand in which aspects the students are consistent with the standard answers and in which aspects there are deviations, and make a more accurate assessment of students' learning situation.

[0073] The comprehension analysis submodule evaluates students’ comprehension of the reading text based on the similarity assessment results and generates answer verification results;

[0074] The specific formula for assessing students' comprehension of a reading text is:

[0075]

[0076] Among them, S represents the student's overall understanding score, C represents the keyword coverage score, S represents the semantic consistency score, I represents the information completeness score, w1 is the weight coefficient of the keyword coverage score, w2 is the weight coefficient of the semantic consistency score, and w3 is the weight coefficient of the information completeness score.

[0077] formula:

[0078]

[0079] Detailed explanation of the formula and the process of formula calculation and derivation:

[0080] The formula is used to comprehensively evaluate students' understanding of the reading text and obtain a quantitative score of students' understanding of the target reading comprehension text.

[0081] Parameter meaning and setting value:

[0082] C is the keyword coverage score, which is assumed to be 85%, reflecting the coverage of the student's answer and the standard answer;

[0083] S is the semantic consistency score, which is assumed to be 90%, indicating how close the student's answer is semantically to the preset answer;

[0084] I is the information completeness score, which is assumed to be 80%, showing the information completeness of the student's answer;

[0085] w1, w2, and w3 are the weight coefficients of keyword coverage, semantic consistency, and information completeness, respectively. Assume that w1 = 0.3, w2 = 0.4, and w3 = 0.3;

[0086] Substitute the parameters into the formula for calculation:

[0087]

[0088] S=85.5;

[0089] The result S=85.5 indicates that the student's total understanding score is 85.5 points. The higher score reflects that the student has a better understanding and grasp of the reading material. The formula evaluates the student's reading comprehension ability in a quantitative way, improving the comprehensiveness and accuracy of the evaluation.

[0090] See also Figure 2 and Figure 4 , the reading pattern analysis module includes:

[0091] The reading behavior monitoring submodule monitors students’ reading and answering behaviors in real time based on the answer verification results, records the students’ reading start and end time points, and obtains reading time measurement data;

[0092] Real-time data tracking technology is used to continuously monitor students' reading and answering activities. The system collects students' reading start and end times through sensors and interface interaction data, and uses timestamps to mark the specific moments of starting and ending reading. The time point data is entered into the database in real time. By analyzing the time points, the total reading time of each student is calculated. The process involves time series analysis, including real-time database query and time difference calculation to ensure the accuracy and real-time nature of time measurement. Real-time monitoring allows educators to observe students' specific reading patterns, how to switch between different paragraphs, and how long they stay on specific content, providing data support for subsequent teaching improvements.

[0093] The reading stop marking submodule detects and identifies students' stop points during the reading process based on the reading time measurement data, records the stop duration and stop frequency, and obtains the stop behavior recognition results;

[0094] Use pattern recognition technology to determine the specific points where students stay in the text. By analyzing students' interactive behaviors in the text, such as mouse movement, scrolling behavior, and page pauses, the time and frequency of the behaviors are automatically recorded. Cluster analysis is performed on the pause points to identify which parts students spend more time on, which usually indicates the difficulty or key areas of reading. Through this technology, the duration and frequency of each pause point are recorded, and statistical analysis methods, such as clustering algorithms, are used to convert the pause behavior into quantitative data. The data reflects the depth and focus of students' reading, providing teachers with a basis for further personalized teaching.

[0095] The behavior pattern recording submodule monitors students’ annotation behavior of the text based on the results of the dwell behavior recognition, identifies the target students’ reading behavior patterns, and generates reading behavior monitoring records;

[0096] Continue to track and analyze students' annotation behaviors, including interactive behaviors such as highlighting, annotating, and adding bookmarks to texts. Through data mining techniques such as sequence analysis, the system records the time and place of the behavior, and identifies students' reading strategies and preferences from behavioral patterns. Each annotation behavior is regarded as an in-depth processing of the reading content. After analysis, the data can reveal students' level of understanding of the text and problem-solving ability. The information is aggregated to form a comprehensive reading behavior monitoring record, which includes the specific content and context of the annotations, and the distribution in the time dimension. It provides teachers with a multi-dimensional view of students' reading behavior, helping teachers to adjust classes and reading materials.

