A senior high school english reading comprehension auxiliary learning method and device

Abstract

The application provides an auxiliary learning system for high school English subject reading comprehension questions, which helps students better improve their English learning ability. First, download our application software through APK file on Android mobile terminal; second, ask the specific learning level of the user, and divide the learning mode of the software into "high", "medium" and "low" three modes according to the evaluation score input by the user, so as to provide different difficulty word and sentence recognition and analysis; then, further ask the learning interest of the user, and recommend high-quality related category good articles for the user in real time through the interface of ChatGPT by inputting the characteristic words of the user; finally, the user can select the necessary words and articles to be collected, and view the content and analysis in the collection folder at any time. The auxiliary learning software for high school English subject reading comprehension questions can accurately carry out personalized teaching on the high school student group, improve the English reading learning efficiency of the user, and promote the improvement of the English ability of the user and the popularization of culture.

Description

Technical Field

[0001] The present invention relates to the field of text target extraction in computer vision, and in particular to a high school English reading comprehension auxiliary learning method and device. Background Art

[0002] Reading comprehension is a crucial component of English language instruction, particularly in exams. Intensive reading is essential for fully understanding and grasping a passage. Most English teachers devote considerable time to providing students with detailed analysis to facilitate comprehension. However, many high-quality extracurricular reading sessions lack the support of teachers, who are often unable to fully immerse themselves in the intensive reading of each student's passages. Therefore, learning assistance software is crucial to meet the needs of a broad student base, acting as a substitute in the absence of teachers.

[0003] There are many question-searching and learning apps available on the market, such as Baidu Question Search and Xiaoyuan Question Search. However, these apps either only provide answers through online matching or offer existing reading comprehension data from a database, failing to provide machine-assisted, in-depth reading of real-world, printed articles. Without a teacher to guide students, students are left with the choice of copying answers or reading articles on the app, which may not be high-quality or up-to-date. When faced with a good printed article, they must laboriously search and comprehend it word by word. Our English reading assistance software, available on smartphones, offers machine-assisted comprehension anytime, anywhere, allowing users to quickly analyze the essence and key points of a text without a teacher or classmates. Summary of the Invention

[0004] The present invention provides a high school English reading comprehension auxiliary learning method and device to solve the problems existing in the prior art.

[0005] In a first aspect, the present invention provides a method for assisting high school English reading comprehension learning, or in other words, a method for processing images or texts containing English reading content, specifically comprising the following steps:

[0006] Step S1: obtaining an evaluation score, dividing the learning mode according to the evaluation score, and identifying and parsing words and sentences of different difficulty modes;

[0007] Step S2: Obtain learning interest feature words, obtain high-quality related articles using the learning interest feature words through the interface of ChatGPT, and highlight the related articles;

[0008] Step S3: Collect important words, article contents and their analyses.

[0009] Preferably, in step S1, the difficulty mode includes "low", "medium" and "high", wherein "low" corresponds to CET-4 words, "medium" corresponds to CET-6 words, and "high" corresponds to IELTS words.

[0010] Preferably, in step S1, an evaluation score is obtained, and after the learning mode is divided according to the evaluation score, words and sentences of different difficulty levels are identified and parsed, which specifically includes the following steps:

[0011] Step S101: Segment and extract the article body and title body based on the uploaded reading comprehension image or text using OCR recognition technology;

[0012] Step S102: traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database, highlight the matched words in red in the article, and add a link to open a window showing the analysis of the word when clicked;

[0013] Step S103: traverse the English phrases in the article and match them one by one with the key phrases required for high school English exams that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the analysis of the phrase when clicked.

[0014] Step S104: traverse the English words in the article and match them with the pre-set database of different difficulty levels according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article, and a link is added to open a window showing the analysis of the word when clicked;

[0015] Step S105: perform similarity matching between the extracted title content text and the sentence text of the article, and underline one or two sentences with the highest similarity;

[0016] Step S106: Segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and change them to italics to highlight them.

[0017] Preferably, in step S101, based on the uploaded reading comprehension picture or text, the article body and the title body are segmented and extracted in combination with OCR recognition technology, which specifically includes the following steps:

[0018] Step S101a: pre-process the uploaded reading comprehension picture to form a pre-processed reading comprehension picture;

[0019] Step S101b: inputting the pre-processed reading comprehension image into a neural network model, and separating the characters in the image through the neural network model to form separated characters;

[0020] Step S101c: converting the separated characters through a machine learning algorithm to form editable text obtained after the reading comprehension image is recognized;

[0021] Step S101d: forming complete text content based on the editable text obtained after recognizing the reading comprehension image and the uploaded reading comprehension text.

[0022] Preferably, the preprocessing in step S101a includes performing operations such as denoising, binarization, and tilt correction on the image to improve the recognition rate.

