An AI-based method and system for evaluating the teaching effect of foreign language teaching

Abstract

The present invention relates to the technical field of teaching effect evaluation, and particularly to an AI-based foreign language teaching effect evaluation method and system. The method includes: performing text paragraph logic segmentation on the foreign language writing assessment content text to obtain the paragraph logic segmentation data of the foreign language writing assessment content; performing mother tongue system interference recognition on the foreign language writing assessment content text based on the paragraph logic segmentation data to obtain the mother tongue system interference recognition data of the content paragraph; performing adaptive language system anomaly recognition optimization according to the mother tongue system interference recognition data of the content paragraph to obtain the AI language system anomaly recognition architecture; performing foreign language teaching effect evaluation according to the AI language system anomaly recognition architecture to obtain the foreign language teaching effect evaluation data, and sending the foreign language teaching effect evaluation data to the terminal. The present invention makes the AI foreign language teaching effect evaluation method more perfect through the optimization process of the AI foreign language teaching effect evaluation method.

Description

Technical Field

[0001] The present invention relates to the technical field of teaching effect evaluation, and particularly to an AI-based foreign language teaching effect evaluation method and system. Background Art

[0002] Previous foreign language teaching evaluation methods mostly relied on teachers' subjective judgments and students' language ability tests. However, these methods often failed to comprehensively and accurately reflect students' true language levels and the difficulties they faced in the process of foreign language learning. With the continuous progress of artificial intelligence (AI) technology, especially the development of technologies such as natural language processing (NLP) and machine learning, an AI-based foreign language teaching effect evaluation method has emerged. This method can analyze students' language output, identify their language errors, interference factors, and learning progress, and then provide a more objective and accurate teaching evaluation. However, a traditional AI-based foreign language teaching effect evaluation method has problems such as inaccurate logical analysis of students' writing information and large errors in identifying the influence of students' native language structures, resulting in low accuracy in teaching effect evaluation. Summary of the Invention

[0003] Based on this, it is necessary to provide an AI-based foreign language teaching effect evaluation method to solve at least one of the above technical problems.

[0004] To achieve the above object, an AI-based foreign language teaching effect evaluation method, the method includes the following steps:

[0005] Step S1: Obtain the text of the foreign language writing assessment content submitted by students; perform text paragraph logical segmentation on the text of the foreign language writing assessment content to obtain the paragraph logical segmentation data of the foreign language writing assessment content;

[0006] Step S2: Quantify the interference gradient of the information entropy sequence for the text of the foreign language writing assessment content based on the paragraph logical segmentation data to obtain the information entropy sequence interference gradient data; identify the interference of the native language system for the text of the foreign language writing assessment content according to the information entropy sequence interference gradient data to obtain the content paragraph native language system interference identification data;

[0007] Step S3: Obtain the underlying learning architecture of the AI tool; optimize the adaptive language system anomaly identification for the underlying learning architecture of the AI tool according to the content paragraph native language system interference identification data to obtain the AI language system anomaly identification architecture;

[0008] Step S4: Evaluate the foreign language teaching effect for the text of the foreign language writing assessment content according to the AI language system anomaly identification architecture to obtain the foreign language teaching effect evaluation data, and send the foreign language teaching effect evaluation data to the terminal.

[0009] Preferably, step S1 includes the following steps:

[0010] Step S11: Obtain the text of the foreign language writing assessment content submitted by the student;

[0011] Step S12: Clean the text of the foreign language writing assessment content to obtain the cleaned text of the foreign language writing assessment content;

[0012] Step S13: Conduct a narrative analysis of the text structure of the cleaned text of the foreign language writing assessment content to obtain the narrative data of the text structure of the writing content;

[0013] Step S14: Perform logical segmentation of the paragraphs of the cleaned text of the foreign language writing assessment content according to the narrative data of the text structure of the writing content to obtain the paragraph logical segmentation data of the foreign language writing assessment content.

[0014] Preferably, step S2 includes the following steps:

[0015] Step S21: Conduct an analysis of parallel / subordinate connections of sentence patterns on the text of the foreign language writing assessment content according to the paragraph logical segmentation data to obtain the data of parallel / subordinate connections of sentence patterns;

[0016] Step S22: Identify the logical jump links of the text of the foreign language writing assessment content based on the data of parallel / subordinate connections of sentence patterns and the paragraph logical segmentation data to obtain the content paragraph logical jump data;

[0017] Step S23: Quantify the interference gradient of the information entropy sequence of the text of the foreign language writing assessment content based on the content paragraph logical jump data to obtain the information entropy sequence interference gradient data;

[0018] Step S24: Identify the interference of the mother tongue system on the text of the foreign language writing assessment content according to the information entropy sequence interference gradient data and the content paragraph logical jump data to obtain the content paragraph mother tongue system interference identification data.

[0019] Preferably, step S23 includes the following steps:

[0020] Step S231: Count the frequency of logical jump connecting words in the text of the foreign language writing assessment content based on the content paragraph logical jump data to obtain the frequency data of logical jump connecting words;

[0021] Step S232: Decompose the semantic span word vector of the connecting words through the frequency data of logical jump connecting words to obtain the semantic span word vector of the connecting words;

[0022] Step S233: Conduct a regression analysis of the grammatical ambiguity of the front and back paragraphs on the text of the foreign language writing assessment content according to the semantic span word vector of the connecting words and the frequency data of logical jump connecting words to obtain the regression data of the grammatical ambiguity of the front and back paragraphs;

[0023] Step S234: Based on the semantic span word vectors of connecting words and the grammatical fuzzy regression data of the preceding and following paragraphs, conduct a simulation analysis of the information entropy sequence dimension constraint disorder state to obtain the information entropy sequence dimension constraint disorder state;

[0024] Step S235: Quantify the interference gradient of the information entropy sequence for the information entropy sequence dimension constraint disorder state to obtain the information entropy sequence interference gradient data.

[0025] Preferably, step S234 includes the following steps:

[0026] Conduct a polysemy clustering analysis on the semantic span word vectors of connecting words to obtain the polysemy clustering data of connecting words;

[0027] Conduct a grammatical boundary fuzzy trend analysis on the grammatical fuzzy regression data of the preceding and following paragraphs to obtain the grammatical boundary fuzzy trend data of the preceding and following paragraphs;

[0028] Based on the polysemy clustering data of connecting words and the grammatical boundary fuzzy trend data of the preceding and following paragraphs, conduct a context change boundary fitting to obtain the context change boundary fitting data before and after the paragraph;

[0029] Based on the Markov chain, conduct a context boundary fitting state simulation on the context change boundary fitting data before and after the paragraph to obtain the context boundary state fitting data;

[0030] According to the context boundary state fitting data, conduct a simulation analysis of the information entropy sequence dimension constraint disorder state to obtain the information entropy sequence dimension constraint disorder state.

[0031] Preferably, step S24 includes the following steps:

[0032] Step S241: Obtain the native language system structure data of different students;

[0033] Step S242: Based on the content paragraph logical jump data and the native language system structure data, conduct a native language transfer feature analysis on the foreign language writing assessment content text to obtain the assessment content native language transfer feature data;

[0034] Step S243: Based on the assessment content native language transfer feature data and the content paragraph logical jump data, conduct a syntactic structure logical deviation identification to obtain the native language syntactic structure logical deviation data;

[0035] Step S244: Conduct a semantic deviation identification on the assessment content native language transfer feature data to obtain the native language semantic feature deviation data;

[0036] Step S245: Based on the native language syntactic structure logical deviation data and the native language semantic feature deviation data, conduct a native language dominant syntactic framework mapping on the foreign language writing assessment content text to obtain the native language dominant syntactic framework mapping data;

[0037] Step S246: Identify the interference in the native language system based on the information entropy sequence interference gradient data and the native language dominant syntactic framework mapping data, and obtain the content paragraph native language system interference identification data.

[0038] Preferably, step S3 includes the following steps:

[0039] Step S31: Obtain the underlying learning architecture of the AI tool;

[0040] Step S32: Normalize the information entropy sequence interference gradient data to obtain the information entropy sequence interference gradient normalized data;

[0041] Step S33: Identify and learn to strengthen the disorder degree of the AI architecture based on the information entropy sequence interference gradient normalized data for the underlying learning architecture of the AI tool, and obtain the AI disorder degree identification and learning strengthened architecture;

[0042] Step S34: Based on the content paragraph native language system interference identification data, perform adaptive language system anomaly identification and optimization on the AI disorder degree identification and learning strengthened architecture to obtain the AI language system anomaly identification architecture.

[0043] Preferably, step S33 includes the following steps:

[0044] Step S331: Perform topological information entropy discretization processing on the information entropy sequence interference gradient normalized data to obtain the information entropy topological discrete feature space;

[0045] Step S332: Perform feature orthogonal mapping and information entropy reconstruction processing on the information entropy topological discrete feature space to obtain the feature mapping information entropy tensor;

[0046] Step S333: Perform interference gradient calibration processing according to the feature mapping information entropy tensor to obtain the interference gradient calibration data;

[0047] Step S334: Based on the interference gradient calibration data, identify and learn to strengthen the disorder degree of the AI architecture for the underlying learning architecture of the AI tool to obtain the AI disorder degree identification and learning strengthened architecture.