[0097] See also Figure 2 and Figure 5 , the attention ability assessment module includes:

[0098] The real-time behavior detection submodule detects students' non-reading behaviors during the reading process based on reading behavior monitoring records, including leaving their seats, screen switching, and irrelevant notes. It also records the frequency and duration of non-reading behaviors and generates non-reading behavior detection records.

[0099] Using sensor technology and user interface tracking technology, students' non-reading related behaviors are monitored. Behavior recognition algorithms, including support vector machines, are used to analyze data collected from students' interactive devices, such as keyboard activity, mouse movement, and camera input. The data helps the system identify the frequency and duration of behaviors such as leaving the seat, screen switching, and taking notes on the reading platform. Each behavior is recorded, including its start and end time points, and the occurrence of the behavior is recorded during the process. The duration and frequency of the behavior are analyzed to accurately divide reading and non-reading time. Detailed records of non-reading behaviors are crucial for understanding students' overall engagement and distraction during the reading process, providing teachers with precise data support to adjust teaching methods and content.

[0100] The attention distraction identification submodule identifies and marks the time periods when students are distracted during reading based on the non-reading behavior detection records, and generates distraction time period information;

[0101] Time series analysis techniques, such as the autoregressive moving average model, are used to identify and mark specific periods when students are distracted during the reading process. By analyzing students' behavioral patterns, it is determined which periods are high-frequency non-reading behaviors and are regarded as distraction periods. Information about the time periods is recorded and marked, including the duration of each distraction segment and its frequency relative to the entire reading period. The analysis helps teachers identify the passages where students are most easily distracted while reading, and observe whether there is any regularity in these distracting behaviors, such as whether they are more frequent in specific types of reading materials or within a certain time period.

[0102] The attention level assessment submodule evaluates students' attention persistence and concentration based on scattered time period information by analyzing students' attention fluctuations on reading tasks of various difficulty levels, and generates attention analysis results;

[0103] In the attention level assessment submodule, based on the scattered time period information obtained from the non-reading behavior detection records, the attention fluctuations of students on various reading tasks are assessed using the formula Calculate the average concentration of students,

[0104] Where A is the average concentration of attention, d i represents the duration of the i-th distraction, w i is a weight adjusted based on the difficulty of the reading task, N is the total number of distracting events identified during the measurement period, i is the index of the distracting event,

[0105] Detailed explanation of the formula and the process of formula calculation and derivation:

[0106] Assume that there are three distraction events in the target measurement period, with durations d1 = 30 seconds, d2 = 45 seconds, and d3 = 15 seconds, and the corresponding fluctuation weights w are w1 = 0.2, w2 = 0.5, and w3 = 0.1, respectively.

[0107] N=3, calculate the student's concentration score:

[0108]

[0109] The calculation result A=10 shows that the average distraction time of students during the reading process is relatively short, indicating that under the current reading materials and environment settings, students can maintain good concentration. The analysis helps educators understand students' attention status and provide effective learning guidance and environmental adjustments.

[0110] See also Figure 2 and Figure 6 , the efficiency score calculation module includes:

[0111] The reading speed assessment submodule extracts the reading time and interruption frequency of multiple students based on the attention analysis results, calculates the reading speed of multiple students, and generates reading time analysis results;

[0112] Record the timestamps from the start to the end of each student's reading, monitor reading interruptions, including page switching or stopping reading, and automatically collect and record data through behavior recognition technology. Apply autoregressive models to process time series data and calculate each student's reading speed. The speed is calculated by dividing the total number of words read by the net reading time. Track and analyze students' reading behavior and generate personalized reading time analysis results for each student. Through analysis, educators can understand in detail the changes in speed and interruptions of each student during the reading process, and further identify possible reading difficulties or areas requiring additional tutoring.

[0113] The reading quality analysis submodule uses the reading time analysis results to evaluate students' reading quality by analyzing the duration and frequency of text annotation behavior and reading pause points, combined with the students' understanding of the reading text, and generates reading behavior evaluation results;

[0114] Through behavioral pattern recognition technologies such as hidden Markov models, students' reading quality is assessed, and the number of annotation behaviors and the length and frequency of dwell time in each reading session are tracked. The data is processed by text analysis tools such as natural language processing to understand students' attention to and depth of understanding of the target text section. Based on the target data, each student's performance on different texts is evaluated and associated with the degree of comprehension, generating comprehensive reading behavior assessment results. The results reflect how students interact with reading materials and provide an intuitive measure of comprehension quality and efficiency, allowing educators to adjust teaching strategies and materials in a targeted manner to improve learning efficiency.

[0115] The behavior efficiency analysis submodule calculates reading efficiency scores for multiple students based on the reading behavior assessment results, according to reading speed and reading quality, and generates reading efficiency assessment results;

[0116] The specific formula for calculating the reading efficiency score for multiple students is:

[0117]

[0118] Among them, E represents the student's reading efficiency score, w v is the weight coefficient of reading speed score, N represents the total number of words in the target reading comprehension text, T represents the total reading time, and w q is the weight coefficient of the reading quality score, C represents the number of correct answers, and Q represents the total number of questions.

[0119] formula:

[0120]

[0121] Detailed explanation of the formula and the process of formula calculation and derivation:

[0122] The formula is used to calculate a reading efficiency score, and the results are used to assess students’ reading efficiency;

[0123] Parameter meaning and setting value:

[0124] N is the total number of words, assuming it is 6000 words;

[0125] T is the total reading time, assuming it is 30 minutes;

[0126] C is the number of correct answers, assuming it is 15;

[0127] Q is the total number of questions, assuming it is 20;

[0128] w v is the weight coefficient of reading speed score, assumed to be 0.6;

[0129] w q is the weight coefficient of the reading quality score, which is assumed to be 0.4;

[0130] Substitute the parameters into the formula for calculation:

[0131]

[0132]

[0133] The result 154.92 indicates that the student's reading efficiency score is 154.92. The score reflects the student's performance in reading speed and comprehension quality. The formula is used to evaluate the target student's reading efficiency for the corresponding reading comprehension material. The result provides a data basis for evaluating students' English reading comprehension ability.

[0134] See also Figure 2 and Figure 7 ,Understanding the obstacle identification module includes:

[0135] The inefficient behavior identification submodule detects inefficient reading behaviors in students during reading based on the reading efficiency assessment results, including staying in paragraphs for too long and frequently looking back at previous texts, and generates inefficient behavior detection results;

[0136] By analyzing the time each student spends on multiple paragraphs and comparing them with the standard reading time model, we can determine whether there are abnormal behaviors of staying too long. We can track and record students' frequent revisiting of previous texts during the reading process. Through pattern recognition technology, such as support vector machines, we can determine whether the frequency of target behavior exceeds the range of normal learning behavior. By collecting specific timestamp data and the number of revisits, we use quantitative methods to determine which reading habits may lead to reduced learning efficiency and generate inefficient behavior detection results for each student. The results provide teachers with data-based insights to adjust teaching strategies or provide personalized tutoring to improve students' reading habits.

[0137] The obstacle point marking submodule identifies key obstacle points that affect reading efficiency in the reading text based on the inefficient behavior detection results and generates reading obstacle point identification results;

[0138] The obstacle point marking submodule uses text analysis technologies, such as deep learning natural language processing algorithms, to extract data from inefficient behavior detection results and identify key obstacles that lead to low reading efficiency. By analyzing students' stop point data and frequent review behaviors in the reading text, the possible reasons behind the target behavior, such as complex grammatical structures or difficult-to-understand vocabulary, are determined. By matching the target behavior with the specific content in the text, the key obstacles that affect reading fluency are accurately marked. The generated obstacle point identification results provide teachers with a clear goal, allowing them to make teaching adjustments or provide additional explanatory materials based on the target obstacle points to help students overcome reading difficulties and optimize learning outcomes.

[0139] The barrier point classification submodule analyzes and classifies multiple barrier points in the text based on the barrier point identification results, including grammatical difficulties, vocabulary difficulties, and phrase difficulties, and generates a barrier point diagnosis record.

[0140] Classification algorithms, such as decision trees, are used to conduct detailed analysis and classification of various reading difficulties. Different categories of reading difficulties are systematically defined, including grammatical difficulties, vocabulary difficulties, and phrase difficulties. The frequency of occurrence of each obstacle point in the text and the degree of its impact on students' reading behavior are analyzed through algorithms. Through this classification, detailed reading difficulty diagnosis records are generated, listing the main problems encountered by students in the reading process and their categories, helping teachers understand the specific difficulties encountered by students in reading and develop more effective teaching plans.