[0048] Preferably, step S34 includes the following steps:

[0049] Step S341: Perform logical analysis of the native language part-of-speech interference anomalies on the content paragraph native language system interference identification data to obtain the native language part-of-speech interference anomaly logical data;

[0050] Step S342: Based on the native language part-of-speech interference anomaly logical data, perform induction of the part-of-speech logical metaphor deviation to obtain the part-of-speech logical metaphor deviation data;

[0051] Step S343: Conduct cross - domain metaphor anomaly analysis on the part - of - speech logical metaphor deviation data to obtain cross - domain metaphor anomaly data for concepts;

[0052] Step S344: Based on the part - of - speech logical metaphor deviation data and the cross - domain metaphor anomaly data for concepts, perform self - adaptive language system anomaly recognition and optimization on the AI disorder degree recognition and learning reinforcement architecture to obtain the AI language system anomaly recognition architecture.

[0053] Preferably, the present invention also provides an AI - based foreign language teaching lecture effect evaluation system for implementing the above - mentioned AI - based foreign language teaching lecture effect evaluation method. The AI - based foreign language teaching lecture effect evaluation system includes:

[0054] A paragraph logic segmentation module, configured to obtain the text of the foreign language writing assessment content submitted by students; perform text paragraph logic segmentation on the text of the foreign language writing assessment content to obtain the paragraph logic segmentation data of the foreign language writing assessment content;

[0055] An interference recognition module, configured to perform information entropy sequence interference gradient quantization on the text of the foreign language writing assessment content based on the paragraph logic segmentation data to obtain information entropy sequence interference gradient data; perform mother - tongue system interference recognition on the text of the foreign language writing assessment content according to the information entropy sequence interference gradient data to obtain content paragraph mother - tongue system interference recognition data;

[0056] An anomaly recognition and optimization module, configured to obtain the underlying learning architecture of the AI tool; perform self - adaptive language system anomaly recognition and optimization on the underlying learning architecture of the AI tool according to the content paragraph mother - tongue system interference recognition data to obtain the AI language system anomaly recognition architecture;

[0057] An effect evaluation module, configured to evaluate the foreign language teaching lecture effect on the text of the foreign language writing assessment content according to the AI language system anomaly recognition architecture to obtain foreign language teaching lecture effect evaluation data, and send the foreign language teaching lecture effect evaluation data to the terminal.

[0058] The beneficial effects of the present invention are as follows. First, by obtaining the text of the foreign language writing assessment content submitted by students, logical segmentation of the text paragraphs is carried out. The key to this process lies in decomposing the students' writing content as a whole into multiple paragraphs with clear logic, so that each paragraph can independently express a specific idea or theme. Through this segmentation, the structure of the text becomes clearer, which helps to more accurately identify the language features, logical organization and expression problems of different paragraphs in subsequent analysis, providing accurate basic data for subsequent interference identification and evaluation. Based on the paragraph logical segmentation data obtained in step S1, in the process of quantifying the interference gradient of the information entropy sequence, by analyzing the information density and complexity in the text, the language features of each paragraph are quantified. Through the change of entropy value, it can reveal which parts of the text have logical jumps or unnatural language use. In addition, according to the information entropy sequence interference gradient data, the system can identify the foreign language expression problems caused by mother tongue interference. This identification helps to detect the mother tongue traces in students' writing, and thus provides a basis for subsequent teaching intervention, effectively improving the pertinence of teaching. The underlying learning architecture of the AI tool is adaptively optimized by analyzing the mother tongue system interference identification data. The AI system adjusts the language system anomaly identification architecture according to different students' writing characteristics and mother tongue interference patterns. By continuously learning and optimizing its identification ability, the AI can improve its sensitivity to subtle problems in foreign language writing, especially for inappropriate language habits and structures, and can make more accurate judgments. This process provides the AI system with the ability of dynamic adjustment, ensuring efficient and accurate language anomaly identification in different students' learning scenarios. Finally, the optimized AI language system anomaly identification architecture will evaluate the teaching effect of the foreign language writing assessment content of students. Through comprehensively analyzing the language features, logical structure and mother tongue interference in students' writing in this step, the system can objectively and comprehensively evaluate the effect of foreign language teaching. The evaluation results will specifically reflect in which aspects students have language expression obstacles or misunderstandings, and put forward improvement suggestions accordingly. The evaluation data are fed back to teachers and students through the terminal, helping teachers adjust teaching strategies, and students can improve their language skills pertinently according to the evaluation results, thus effectively improving the foreign language learning effect. Therefore, the present invention makes an optimization treatment for a traditional AI-based foreign language teaching effect evaluation method, solves the problems existing in the traditional AI-based foreign language teaching effect evaluation method, such as inaccurate logical analysis of students' writing information, large error in identifying the influence of students' mother tongue structure, resulting in low accuracy of teaching effect evaluation, improves the accuracy of logical analysis of students' writing information, reduces the error in identifying the influence of students' mother tongue structure, and improves the accuracy of teaching effect evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1Schematic diagram of the step process of an AI-based foreign language teaching lecture effect evaluation method;

[0060] Figure 2 It is Figure 1 Schematic diagram of the detailed implementation steps of step S2 in;

[0061] Figure 3 It is Figure 1 Schematic diagram of the detailed implementation steps of step S3 in. Specific implementation manner

[0062] Please refer to Figures 1 to 3 , an AI-based foreign language teaching lecture effect evaluation method, the method includes the following steps:

[0063] Step S1: Obtain the text of the foreign language writing assessment content submitted by students; perform logical segmentation of the text paragraphs of the foreign language writing assessment content to obtain the paragraph logical segmentation data of the foreign language writing assessment content;

[0064] Step S2: Quantify the information entropy sequence interference gradient of the foreign language writing assessment content text based on the paragraph logical segmentation data to obtain information entropy sequence interference gradient data; identify the interference of the native language system on the foreign language writing assessment content text according to the information entropy sequence interference gradient data to obtain the native language system interference identification data of the content paragraphs;

[0065] Step S3: Obtain the underlying learning architecture of the AI tool; optimize the adaptive language system anomaly identification of the underlying learning architecture of the AI tool according to the native language system interference identification data of the content paragraphs to obtain the AI language system anomaly identification architecture;

[0066] Step S4: Evaluate the foreign language teaching lecture effect of the foreign language writing assessment content text according to the AI language system anomaly identification architecture to obtain the foreign language teaching lecture effect evaluation data, and send the foreign language teaching lecture effect evaluation data to the terminal.

[0067] In the embodiment of the present invention, referring to Figure 1 As described, it is a schematic diagram of the step process of an AI-based foreign language teaching lecture effect evaluation method of the present invention. In this example, the AI-based foreign language teaching lecture effect evaluation method includes the following steps:

[0068] Step S1: Obtain the text of the foreign language writing assessment content submitted by students; perform logical segmentation of the text paragraphs of the foreign language writing assessment content to obtain the paragraph logical segmentation data of the foreign language writing assessment content;

[0069] In the embodiments of the present invention, first, it is necessary to obtain the text of the foreign language writing assessment content submitted by students. This operation can be obtained through a network interface or file upload to ensure the integrity and correctness of the text data. Next, the text is segmented according to paragraph logic to ensure that the segmented text paragraphs can reflect the logical relationship of each paragraph in the article. Specifically, in implementation, first, a basic text cleaning operation is performed on the text submitted by students, including removing irrelevant characters, standardizing punctuation marks, and excluding garbled information. Then, an algorithm based on natural language processing is used, combined with syntactic analysis and semantic analysis methods to segment the text into paragraphs. The basis for paragraph segmentation is the logical structure. Techniques such as Dependency Parsing are used to identify the topic sentences and core semantics of each paragraph, thereby segmenting paragraphs with independent semantic units. The process of this paragraph segmentation needs to use a hierarchical grammar rule tree to define the boundaries of paragraphs based on the logical relevance between the main clause and subordinate clauses in each paragraph, ensuring that each paragraph can completely convey a set of ideas or arguments. The generation of paragraph segmentation data can be achieved through the construction of a text matrix. In this process, the sentence and paragraph structures of the text are converted into the form of a data matrix for subsequent processing.

[0070] Step S2: Quantify the information entropy sequence interference gradient for the foreign language writing assessment content text based on the paragraph logic segmentation data to obtain information entropy sequence interference gradient data; identify the native language system interference for the foreign language writing assessment content text according to the information entropy sequence interference gradient data to obtain content paragraph native language system interference identification data;

[0071] In the embodiments of the present invention, first, the information entropy sequence interference gradient is quantified for the foreign language writing assessment content text based on the paragraph logic segmentation data. For this purpose, the calculation method of information entropy is used to measure the amount of information in each paragraph of the text. Specifically, first, the word frequency of each paragraph is counted, the occurrence frequency of each word is calculated, and the information entropy is calculated using the following formula: , where represents the term 's occurrence probability in the paragraph, is the entropy value of the paragraph, reflecting the complexity and uncertainty of the paragraph, Indicates the number of words. Then, based on the calculated information entropy value, further analyze the interference gradient of the text. The main purpose of quantifying the interference gradient is to identify the change in information density in the text, and by calculating the change rate of the entropy value, obtain the interference gradient data of the entropy sequence. This interference gradient represents the degree of distortion of information transfer between paragraphs. Next, based on the interference gradient data of the information entropy sequence, conduct native language system interference recognition. By comparing the entropy gradient of the paragraph with the known native language system interference model, identify the existing native language interference. For example, if the entropy gradient of certain paragraphs matches the native language pattern highly, it indicates that there is native language interference in that paragraph. Through this process, finally obtain the native language system interference recognition data of the content paragraph. This data not only includes the position and type of interference of the paragraph, but also the intensity and degree of influence of the interference, and is output in the form of structured data.