[0141] See also Figure 2 and Figure 8 , the reading ability assessment modules include:

[0142] The behavioral performance analysis submodule analyzes and records students' performance in various reading difficulty tasks based on dyslexia diagnosis records, including text comprehension, attention level, reading efficiency, and dyslexia, to obtain a student performance dataset;

[0143] Using data analysis techniques, such as cluster analysis, to process detailed records from dyslexia diagnosis involves collecting and organizing data on students' performance in different reading tasks, including multiple dimensions such as text comprehension, attention level, reading efficiency, and identified dyslexia. Pattern recognition is performed on each student's behavioral data, and time series analysis is used to track students' performance trends. The data is aggregated into a student performance dataset. By analyzing students' performance on different types of texts and difficulty levels, the dataset reveals students' learning habits and potential learning barriers, providing data support for subsequent teaching interventions and learning support.

[0144] The comprehensive ability assessment submodule is based on the student performance dataset. It analyzes students' knowledge level by evaluating their response to reading tasks of various difficulty levels and generates reading ability analysis results.

[0145] Machine learning algorithms, including random forests, are used to analyze student performance datasets and evaluate students' responses to various reading tasks. The process involves scoring each student's reading speed, comprehension ability, and depth of task completion, and normalizing the data to eliminate measurement bias. Through target analysis, we can understand the students' knowledge level in detail, and combine the difficulty of various reading tasks to generate personalized reading ability analysis results, helping educators understand the performance differences of each student when faced with different types of reading materials, and better adjust teaching content and methods to meet the specific needs of students.

[0146] The ability level assessment submodule evaluates students' English reading comprehension ability in real time based on the reading ability analysis results and generates a reading comprehension ability score;

[0147] Using real-time analysis technologies, such as the real-time data stream processing framework Apache Kafka, combined with the obtained reading ability analysis results, students' reading comprehension ability can be instantly assessed. By continuously monitoring students' reading behavior and test scores, their reading comprehension ability scores are dynamically calculated. During the scoring process, multiple factors such as students' reading speed, error rate, depth of comprehension, and duration are combined through a weighted algorithm to form an ability score. The score provides teachers with a quantitative tool to evaluate the improvement or decline of students' reading ability, and to adjust teaching strategies in a timely manner to help students make progress in reading comprehension.

[0148] The above embodiments can be implemented in whole or in part through software, hardware (such as circuits), firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired method (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0149] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0150] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0151] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0152] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0153] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0154] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, 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 interface, indirect coupling or communication connection of the device or unit, which can be electrical, mechanical or other forms.

[0155] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0156] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0157] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0158] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. English reading comprehension ability analysis system based on deep learning, characterized by: The system comprises: The answer text analysis module extracts keywords from the student answer text and combines it with pre-set answer information. It then evaluates the accuracy, completeness, and relevance of the student answer, analyzes the student's understanding of the reading text, and generates answer verification results. The reading pattern analysis module identifies the target student's reading behavior pattern based on the answer verification result by monitoring the student's reading and answering behavior in real time, and obtains a reading behavior monitoring record; The attention ability assessment module evaluates the persistence and concentration of the student's attention by identifying the time periods of distraction during the reading process based on the reading behavior monitoring records, and generates an attention analysis result; The efficiency score calculation module analyzes the reading fluency and text processing ability of multiple students based on the attention analysis results and the students' reading behaviors, calculates the reading efficiency scores of the multiple students, and forms a reading efficiency evaluation result; The comprehension barrier identification module detects the student's inefficient reading behavior based on the reading efficiency evaluation results, identifies the target student's comprehension barrier points, classifies the barrier points, and generates a reading barrier diagnosis record; The reading ability assessment module evaluates the student's reading comprehension ability in real time based on the reading disability diagnosis record and the student's performance in reading tasks of various difficulty levels, and generates a reading comprehension ability score.

2. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The answer verification results specifically include keyword coverage analysis data, semantic consistency analysis results, and information integrity assessment information. The reading behavior monitoring records include reading start and end time points, stop point behavior marks, and annotation behavior monitoring records. The attention analysis results specifically refer to attention concentration time period marking information, distracted time period identification results, and persistence level. The reading efficiency assessment results specifically include reading time, annotation behavior frequency, and stop point duration. The dyslexia diagnosis records include grammatical difficulties, vocabulary difficulties, and phrase difficulties. The reading comprehension ability score specifically refers to the difficulty of the reading task, the degree of text comprehension, and reading behavior performance.

3. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The answer text analysis module includes: The answer keyword recognition submodule extracts keywords from the student answer and the preset answer text based on the student answer text and combines it with the preset answer information to obtain a keyword list; The text similarity analysis submodule compares the similarity of text keywords based on the keyword list, analyzes the keyword coverage, semantic consistency and information completeness of the student answers, evaluates the accuracy, completeness and relevance of the student answers, and obtains a similarity evaluation result; The comprehension level analysis submodule evaluates the student's comprehension level of the reading text based on the similarity evaluation result and generates an answer verification result.

4. The English reading comprehension ability analysis system based on deep learning according to claim 3 is characterized in that: The specific formula for evaluating students' understanding of the reading text is: Among them, S represents the student's overall understanding score, C represents the keyword coverage score, S represents the semantic consistency score, I represents the information completeness score, w1 is the weight coefficient of the keyword coverage score, w2 is the weight coefficient of the semantic consistency score, and w3 is the weight coefficient of the information completeness score.

5. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The reading pattern analysis module includes: The reading behavior monitoring submodule monitors the student's reading and answering behavior in real time based on the answer verification result, records the student's reading start and end time points, and obtains reading time measurement data; The reading stop marking submodule detects and identifies the student's stop points during the reading process based on the reading time measurement data, records the stop duration and stop frequency, and obtains the stop behavior recognition result; The behavior pattern recording submodule monitors the students' annotation behavior of the text based on the dwell behavior recognition result, identifies the target students' reading behavior pattern, and generates a reading behavior monitoring record.

6. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The attention ability assessment module includes: The real-time behavior detection submodule detects the student's non-reading behaviors during the reading process based on the reading behavior monitoring records, including seat-leaving, screen switching, and irrelevant notes, records the frequency and duration of the non-reading behaviors, and generates a non-reading behavior detection record; The attention distraction identification submodule identifies and marks the student's attention distraction time period during the reading process based on the non-reading behavior detection record, and generates distraction time period information; The attention level assessment submodule evaluates the persistence and concentration of students' attention based on the dispersed time period information by analyzing the fluctuation of students' attention on reading tasks of various difficulty levels, and generates attention analysis results.

7. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The efficiency score calculation module includes: The reading speed assessment submodule extracts the reading time and the frequency of interruptions during reading of multiple students based on the attention analysis results, calculates the reading speeds of the multiple students, and generates a reading time analysis result; The reading quality analysis submodule uses the reading time analysis results to evaluate students' reading quality by analyzing the duration and frequency of text annotation behavior and reading pause points, combined with the students' understanding of the reading text, and generates a reading behavior evaluation result; The behavior efficiency analysis submodule calculates reading efficiency scores for multiple students based on the reading behavior evaluation results, according to reading speed and reading quality, and generates a reading efficiency evaluation result.

8. The English reading comprehension ability analysis system based on deep learning according to claim 7 is characterized in that: The specific formula for calculating the reading efficiency scores of multiple students is: Among them, E represents the student's reading efficiency score, w v is the weight coefficient of reading speed score, N represents the total number of words in the target reading comprehension text, T represents the total reading time, and w q is the weight coefficient of the reading quality score, C represents the number of correct answers, and Q represents the total number of questions.

9. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The comprehension barrier identification module includes: The inefficient behavior identification submodule detects inefficient reading behaviors of students during reading, including excessively long pauses on paragraphs and frequent rereading of previous texts, based on the reading efficiency evaluation results, and generates inefficient behavior detection results. The obstacle point marking submodule identifies key obstacle points that affect reading efficiency in the reading text based on the inefficient behavior detection result, and generates a reading obstacle point identification result; The barrier point classification submodule analyzes and classifies multiple barrier points in the text based on the barrier point identification results, including grammatical difficulties, vocabulary difficulties and phrase difficulties, and generates a barrier point diagnosis record.

10. The English reading comprehension ability analysis system based on deep learning according to claim 1 is characterized in that: The reading ability assessment module includes: The behavioral performance analysis submodule analyzes and records the student's performance in various reading difficulty tasks based on the dyslexia diagnosis record, including text comprehension, attention level, reading efficiency, and dyslexia, to obtain a student performance dataset; The comprehensive ability assessment submodule is based on the student performance data set, analyzes the student's knowledge level by evaluating the student's response to reading tasks of various difficulty levels, and generates a reading ability analysis result; The ability level assessment submodule assesses the student's English reading comprehension ability level in real time based on the reading ability analysis results and generates a reading comprehension ability score.

Images