[0072] Step S3: Obtain the underlying learning architecture of the AI tool; perform adaptive language system anomaly recognition optimization on the underlying learning architecture of the AI tool according to the native language system interference recognition data of the content paragraph to obtain the AI language system anomaly recognition architecture;

[0073] In the embodiment of the present invention, in step S3, first obtain the underlying learning architecture of the AI tool. Specifically, this architecture can be a neural network model based on deep learning or a feature extraction method based on traditional machine learning. Assuming that a neural network-based architecture is used, first load a pre-trained language model (such as BERT, GPT, etc.). Then, use the native language system interference recognition data obtained previously in step S2 to perform adaptive language system anomaly recognition optimization on the underlying learning architecture of the AI tool. To achieve this optimization, first fine-tune the model according to the interference recognition data, adjust the parameters of the model so that it can better identify the language system anomalies in the text. The fine-tuning process can use the gradient descent method, and the specific optimization process is as follows: , where are the learning weight parameters of the model, is the time step parameter, is the learning rate, is the loss function. In this embodiment, the design of the loss function should be able to effectively capture the characteristics of language system anomalies. During the fine-tuning process, mainly update the network weights through the backpropagation algorithm, so that the model's ability to recognize native language interference is enhanced. The optimized model can more accurately identify the language system anomalies in foreign language writing texts and finally output the optimized AI language system anomaly recognition architecture.

[0074] Step S4: Evaluate the foreign language teaching and lecturing effect on the foreign language writing assessment content text according to the AI language system anomaly recognition architecture to obtain the foreign language teaching and lecturing effect evaluation data, and send the foreign language teaching and lecturing effect evaluation data to the terminal.

[0075] In the embodiment of the present invention, based on the AI language system anomaly recognition architecture, the foreign language teaching effect of the foreign language writing assessment content text is evaluated. First, the optimized AI language system anomaly recognition architecture is used to analyze each paragraph to detect language system anomalies in the text. This process mainly evaluates the foreign language writing level of students by calculating the differences between the grammar, semantics and other features of the paragraph and the standard foreign language model. By analyzing factors such as grammar errors, inappropriate word usage, and sentence patterns in the text, the foreign language teaching effect evaluation data is finally obtained. The evaluation results will include information such as the type of language errors, error frequency, and severity of errors in each paragraph. These data will be converted into the final teaching effect evaluation score through a standardized scoring model. All calculations and scoring during the evaluation process are automatically completed by algorithms without manual intervention. Finally, the foreign language teaching effect evaluation data is sent to the terminal device through the network for teachers and students to view. The data transmission process uses encryption algorithms to ensure data security and the privacy and reliability of the evaluation results.

[0076] Step S1 includes the following steps:

[0077] Step S11: Obtain the foreign language writing assessment content text submitted by the student;

[0078] Step S12: Clean the foreign language writing assessment content text to obtain the cleaned foreign language writing assessment content text;

[0079] Step S13: Perform narrative analysis of the text structure of the cleaned foreign language writing assessment content text to obtain the narrative data of the text structure of the writing content;

[0080] Step S14: Logically segment the paragraphs of the cleaned foreign language writing assessment content text according to the narrative data of the text structure of the writing content to obtain the paragraph logical segmentation data of the foreign language writing assessment content.

[0081] In the embodiments of the present invention, first, the text of the foreign language writing assessment content submitted by students is obtained from the system. This text is usually in the form of an electronic document, such as standard formats like Word documents, PDF files, or TXT files. By reading the content of the file, it is converted into plain text format. At this time, a file parsing tool is used to extract the text content, ensuring format standardization and avoiding any potential document encoding issues. After the text is obtained, the character set of the text (such as UTF-8 encoding) is further confirmed to ensure its compatibility and avoid garbled problems during subsequent processing. After obtaining the text, it is stored in a text data structure and prepared to enter the subsequent processing flow. For the processing of different document formats, corresponding parsing libraries are used for adaptation, such as a PDF parsing library or a Word document parsing tool, to ensure that the obtained text data is clear and complete. In step S12, text cleaning of the foreign language writing assessment content text is performed. The goal of text cleaning is to remove irrelevant symbols, formatting errors, and noise data to ensure the accuracy of subsequent analysis. In the specific implementation process, first, all non-language characters in the text are removed, such as page numbers, headers, and footers, and special symbols and punctuation marks are removed using regular expressions. Then, redundant spaces and line breaks in the text are further removed to ensure that each line of text is valid language data. During the processing, part-of-speech tagging is performed on the words in the text through a syntax analysis library, and irrelevant stop words are removed, such as "de", "shi", "le", etc., which have no practical meaning for subsequent analysis. The cleaning process is implemented through text processing tools such as the nltk and re libraries in Python to ensure that only the parts of the text useful for semantic analysis are retained in the final output. In addition, during the cleaning process, it is also necessary to ensure the handling of text encoding issues to avoid garbled problems caused by inconsistent character sets. After cleaning, the cleaned text data is generated for subsequent analysis. In step S13, discourse structure narrative analysis is performed. The purpose of discourse structure analysis is to identify the overall organization of the text and the structural levels of each paragraph. Specifically, first, a rule-based discourse analysis algorithm is used to process the cleaned text. This algorithm divides different discourse units based on punctuation marks (such as full stops, commas, question marks, etc.) in the text and the grammatical structure of sentences. During the analysis, dependency syntactic analysis is used to identify the relationships between individual sentences in the text. For example, by parsing the basic grammatical components such as the subject, predicate, and object in a sentence, the connectivity of each sentence with the preceding and following sentences is further judged. During the discourse structure narrative analysis process, some text-based logical structure models are also combined, such as the common discourse structure of "introduction - body - conclusion", and the discourse structure in the text is inferred through pattern matching techniques. This process also relies on the topic words in the text to judge the core topic of each paragraph and its position in the discourse.The final output is a narrative data of the text structure, which records the logical levels, themes, and syntactic structures of each paragraph, as well as the relationship chain between paragraphs, providing a reference for subsequent logical segmentation of text paragraphs. In step S14, based on the narrative data of the text structure from the previous step, logical segmentation of text paragraphs is performed. In specific operations, first, using the aforementioned narrative data of the text structure, the entire text is decomposed into multiple paragraphs or text units. The basis for paragraph division is the conversion of themes in the text, changes in grammatical structures, and the logical relationships marked in the text structure. For example, if the text structure data indicates that a certain paragraph is a further elaboration or summary of the content of the previous paragraph, it is regarded as an independent paragraph. By analyzing the cohesive words (such as "therefore", "in addition", "for example", etc.) that appear in the text, the logical relationships between sentences are judged to further determine the boundaries of paragraphs. To ensure the accuracy of segmentation, deep learning algorithms such as long short-term memory network (LSTM) or recurrent neural network (RNN) can be used to assist in judging the logical relationships between sentences and paragraphs, and the paragraph boundaries are adjusted according to the prediction results of these algorithms. After segmentation, the starting position, ending position, and content of each paragraph will be saved, forming paragraph logical segmentation data. This data is usually saved in JSON or CSV format, recording the detailed information of each paragraph, including the paragraph content, the position of the paragraph in the text, the theme information of the paragraph, etc. Finally, this data provides structured text input for subsequent evaluation of the teaching effect of foreign language teaching.

[0082] Step S2 includes the following steps:

[0083] Step S21: Analyze the text of the foreign language writing assessment content for parallel / subordinate connections of sentence patterns according to the paragraph logical segmentation data, and obtain parallel / subordinate connection data of sentence patterns;

[0084] Step S22: Identify the logical jump links of the text of the foreign language writing assessment content based on the parallel / subordinate connection data of sentence patterns and the paragraph logical segmentation data, and obtain content paragraph logical jump data;

[0085] Step S23: Quantify the interference gradient of the information entropy sequence for the text of the foreign language writing assessment content based on the content paragraph logical jump data, and obtain information entropy sequence interference gradient data;

[0086] Step S24: Identify the interference of the mother tongue system for the text of the foreign language writing assessment content according to the information entropy sequence interference gradient data and the content paragraph logical jump data, and obtain content paragraph mother tongue system interference identification data.

[0087] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0088] Step S21: Analyze the parallel / subordinate connection of sentences in the foreign language writing assessment content text according to the paragraph logic, and obtain the parallel / subordinate connection data of sentences;

[0089] In the embodiment of the present invention, in step S21, the parallel / subordinate connection of sentences is analyzed to identify the sentence structures in the foreign language writing assessment content text, especially the usage of parallel and subordinate sentences. First, use a syntactic analysis tool (such as a tool based on dependency syntactic analysis) to parse the syntactic structure of each sentence in the text and construct a dependency relationship tree in the sentence. In the dependency relationship tree, the relationships between sentence components such as the subject, predicate, and object are marked. By matching the syntactic analysis results of each sentence, parallel structures and subordinate structures are identified. Parallel sentences are usually connected by conjunctions such as "and", "or", etc. to form an equivalent relationship between sentences; subordinate sentences are connected by subordinate conjunctions such as "because", "although", etc., and the subordinate clauses are attached to the main clause to form a hierarchical structure. For the parallel and subordinate relationships of each sentence, they are marked and recorded in the parallel / subordinate connection data of sentences. This data includes the connection type (parallel or subordinate) of each sentence, the position of the conjunction, the connected sentence components, and the grammatical hierarchical relationship between sentences. The tools used may include the dependency syntactic analysis module based on natural language processing libraries such as SpaCy to ensure the accuracy and grammatical compliance of the analysis results.

[0090] Step S22: Identify the logical jump links in the foreign language writing assessment content text based on the parallel / subordinate connection data of sentences and the paragraph logic segmentation data, and obtain the content paragraph logical jump data;

[0091] In the embodiment of the present invention, in step S22, according to the clause parallel / subordinate connection data and paragraph logical segmentation data obtained in step S21, logical jump links in the foreign language writing assessment content text are identified. First, by analyzing the paragraphs of the text, it is judged whether there are logical jumps between sentences or paragraphs based on the content logical relationship between paragraphs. Logical jumps usually manifest as sudden changes in the discussion in the text, or the lack of obvious connecting words or sentence connection methods between paragraphs. Using the logical link identification algorithm, combined with the clause parallel / subordinate connection data, analyze whether the transition between sentences or paragraphs between paragraphs is smooth, and check whether there are obvious logical gaps or irrelevant jumps. The algorithms used in this process include rule-based text relationship reasoning, and judge the rationality of the logical link by combining the connecting words in the sentence and the semantic relationship of the context. For each paragraph, by matching the topic words and emotional colors in the sentence, judge whether there is a jump between this paragraph and the previous and subsequent paragraphs, and then determine whether there is a logical jump link. The identified logical jump links will be recorded as content paragraph logical jump data, and the data includes information such as the location where the jump occurs, the paragraphs involved, and the type of jump. This data provides a basis for subsequent interference gradient quantization.

[0092] Step S23: Quantize the interference gradient of the information entropy sequence for the foreign language writing assessment content text based on the content paragraph logical jump data to obtain information entropy sequence interference gradient data;

[0093] In the embodiment of the present invention, in step S23, according to the content paragraph logical jump data obtained in step S22, the interference gradient quantization of the information entropy sequence is performed. Information entropy is a measure of the uncertainty or complexity of information. In this step, information entropy is used to perform quantitative analysis on the logical jumps in the text. First, windowing processing is performed on the foreign language writing assessment content text, and it is defined that each window contains several sentences or paragraphs. Then, the information entropy of each window is calculated according to the sentence structure, logical relationship and semantic information within the window. The calculation formula of information entropy is as follows: , where represents the term the occurrence probability in the paragraph, is the entropy value of the paragraph, represents the number of words. By analyzing the distribution of each unit within the window, the information entropy of the window is calculated. If the information entropy is high, it indicates that there is a large information fluctuation within the sentences or paragraphs of this window, which is caused by logical jumps. Then, the information entropy of each window is used as the basis for interference gradient quantization. By analyzing the change situation within each paragraph, the information entropy sequence of the entire text is calculated, and information entropy sequence interference gradient data is generated. This data includes the entropy value of each paragraph, the corresponding jump type, and the entropy change situation between paragraphs. This data can be used to quantify the degree of logical jump interference in the text and provide a basis for subsequent native language system interference identification.

[0094] Step S24: Identify the mother tongue system interference in the foreign language writing assessment content text based on the information entropy sequence interference gradient data and the logical jump data of the content paragraphs, and obtain the mother tongue system interference identification data of the content paragraphs.

[0095] In the embodiments of the present invention, the mother tongue system interference is identified based on the information entropy sequence interference gradient data and the logical jump data of the content paragraphs. The goal of mother tongue system interference identification is to detect whether there are phenomena of logical incoherence or improper language structure caused by mother tongue interference in foreign language writing. First, by analyzing the logical jump data of the content paragraphs and combining the information entropy sequence interference gradient data, it is judged whether the grammatical structure and logical jumps in the paragraphs conform to the standard writing norms of the target foreign language. By comparing the grammatical and syntactic differences between the mother tongue system and the foreign language system, the abnormal language structures caused by mother tongue interference are identified. For example, the mother tongue tends to use simple sentences in sentence patterns, while the target foreign language prefers to use complex sentences. During the identification process, a rule-based interference model is used to detect whether there are features affected by the mother tongue in the foreign language writing. This model judges which paragraphs or sentences are affected by the mother tongue by checking the sentence patterns, grammatical markers, and the use of conjunctions in the text, combined with the fluctuations of the information entropy values. Finally, the identified interference information is recorded as the mother tongue system interference identification data of the content paragraphs, including information such as the interference type, interference location, and degree of interference. These data will provide detailed feedback for the subsequent evaluation of the teaching effect of foreign language teaching, helping to identify the language structure problems existing in students' writing.

[0096] Step S23 includes the following steps:

[0097] Step S231: Statistically count the frequencies of logical jump conjunctions in the foreign language writing assessment content text based on the logical jump data of the content paragraphs, and obtain the logical jump conjunction frequency data;

[0098] Step S232: Decompose the semantic span word vectors of the conjunctions through the logical jump conjunction frequency data to obtain the semantic span word vectors of the conjunctions;

[0099] Step S233: Perform a regression analysis on the grammatical ambiguity of the front and back paragraphs of the foreign language writing assessment content text based on the semantic span word vectors of the conjunctions and the logical jump conjunction frequency data, and obtain the grammatical ambiguity regression data of the front and back paragraphs;

[0100] Step S234: Perform a simulation analysis on the disorder state of the information entropy sequence dimension constraint based on the semantic span word vectors of the conjunctions and the grammatical ambiguity regression data of the front and back paragraphs, and obtain the disorder state of the information entropy sequence dimension constraint;

[0101] Step S235: Perform information entropy sequence interference gradient quantization on the disorder state of the information entropy sequence dimension constraints to obtain information entropy sequence interference gradient data.

[0102] In the embodiment of the present invention, in step S231, the frequency of logical jump conjunctions is counted to analyze the usage of conjunctions in the text of foreign language writing assessment content, especially those conjunctions that cause logical jumps between paragraphs. First, through syntactic analysis and natural language processing tools (such as SpaCy or NLTK), lexical and syntactic analysis is performed on the text of foreign language writing assessment content to extract all conjunctions. Conjunctions include but are not limited to "however", "therefore", "because", etc. These words usually appear between paragraphs, indicating a change in logical relationship. Then, by writing a statistical algorithm, the frequency of each conjunction appearing in the text is calculated, and its appearance position is recorded. For each paragraph, the total number of occurrences of the same conjunction and its relative position in the text are counted, so as to obtain the frequency data of logical jump conjunctions. These data not only record the frequency of each conjunction, but also include their positions in the text, helping to analyze the possibility and intensity of logical jumps in the text. In step S232, semantic span word vector decomposition of conjunctions is performed based on the frequency data of logical jump conjunctions obtained in step S231. First, a word vectorization model (such as Word2Vec or GloVe) is used to learn the semantic representation of each conjunction. The word vectorization model maps each conjunction to a multi-dimensional vector, and each dimension of the vector represents the semantic features of the word. By analyzing the context of each conjunction in the text, word vectors are generated, and a vector representation is assigned to each conjunction. These word vectors can capture the semantic span of conjunctions, that is, the semantic extensibility of conjunctions in sentences or paragraphs. Then, based on the frequency data, the conjunctions are decomposed to generate the semantic span of word vectors. This process requires dimensionality reduction of word vectors through methods such as principal component analysis (PCA) to obtain a more concise and expressive semantic span word vector. Through this process, the semantic influence range of conjunctions in the text can be quantified, providing key data support for subsequent regression analysis. According to the semantic span word vectors of conjunctions and the frequency data of logical jump conjunctions obtained in step S232, regression analysis of the grammatical ambiguity between the front and back paragraphs is performed. Grammatical ambiguity refers to the situation where it is difficult to understand due to unclear grammar or loose structure in the text. In this step, first, grammatical analysis is performed on each paragraph in the text, and a tool based on dependency syntactic analysis (such as SpaCy or Stanford Parser) is used to identify the syntactic structure of the paragraph. Then, according to the syntactic structures of the front and back paragraphs, combined with the semantic span and frequency data of logical jump conjunctions, a regression analysis model (such as linear regression or polynomial regression) is applied to calculate the grammatical ambiguity between paragraphs. Specifically, for each pair of adjacent paragraphs, evaluate whether the logical connection between them is smooth, and whether there are abrupt grammatical transitions or structural differences.According to the results of regression analysis, the grammatical fuzzy regression data of the previous and subsequent paragraphs are obtained. The data contains the grammatical fuzzy degree values between each pair of paragraphs, which represents the grammatical coherence and the clarity of the grammatical structure between the two paragraphs. Based on the semantic span word vectors of connecting words and the grammatical fuzzy regression data of the previous and subsequent paragraphs, a simulation analysis of the disorder state of the information entropy sequence dimension constraint is carried out. Information entropy is a standard for measuring the disorder and uncertainty of the amount of information in a text. In this step, first, combined with the semantic span word vectors obtained in the previous step, an information entropy sequence of the text is constructed. This sequence represents the semantic uncertainty within the paragraph by calculating the information entropy value of each paragraph. By defining a sliding window, the grammatical fuzzy degree and information entropy of the paragraph are calculated within each window, and its dimension is further analyzed. The dimension constraint represents the fluctuation range of the information entropy in time or space. Using this constraint, the disorder state in the text is simulated, that is, whether there are sudden logical jumps or grammatical inconsistencies between paragraphs. For this purpose, the information entropy formula: is used. , where represents the term 's occurrence probability in the paragraph, is the entropy value of the paragraph, represents the number of words, describing the semantic uncertainty within the paragraph. By calculating the entropy value sequence of the entire text, it is identified which parts of the text show a strong disorder state, indicating that there are large logical jumps or unclear grammar between paragraphs. Finally, through model simulation, the disorder state of the information entropy sequence dimension constraint is obtained, and the entropy value and disorder degree of each paragraph are recorded, providing data support for the subsequent steps. The disorder state of the information entropy sequence dimension constraint is quantified by the interference gradient of the information entropy sequence to obtain the interference gradient data of the information entropy sequence. The purpose of this step is to measure the degree of interference of the logical jumps and grammatical fuzzy degrees in the text on the text structure by quantifying the changes in the information entropy sequence. First, using the disorder state obtained in the previous stage, by analyzing the fluctuation of the information entropy sequence, the interference gradient of each paragraph is calculated. The interference gradient can be calculated by the following formula: , where is the interference gradient at time point , and represent the information entropy values at the previous and subsequent moments respectively. By calculating the change in the information entropy at each moment, the interference gradient between each paragraph is obtained, indicating the degree of influence caused by logical jumps or grammatical fuzzy degrees between paragraphs. Finally, all the interference gradient data are integrated into the interference gradient data of the information entropy sequence, providing a quantitative basis for the subsequent identification of the interference of the mother tongue system and the evaluation of the teaching effect of foreign language teaching. These interference gradient data can reveal which parts of the writing are strongly affected by logic or grammar.

[0103] Step S234 includes the following steps:

[0104] Perform polysemy clustering analysis on the semantic span word vectors of conjunctions to obtain polysemy clustering data of conjunctions;

[0105] Perform a grammatical boundary fuzziness trend analysis on the grammatical fuzziness regression data of the previous and subsequent paragraphs to obtain the grammatical boundary fuzziness trend data of the previous and subsequent paragraphs;

[0106] Based on the polysemy clustering data of conjunctions and the grammatical boundary fuzziness trend data of the previous and subsequent paragraphs, perform context change boundary fitting to obtain the context change boundary fitting data before and after the paragraph;

[0107] Based on the Markov chain, perform a context boundary fitting state simulation on the context change boundary fitting data before and after the paragraph to obtain the context boundary state fitting data;

[0108] According to the context boundary state fitting data, perform a simulation analysis of the disorder state of the information entropy sequence dimension constraint to obtain the disorder state of the information entropy sequence dimension constraint.

[0109] In the embodiment of the present invention, in step S234, polysemy clustering analysis of the semantic span word vectors of the connecting words is performed on the premise that the semantic span word vectors of the connecting words have been obtained through step S232. First, the K-means clustering algorithm is used to perform clustering analysis on the semantic span word vectors of the connecting words. Specifically, the Euclidean distance is used to measure the similarity between word vectors, and classification is performed according to the polysemy of the connecting words in the context. Through multiple iterations, the K-means algorithm divides the connecting words into multiple categories, and each category represents a different semantic representation of the connecting words. Then, a label is assigned to each category as the "polysemy clustering data of the connecting words". For example, if the semantic manifestations of the word "because" in different paragraphs are causal relationships, explanations, etc., then this word will be divided into multiple categories in the polysemy clustering. The clustering result generates a polysemy category data for each connecting word, and this data can be used for subsequent context change analysis. Next, a grammatical boundary fuzziness trend analysis is performed on the grammatical fuzziness regression data of the front and back paragraphs. This analysis aims to reveal whether the grammatical connection between paragraphs in the text gradually becomes fuzzy. First, based on the grammatical fuzziness regression data of the front and back paragraphs, a time series analysis method is used to model the change trend of the grammatical fuzziness. The weighted sliding window is used to calculate the change in fuzziness of each paragraph, with particular attention paid to the boundary changes between paragraphs. By calculating the gradient of the paragraph grammatical fuzziness, the change trend of the grammatical fuzziness boundary between paragraphs, that is, the "grammatical boundary fuzziness trend data of the front and back paragraphs", is obtained. This data shows how the connection between paragraphs gradually becomes fuzzy or clear as the text progresses. Based on the above polysemy clustering data of the connecting words and the grammatical boundary fuzziness trend data of the front and back paragraphs, a context change boundary fitting analysis is performed. This analysis identifies potential context changes in the text by fitting the change in the grammatical fuzziness between paragraphs. First, using the curve fitting technique based on the least squares method, the polysemy clustering data of the connecting words is combined with the grammatical boundary fuzziness trend data of the paragraphs to construct a model of the context change boundary. By optimizing the parameters in the model, the best match between the change in the paragraph grammatical fuzziness and the semantic span of the connecting words is found, thereby obtaining the fitting boundary data of the context change before and after the paragraphs. The goal of the fitting process is to minimize the fitting error to obtain the most accurate context change boundary model. This data provides effective boundary data for subsequent context boundary state simulation. Based on the context boundary fitting data, a Markov chain is used to perform state fitting simulation on the context change boundary before and after the paragraphs. First, the state space of the Markov chain is defined as all possible states of the context change between text paragraphs. Each state represents a specific context change boundary. The transition matrix of the Markov chain represents the probability relationship of the context change between paragraphs. By modeling the probability of the context change before and after the paragraphs, the impact of the context change of a certain paragraph in the text on the subsequent paragraphs can be predicted.Fit the data using the known context change boundaries, and perform simulation calculations using the state transition algorithm of the Markov chain to obtain the context boundary state fitting data in the text. This step further improves the prediction accuracy of context changes by simulating the state transition process of context changes. Finally, based on the obtained context boundary state fitting data, perform a simulation analysis of the disorder state of the information entropy sequence dimension. This analysis aims to reveal the degree of disorder of the text structure by quantifying the information entropy changes in the text. First, by calculating the context boundary changes between paragraphs, use the formula of information entropy to measure the information uncertainty in the text. The calculation formula of information entropy is as follows: , where represents the term in the paragraph, is the entropy value of the paragraph, represents the number of words. By combining the information entropy with the context boundary state fitting data, use multiple regression analysis or information gain algorithm to further constrain the dimension of the entropy value. This step will calculate a disorder degree value for each paragraph or context state, and finally obtain the disorder state data of the information entropy sequence dimension constraint. This data can reveal the grammar ambiguity and information entropy fluctuation caused by factors such as context changes and paragraph logic jumps in the text, providing an important basis for the subsequent evaluation of the teaching effect of foreign language teaching.

[0110] Step S24 includes the following steps:

[0111] Step S241: Obtain the native language system structure data of different students;

[0112] Step S242: Analyze the native language transfer characteristics of the foreign language writing assessment content text based on the content paragraph logic jump data and the native language system structure data to obtain the native language transfer characteristic data of the assessment content;

[0113] Step S243: Identify the syntactic structure logic deviation based on the native language transfer characteristic data of the assessment content and the content paragraph logic jump data to obtain the native language syntactic structure logic deviation data;

[0114] Step S244: Identify the semantic deviation of the native language transfer characteristic data of the assessment content to obtain the native language semantic characteristic deviation data;

[0115] Step S245: Map the foreign language writing assessment content text based on the native language syntactic structure logic deviation data and the native language semantic characteristic deviation data to obtain the native language dominant syntactic framework mapping data;

[0116] Step S246: Identify the native language system interference based on the information entropy sequence interference gradient data and the native language dominant syntactic framework mapping data to obtain the content paragraph native language system interference identification data.

[0117] In the embodiments of the present invention, first, it is necessary to obtain the native language system structure data of different students. This data is usually obtained by collecting the language structure information used by students in their native language, including but not limited to vocabulary, syntactic structure, word order, grammar rules, etc. Through large-scale linguistic datasets, such as the United Nations Language Dataset or various known native language annotated corpora, the core syntactic structure features of each student's native language are extracted. Specifically, when implementing, a dependency syntactic analysis algorithm is used to perform syntactic analysis on the native language text provided by each student. Analysis tools such as Stanford Parser can be used to parse the hierarchical structure of sentence components, extract various syntactic components (such as subject, predicate, object, modifying components, etc.), and record the position and frequency of each component in the sentence. By extracting and encoding these syntactic elements, the structure data of each student's native language system, such as different sentence patterns, common grammar rules, etc., is obtained. This data provides a basis for subsequent analysis of native language transfer features. In step S242, based on the content paragraph logical jump data and the native language system structure data, an analysis of the native language transfer features of the foreign language writing assessment content is carried out. First, the paragraph logical jump data obtained previously is combined with the native language system data of each student to analyze whether the grammar structure in foreign language writing is affected by the native language structure. Specifically, when implementing, a pattern matching algorithm is used to identify the similarity between the grammar structure in the foreign language writing text and the structure in the native language system. For example, when the subject-verb-object structure of a foreign language sentence is similar to the common structure of a native language sentence, it can be considered as one of the features of native language transfer. By comparing the differences between the sentence structures in each student's foreign language writing text and the common patterns of their native language system, the potential influence of the native language on foreign language writing is identified, and "native language transfer feature data" is generated. These data identify which grammar structures are affected by the native language, further revealing the native language transfer phenomenon in students' writing. Step S243 involves identifying the syntactic structure logical deviation of the foreign language writing assessment content based on the native language transfer feature data and the paragraph logical jump data. In this step, first, it is necessary to analyze the grammar deviation in foreign language writing, that is, the syntactic structure errors or logical inconsistencies that occur in students' foreign language writing. By comparing the native language transfer feature data with the syntactic structure in foreign language writing, syntactic analysis tools (such as NLTK, spaCy, etc.) can be used to perform structured parsing on the sentences in the writing and detect the syntactic deviations. For the identification of deviations, a method of matching and comparing syntactic trees is adopted. For example, by calculating the similarity of syntactic trees to determine whether there are syntactic structures that do not conform to foreign language rules. In addition, by introducing a logical deviation analysis method (such as logistic regression or decision tree), the logical jump deviation between sentences is identified according to the logical relationship between paragraphs. These steps finally obtain "mother tongue syntactic structure logical deviation data", which indicates the syntactic logic inconsistency in foreign language writing due to the influence of the native language. In step S244, semantic deviation identification is carried out.The goal of semantic deviation identification is to analyze the expression errors caused by mother tongue transfer in foreign language writing, especially the meaning communication deviation that occurs when students use their mother tongue thinking framework to construct foreign language sentences. To this end, semantic analysis technology (such as a word vector model based on Word2Vec or GloVe) is used to analyze foreign language writing at the semantic level. By comparing the semantic associations between words in the student writing text and their mother tongue, the parts with miscommunication of meaning are identified. For example, when the semantic span of some words in the mother tongue and the foreign language is very different, semantic deviation is caused. The degree of semantic deviation of each word is calculated using an algorithm based on semantic similarity (such as cosine similarity), and the degree of semantic deviation of each paragraph is quantified by regression analysis. These calculations result in "mother tongue semantic feature deviation data", which reflects the meaning communication problems caused by the influence of the mother tongue in foreign language writing. In step S245, the mother tongue dominant syntactic framework mapping is performed in combination with the mother tongue syntactic structure logic deviation data and the mother tongue semantic feature deviation data. This step creates a mapping model by comprehensively analyzing the syntactic structure and semantic deviation in the foreign language text, and matches the syntactic structure in the foreign language text with the syntactic framework of the native language. The specific method adopted is to fit the relationship between the native syntactic framework and the syntactic structure in the foreign language writing based on a mapping algorithm (such as linear regression, polynomial regression, etc.). Through this step, it is possible to identify which parts of the foreign language text are completely consistent with the syntactic structure of the native language and which parts deviate from the logic of the native language. Finally, the "native language dominant syntactic framework mapping data" is obtained to show how the foreign language sentences in the students' writing are affected by the native language structure, and to represent the adaptation of the foreign language sentences to the native language syntactic structure in a graphical or numerical way. In step S246, native language system interference identification is performed according to the information entropy sequence interference gradient data and the native language dominant syntactic framework mapping data. This step combines the information entropy sequence interference gradient data with the native language dominant syntactic framework mapping data to further analyze how the native language system affects the overall structure of foreign language writing. First, by using information entropy to quantify the degree of interference in the text, the interference gradient data obtained previously was used to conduct interference analysis on foreign language writing, revealing the interference areas of the native language in foreign language writing. Then, combined with the native language dominant syntactic framework mapping data, clustering analysis methods (such as K-means clustering) were used to cluster foreign language writing texts to identify the sentence areas most significantly affected by the native language. These analyses ultimately generated "content paragraph native language system interference identification data" to identify potential error areas in foreign language writing caused by native language interference and provide a basis for subsequent teaching evaluation.

[0118] Step S3 includes the following steps:

[0119] Step S31: Obtain the underlying learning architecture of the AI ​​tool;

[0120] Step S32: Normalize the information entropy sequence interference gradient data to obtain the normalized information entropy sequence interference gradient data;

[0121] Step S33: Identify and enhance the disorder degree of the AI architecture for the underlying learning architecture of the AI tool based on the normalized information entropy sequence interference gradient data to obtain the AI disorder degree identification and learning enhanced architecture;

[0122] Step S34: Perform adaptive language system anomaly identification and optimization on the AI disorder degree identification and learning enhanced architecture based on the content paragraph mother tongue system interference identification data to obtain the AI language system anomaly identification architecture.

[0123] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0124] Step S31: Obtain the underlying learning architecture of the AI tool;

[0125] In the embodiment of the present invention, first, the underlying learning architecture of the AI tool needs to be obtained. The underlying learning architecture of the AI tool usually consists of a neural network, a deep learning framework, an optimization algorithm, etc. In a specific implementation, the underlying learning architecture uses a deep neural network (DNN) or a convolutional neural network (CNN) as the basic structure to process the input data and perform learning and prediction. The architecture includes multiple neural network layers, each layer contains a different number of neurons, and performs non-linear transformation through activation functions (such as ReLU, Sigmoid). According to different tasks, the learning architecture also includes a recurrent neural network (RNN) for processing time series data, or a long short-term memory network (LSTM) for processing long-term dependencies. The design of the underlying learning architecture needs to be optimized according to specific tasks, and usually selects hyperparameters of the training model, such as the learning rate, batch size, optimizer type (such as Adam or SGD), etc., to ensure that the model can be effectively trained on the given data set.

[0126] Step S32: Normalize the information entropy sequence interference gradient data to obtain the normalized information entropy sequence interference gradient data;

[0127] In the embodiment of the present invention, the information entropy sequence interference gradient data is normalized to obtain the normalized information entropy sequence interference gradient data. The information entropy sequence interference gradient data represents the uncertainty and interference degree in the sequence, and the purpose of normalization is to scale the data to a standard range to eliminate the influence of different data ranges on subsequent analysis. Specifically, a standardization method (such as Z-score standardization or Min-Max normalization) is used to process the data.

[0128] Step S33: Based on the information entropy sequence interference gradient normalized data, perform AI architecture disorder degree recognition learning reinforcement on the underlying learning architecture of the AI tool to obtain an AI disorder degree recognition learning reinforcement architecture;

[0129] In the embodiment of the present invention, based on the information entropy sequence interference gradient normalized data, perform AI architecture disorder degree recognition learning reinforcement on the underlying learning architecture of the AI tool. The core objective of this step is to use the normalized interference gradient data to strengthen the disorder degree recognition ability of the AI architecture. Specifically, first, the parameters in the AI model (such as the weight matrix) need to be adjusted so that it can better recognize the disorder patterns in the data. The network is trained by using the backpropagation algorithm to adjust the weights in the network to minimize the loss function of disorder degree recognition. The loss function can be defined based on information entropy or other uncertainty metrics. Through the backpropagation optimization algorithm (such as the gradient descent method), the weights of the network are adjusted until the network can accurately recognize the disorder patterns in the data, thereby obtaining the strengthened AI architecture. In this process, the AI model not only trains the prediction target but also strengthens the ability to recognize disorder patterns, enabling it to maintain efficient learning ability when facing complex input data.

[0130] Step S34: Based on the content paragraph mother tongue system interference recognition data, perform adaptive language system anomaly recognition optimization on the AI disorder degree recognition learning reinforcement architecture to obtain an AI language system anomaly recognition architecture.

[0131] In the embodiment of the present invention, based on the content paragraph mother tongue system interference recognition data, perform adaptive language system anomaly recognition optimization on the AI disorder degree recognition learning reinforcement architecture. The objective of this step is to further optimize the architecture of the AI model through the mother tongue interference recognition data so that it can more effectively recognize and process anomalies in different language systems. First, the mother tongue system interference recognition data is used as additional input information, and the AI model is optimized through an adaptive learning algorithm (such as adaptive gradient descent). In this process, the mother tongue interference data is regarded as a kind of noise or interference, and the AI architecture needs to learn how to extract effective information from these interferences. The specific operation method is to dynamically adjust the parameters in the learning process by adjusting the learning rate and the optimization algorithm, so that the AI architecture can better adapt to the anomaly patterns brought by different language systems. The form of the loss function adopted by the optimization algorithm is: , where is the original loss function, which is the loss calculated by the AI model based on the input data and is usually related to the difference between the predicted result and the true label (for example, cross-entropy loss or mean square error), is the regularization coefficient, which is used to balance the influence of the original loss function and the mother tongue interference term, Denotes the sum of native language interference recognition errors, where: Is the true value of the native language interference data, the th native language interference marker of the data point (which can be the actual interference pattern, representing the influence of the native language on foreign language writing), Is the predicted value of the AI model for the th data point, that is, the native language interference predicted by the model, Is the number of data points, that is, how many native language interference data are used to train the model. Finally, based on this optimization process, the "AI language system anomaly recognition architecture" obtained can more accurately identify the abnormal language patterns caused by native language interference, thus effectively optimizing the recognition ability of AI tools in processing foreign language writing tasks.

[0132] Step S33 includes the following steps:

[0133] Step S331: Perform topological information entropy discretization processing on the information entropy sequence interference gradient normalized data to obtain the information entropy topological discrete feature space;

[0134] Step S332: Perform feature orthogonal mapping and information entropy reconstruction processing on the information entropy topological discrete feature space to obtain the feature mapping information entropy tensor;

[0135] Step S333: Perform interference gradient calibration processing according to the feature mapping information entropy tensor to obtain the interference gradient calibration data;

[0136] Step S334: Based on the interference gradient calibration data, perform AI architecture disorder degree recognition learning reinforcement on the underlying learning architecture of the AI tool to obtain the AI disorder degree recognition learning reinforcement architecture.

[0137] In the embodiments of the present invention, first, the input information entropy sequence interference gradient normalized data is subjected to topological information entropy discretization processing. Specifically, the information entropy sequence interference gradient normalized data is represented as an ordered data set, which reflects the gradient change of the data and its distribution state under the information entropy metric. The topological information entropy discretization process discretizes the topological structure of these data and divides it into a finite number of discrete intervals in the continuous space. This process uses the Voronoi diagram for discretization, dividing the data space into multiple non-overlapping regions, representing the characteristics of the information entropy in each region. Through this discretization process, the originally continuous interference gradient data can be mapped to a discrete feature space, which can better represent the implicit topological relationships in the data. The discretized data is the information entropy topological discrete feature space, and the coordinates of each point correspond to the eigenvalue in a specific region, facilitating subsequent processing and analysis of the information entropy features. Using the feature orthogonal mapping method, the information entropy topological discrete feature space obtained in step S331 is subjected to mapping processing. First, through the principal component analysis (PCA) algorithm for feature orthogonal mapping, the relevant features in the feature space are orthogonalized to ensure the independence of the features and avoid the problem of multicollinearity. In this process, by constructing the covariance matrix and calculating the eigenvalues and eigenvectors, the mapped feature vector space is obtained. Then, information entropy reconstruction is performed, that is, the information entropy data is reconstructed by methods such as Fourier transform or wavelet transform to retain its original spectral characteristics. This processing converts the information entropy features in the feature space into a high-dimensional tensor structure, enabling each feature dimension to express more information and being more suitable for complex multi-dimensional data modeling. The finally obtained feature mapping information entropy tensor not only retains the global characteristics of the information entropy but also introduces the orthogonal relationship between each feature, providing a more accurate representation for subsequent data calibration and learning. Based on the feature mapping information entropy tensor, interference gradient calibration processing is performed. The specific operation is to first calculate the gradient of the mapped tensor through the gradient descent method to detect the impact of the interference factors in the current feature space on the model training. This process uses the backpropagation algorithm to calculate the error in the model and update the gradient. At this time, the output of the model will be dynamically adjusted according to the information entropy features in each dimension, gradually reducing the unnecessary interference effects. By gradually reducing the interference, the calibration of the interference gradient is finally achieved. During the calibration process, the Adam optimization algorithm is used to automatically adjust the learning rate to accelerate the convergence speed and ensure that the interference gradient is effectively corrected. The calibrated data is the interference gradient calibration data. The core of this step is to ensure that the model responds more precisely in each feature dimension through the adaptive optimization algorithm, reduce the influence of interference signals, and enable the model to better learn the target task. Based on the interference gradient calibration data obtained in step S333, the underlying learning architecture of the AI tool is strengthened in terms of learning. Specifically, the method of reinforcement learning is adopted, and the learning process of the AI model is gradually improved through environmental interaction.At this time, the AI model learns how to maintain a high learning efficiency and accuracy in the face of interfering data by identifying the degree of disorder (i.e., data inconsistency and abnormal patterns) in the training data and adjusting its learning strategy. In this process, the Q-learning algorithm is used to identify the degree of disorder, and the model continuously optimizes its parameters by evaluating the reward value of each operation. Through reinforcement learning, the AI architecture can gradually adapt to different degrees of disordered data and can self-adjust in a new data environment to achieve stronger learning adaptability. The finally obtained AI disorder degree recognition learning reinforcement architecture can better cope with various data perturbations and improve the stability and accuracy of the model in complex environments.

[0138] Step S34 includes the following steps:

[0139] Step S341: Conduct a mother tongue part-of-speech interference abnormal logic analysis on the mother tongue system interference recognition data of the content paragraph to obtain mother tongue part-of-speech interference abnormal logic data;

[0140] Step S342: Based on the mother tongue part-of-speech interference abnormal logic data, conduct a part-of-speech logic metaphor deviation induction to obtain part-of-speech logic metaphor deviation data;

[0141] Step S343: Conduct a concept cross-domain metaphor abnormal analysis on the part-of-speech logic metaphor deviation data to obtain concept cross-domain metaphor abnormal data;

[0142] Step S344: Based on the part-of-speech logic metaphor deviation data and the concept cross-domain metaphor abnormal data, conduct an adaptive language system abnormal recognition optimization on the AI disorder degree recognition learning reinforcement architecture to obtain an AI language system abnormal recognition architecture.

[0143] In the embodiments of the present invention, the interference recognition data of the content paragraph in the mother tongue system is processed. Through the abnormal logic analysis of the mother tongue part-of-speech interference, the situations of non-standard or inconsistent part-of-speech usage caused by mother tongue interference in foreign language writing are identified. Specifically, the mother tongue system interference data includes the parts affected by the mother tongue in foreign language writing, especially at the levels of vocabulary and grammatical structure. Through syntactic analysis and dependency syntactic analysis, combined with the part-of-speech tagging in linguistics, logical analysis is performed on these data. The conditional random field (CRF) algorithm is used to refine the part-of-speech tagging to find out those part-of-speech conversions or mislabelings that do not conform to the target foreign language rules. Then, combined with the language model, by calculating the probability of each part-of-speech conversion, the abnormal logical patterns are identified. These abnormal patterns usually manifest as the misuse of parts of speech under the influence of the mother tongue in certain contexts. Through this analysis, the abnormal logical data of the mother tongue part-of-speech interference is obtained, which represents the rules and abnormal manifestations of the part-of-speech interference affected by the mother tongue in foreign language writing, providing key data for subsequent deviation induction and anomaly optimization. Based on the abnormal logical data of the mother tongue part-of-speech interference obtained in step S341, the deviation induction of the part-of-speech logical metaphor is carried out. The purpose of the deviation induction of the part-of-speech logical metaphor is to identify the implicit deviation of the part-of-speech usage in the target foreign language caused by the transformation of the mother tongue structure or expression. Through the metaphor analysis method, first, the metaphorical expressions in the text are identified, especially those metaphorical errors where the change of part of speech leads to the conversion of meaning. For example, some words have a certain grammatical function in the mother tongue but do not have the same function in the foreign language, or their part-of-speech conversion leads to a significant deviation in the meaning of the sentence. Then, an inference algorithm is used to conduct inductive analysis on these metaphorical deviations, classify them into specific types of logical errors, and further refine the analysis according to the types of part-of-speech transformations. The decision tree algorithm is used to classify the induced metaphorical deviations to obtain the specific part-of-speech logical metaphor deviation data, which provides a quantitative analysis for the metaphorical conversion errors of different parts of speech and can accurately show the potential error types and their occurrence conditions under the influence of the mother tongue. Based on the part-of-speech logical metaphor deviation data obtained in step S342, the cross-domain metaphor anomaly analysis of concepts is further carried out. The cross-domain metaphor of concepts refers to the situation in foreign language writing where, due to the influence of the mother tongue system, the usage of parts of speech or concepts goes beyond their original context or domain, resulting in abnormal or illogical language expressions. For example, during translation, some concepts in the mother tongue are directly applied to the foreign language, and this cross-domain metaphor will lead to semantic incoordination. By expanding the metaphor recognition and combining the semantic network analysis method, based on dictionary resources such as WordNet, the reasonable meanings of words in different contexts are determined, and the cross-domain concept metaphors that do not conform to the foreign language usage habits are identified. The hierarchical clustering analysis (Hierarchical Clustering) is used to classify the metaphor anomalies according to their cross-domain degrees, especially paying attention to the part-of-speech usage problems caused by concept migration.The final concept cross-domain metaphor anomaly data can accurately capture the cross-domain metaphor deviations and abnormal patterns that appear in foreign languages ​​due to native language interference, and demonstrate the cross-domain metaphor deviations and abnormal patterns that appear in different domains at the language level and the implicit impact they produce. Based on the part-of-speech logic metaphor deviation data and concept cross-domain metaphor abnormality data obtained in steps S342 and S343, the AI ​​disorder degree recognition learning reinforcement architecture is adaptively optimized for language system abnormality recognition. At this time, the AI ​​architecture needs to be adaptively optimized for various abnormal patterns that appear in the foreign language learning process. First, these two types of data are used as training inputs, and data fusion processing is performed through the convolutional neural network (CNN) in deep learning. Combined with the long short-term memory network (LSTM) model, through time series modeling, the dynamic changes of metaphor deviations and cross-domain metaphor anomalies are captured, and the weight distribution in the learning process is optimized. Then, through the Q-learning method in reinforcement learning, the AI ​​architecture gradually improves the ability to recognize language system anomalies based on the reward and punishment mechanism for misidentification. In this way, the AI ​​architecture can automatically adjust its parameters during the training process and enhance the ability to recognize implicit anomalies in the language system. Ultimately, the resulting AI language system anomaly recognition architecture can adjust the model’s learning path in real time, more efficiently identify abnormal patterns in the language, and improve the accuracy of foreign language writing assessment.

[0144] The present invention also provides an AI-based foreign language teaching effect evaluation system, which is used to execute the AI-based foreign language teaching effect evaluation method as described above. The AI-based foreign language teaching effect evaluation system includes:

[0145] The paragraph logic segmentation module is used to obtain the foreign language writing assessment content text submitted by the students; perform paragraph logic segmentation on the foreign language writing assessment content text to obtain paragraph logic segmentation data of the foreign language writing assessment content;

[0146] The interference recognition module is used to quantify the information entropy sequence interference gradient of the foreign language writing assessment content text based on the paragraph logic segmentation data to obtain the information entropy sequence interference gradient data; perform native language system interference recognition on the foreign language writing assessment content text based on the information entropy sequence interference gradient data to obtain the content paragraph native language system interference recognition data;

[0147] The anomaly recognition optimization module is used to obtain the underlying learning architecture of the AI ​​tool; the underlying learning architecture of the AI ​​tool is optimized for adaptive language system anomaly recognition based on the native language system interference recognition data of the content paragraphs to obtain the AI ​​language system anomaly recognition architecture;

[0148] An effect evaluation module is used to evaluate the foreign language teaching effect of the foreign language writing assessment content text according to the AI language system anomaly recognition architecture, obtain the foreign language teaching effect evaluation data, and send the foreign language teaching effect evaluation data to the terminal.

[0149] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An AI-based method for evaluating the teaching effect of foreign language teaching, characterized in that, It includes the following steps: Step S1: Obtain the text of the foreign language writing assessment content submitted by the student; perform logical segmentation of the text paragraphs of the foreign language writing assessment content to obtain the paragraph logical segmentation data of the foreign language writing assessment content; Step S2: Perform information entropy sequence interference gradient quantization on the foreign language writing assessment content text based on the paragraph logical segmentation data to obtain information entropy sequence interference gradient data; Perform native language system interference recognition on the foreign language writing assessment content text according to the information entropy sequence interference gradient data to obtain content paragraph native language system interference recognition data; among them, Step S2 includes: Step S21: Perform analysis of sentence pattern parallelism / subordination connection on the foreign language writing assessment content text according to the paragraph logical segmentation data to obtain sentence pattern parallelism / subordination connection data; Step S22: Perform logical jump link recognition on the foreign language writing assessment content text based on the sentence pattern parallelism / subordination connection data and the paragraph logical segmentation data to obtain content paragraph logical jump data; Step S23: Perform information entropy sequence interference gradient quantization on the foreign language writing assessment content text based on the content paragraph logical jump data to obtain information entropy sequence interference gradient data; Step S23 includes the following steps: Step S231: Perform statistics on the frequency of logical jump connection words on the foreign language writing assessment content text based on the content paragraph logical jump data to obtain logical jump connection word frequency data; Step S232: Decompose the semantic span word vector of the connection word through the logical jump connection word frequency data to obtain the semantic span word vector of the connection word; Step S233: Perform regression analysis on the grammatical ambiguity of the front and back paragraphs of the foreign language writing assessment content text according to the semantic span word vector of the connection word and the logical jump connection word frequency data to obtain the grammatical ambiguity regression data of the front and back paragraphs; Step S234: Perform simulation analysis on the disorder state of information entropy sequence dimension constraint based on the semantic span word vector of the connection word and the grammatical ambiguity regression data of the front and back paragraphs to obtain the disorder state of information entropy sequence dimension constraint; Step S235: Perform information entropy sequence interference gradient quantization on the disorder state of information entropy sequence dimension constraint to obtain information entropy sequence interference gradient data; Step S24: Perform native language system interference recognition on the foreign language writing assessment content text according to the information entropy sequence interference gradient data and the content paragraph logical jump data to obtain content paragraph native language system interference recognition data; Step S3: Obtain the underlying learning architecture of the AI tool; perform adaptive language system anomaly recognition and optimization on the underlying learning architecture of the AI tool according to the content paragraph native language system interference recognition data to obtain the AI language system anomaly recognition architecture; Step S4: Perform an evaluation of the foreign language teaching effect on the foreign language writing assessment content text according to the AI language system anomaly recognition architecture to obtain foreign language teaching effect evaluation data, and send the foreign language teaching effect evaluation data to the terminal.

2. The AI-based foreign language teaching lecture effect evaluation method according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Obtain the text of the foreign language writing assessment content submitted by the student; Step S12: Perform text cleaning on the foreign language writing assessment content text to obtain the cleaned text of the foreign language writing assessment content; Step S13: Perform discourse structure narrative analysis on the cleaned text of the foreign language writing assessment content to obtain discourse structure narrative data of the writing content; Step S14: Perform text paragraph logical segmentation on the cleaned text of the foreign language writing assessment content according to the discourse structure narrative data of the writing content to obtain paragraph logical segmentation data of the foreign language writing assessment content.

3. The AI-based foreign language teaching lecture effect evaluation method according to claim 1, characterized in that Step S234 includes the following steps: Perform polysemy clustering analysis on the semantic span word vectors of conjunctions to obtain polysemy clustering data of conjunctions; Perform grammatical boundary fuzzy trend analysis on the grammatical fuzzy regression data of the front and back paragraphs to obtain grammatical boundary fuzzy trend data of the front and back paragraphs; Perform context change boundary fitting based on the polysemy clustering data of conjunctions and the grammatical boundary fuzzy trend data of the front and back paragraphs to obtain context change boundary fitting data before and after paragraphs; Perform context boundary fitting state simulation on the context change boundary fitting data before and after paragraphs based on Markov chain to obtain context boundary state fitting data; Perform information entropy sequence dimension constraint disorder state simulation analysis according to the context boundary state fitting data to obtain information entropy sequence dimension constraint disorder state.

4. The AI-based foreign language teaching effect evaluation method according to claim 1, wherein Step S24 includes the following steps: Step S241: Obtain the native language system structure data of different students; Step S242: Perform native language transfer feature analysis on the text of the foreign language writing assessment content according to the content paragraph logical jump data and the native language system structure data to obtain native language transfer feature data of the assessment content; Step S243: Perform syntactic structure logical deviation identification based on the native language transfer feature data of the assessment content and the content paragraph logical jump data to obtain native language syntactic structure logical deviation data; Step S244: Perform semantic deviation identification on the native language transfer feature data of the assessment content to obtain native language semantic feature deviation data; Step S245: Perform native language dominant syntactic framework mapping on the text of the foreign language writing assessment content according to the native language syntactic structure logical deviation data and the native language semantic feature deviation data to obtain native language dominant syntactic framework mapping data; Step S246: Perform native language system interference identification according to the information entropy sequence interference gradient data and the native language dominant syntactic framework mapping data to obtain native language system interference identification data of the content paragraph.

5. The AI-based foreign language teaching lecture effect evaluation method according to claim 1, characterized in that Step S3 includes the following steps: Step S31: Obtain the underlying learning architecture of the AI tool; Step S32: Perform normalization processing on the information entropy sequence interference gradient data to obtain normalized information entropy sequence interference gradient data; Step S33: Perform AI architecture disorder degree identification learning reinforcement on the underlying learning architecture of the AI tool according to the normalized information entropy sequence interference gradient data to obtain an AI disorder degree identification learning reinforcement architecture; Step S34: Perform adaptive language system anomaly identification optimization on the AI disorder degree identification learning reinforcement architecture based on the native language system interference identification data of the content paragraph to obtain an AI language system anomaly identification architecture.

6. The AI-based foreign language teaching lecture effect evaluation method according to claim 5, characterized in that, Step S33 includes the following steps: Step S331: Perform topological information entropy discretization processing on the normalized information entropy sequence interference gradient data to obtain an information entropy topological discrete feature space; Step S332: Perform feature orthogonal mapping and information entropy reconstruction processing on the information entropy topological discrete feature space to obtain a feature mapping information entropy tensor; Step S333: Perform interference gradient calibration processing based on the feature mapping information entropy tensor to obtain interference gradient calibration data; Step S334: Based on the interference gradient calibration data, perform AI architecture disorder degree recognition learning enhancement on the underlying learning architecture of the AI tool to obtain an AI disorder degree recognition learning enhanced architecture.

7. The AI-based foreign language teaching lecture effect evaluation method according to claim 5, wherein Step S34 includes the following steps: Step S341: Perform mother tongue part-of-speech interference abnormal logic analysis on the mother tongue system interference recognition data of the content paragraph to obtain mother tongue part-of-speech interference abnormal logic data; Step S342: Based on the mother tongue part-of-speech interference abnormal logic data, perform part-of-speech logic metaphor deviation induction to obtain part-of-speech logic metaphor deviation data; Step S343: Perform concept cross-domain metaphor abnormal analysis on the part-of-speech logic metaphor deviation data to obtain concept cross-domain metaphor abnormal data; Step S344: Based on the part-of-speech logic metaphor deviation data and the concept cross-domain metaphor abnormal data, perform adaptive language system abnormal recognition optimization on the AI disorder degree recognition learning enhanced architecture to obtain an AI language system abnormal recognition architecture.

8. An AI-based foreign language teaching effect evaluation system, characterized in that, For implementing the AI-based foreign language teaching lecture effect evaluation method as described in claim 1, the AI-based foreign language teaching lecture effect evaluation system includes: A paragraph logic segmentation module, configured to obtain the text of the foreign language writing assessment content submitted by the student; perform text paragraph logic segmentation on the text of the foreign language writing assessment content to obtain the paragraph logic segmentation data of the foreign language writing assessment content; An interference recognition module, configured to perform information entropy sequence interference gradient quantization on the text of the foreign language writing assessment content based on the paragraph logic segmentation data to obtain information entropy sequence interference gradient data; perform mother tongue system interference recognition on the text of the foreign language writing assessment content according to the information entropy sequence interference gradient data to obtain the mother tongue system interference recognition data of the content paragraph; An abnormal recognition optimization module, configured to obtain the underlying learning architecture of the AI tool; perform adaptive language system abnormal recognition optimization on the underlying learning architecture of the AI tool according to the mother tongue system interference recognition data of the content paragraph to obtain an AI language system abnormal recognition architecture; An effect evaluation module, configured to perform foreign language teaching lecture effect evaluation on the text of the foreign language writing assessment content according to the AI language system abnormal recognition architecture to obtain foreign language teaching lecture effect evaluation data, and send the foreign language teaching lecture effect evaluation data to the terminal.

Images