[0023] Preferably, in step S105, the similarity matching is implemented by an NLP algorithm to calculate the semantic similarity between the two texts based on semantic vectors, which specifically includes the following steps:

[0024] Step S105a: segment the text using the Jieba library to obtain a vocabulary list for each text;

[0025] Step S105b: Convert each word into a vector according to the value in the vocabulary list through the Word2Vec model;

[0026] Step S105c: performing weighted averaging on the vocabulary vectors of each text to form a semantic vector for each text;

[0027] Step S105d: Calculate the semantic similarity between two texts based on the semantic vector of each text.

[0028] Preferably, in step S105a, the vocabulary list of each text includes information such as the number of times the text appears and the context.

[0029] Preferably, in step S105d, the similarity between the semantic vectors of the two texts is calculated, and the specific formula is as follows:

[0030]

[0031] Among them, similarity semantic (text i ,text j ) represents text i and text j The semantic similarity between texti Represents text i The semantic vector of ||v texti || represents the semantic vector v texti The norm of <v texti ,v textj > represents the inner product of two semantic vectors.

[0032] Preferably, in step S106, syntactic analysis is performed based on a maximum margin Markov network, specifically including the following steps:

[0033] Step S106a: perform word segmentation and part-of-speech tagging on each input sentence using the Jieba function library;

[0034] Step S106b: extract the current word, part of speech and dependency data provided by the jieba function library from each sentence to form feature data;

[0035] Step S106c: constructing a graph structure according to the feature data;

[0036] Step S106d: Calculate the score of each dependency relationship through the maximum margin Markov model, and find the optimal dependency tree structure through the decoding algorithm;

[0037] Step S106e: Convert the optimal dependency tree structure into a syntactic analysis result.

[0038] Preferably, in step S106c, the graph structure includes nodes and edges; wherein the nodes represent each word in the sentence, and the edges represent the dependency relationship between words.

[0039] Preferably, in step S106d, the decoding algorithm includes a Viterbi algorithm, and the calculation formula of the Viterbi algorithm is as follows:

[0040] δ t (i) = max P(i t =i,i t-1 ,…,i1,o t ,…,o1|λ), i=1,2,…,N

[0041] Among them, δ t (i) represents the maximum probability value of all nodes in state i at time t (i.e., the shortest path from left to right in a similar graph structure), i t represents the shortest path, o t represents the observation matrix parameters, o t is a constant, λ represents the probability matrix parameter, and λ is a constant.

[0042] Preferably, in step S106e, the optimal dependency tree structure is converted into a syntactic analysis result, and the specific calculation formula is as follows:

[0043]

[0044] Among them, Φ(x, y) represents the feature vector of the syntactic tree y corresponding to x, w represents the feature weight, and G(x) represents the syntactic rule to which x belongs.

[0045] Preferably, in step S2, learning interest feature words are obtained, high-quality related articles are obtained by accessing the ChatGPT interface for the learning interest feature words, and the related articles are highlighted, which specifically includes the following steps:

[0046] Step S201: After acquiring the learning interest feature words, call the ChatGPT interface and send a query request to ChatGPT;

[0047] Step S202: After receiving the query request, ChatGPT processes the query request, obtains high-quality articles related to the learning interest feature words, and returns them;

[0048] Step S203: obtaining high-quality articles related to the learning interest feature words obtained through ChatGPT through an interface, and highlighting the related articles.

[0049] Preferably, in step S201, the query request includes a query request formed based on learning interest feature words for guiding ChatGPT to obtain high-quality articles related to the learning interest feature words.

[0050] Preferably, in step S203, highlighting the relevant articles specifically includes the following steps:

[0051] Step S203a: Obtain key words that must be tested in high school English and import them into the database. Then, traverse the English words in the article and match them one by one with the words in the database. The matched words are highlighted in red in the article. A link is added and clicked to open a window showing the analysis of the word.

[0052] Step S203b: Obtain the key phrases required for high school English exams and import them into the database. Then, traverse the English phrases in the article and match them one by one with the phrases in the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the analysis of the phrase when clicked.

[0053] Step S203c: Obtain CET-4 words, CET-6 words, and IELTS words and import them into different databases respectively. Traverse the English words in the article and match them with different databases according to the selected difficulty mode. Highlight the matched words in blue in the article. Add a link and click it to open a window showing the analysis of this word.

[0054] Preferably, in step S3, the saved content and analysis can be viewed at any time.

[0055] In a second aspect, the present invention further provides a high school English reading comprehension auxiliary learning device, or a device for processing pictures or texts containing English reading content, which specifically includes the following modules:

[0056] A sentence recognition and parsing module is used to obtain evaluation scores, classify learning modes according to the evaluation scores, and then recognize and parse words and sentences of different difficulty modes;

[0057] A highlighting processing module is used to obtain learning interest feature words, obtain high-quality related articles by accessing the ChatGPT interface of the learning interest feature words, and highlight the related articles;

[0058] The collection processing module is used to collect important words, article content and their analysis.

[0059] Preferably, in the sentence recognition and parsing module, the difficulty mode includes "low", "medium" and "high", wherein "low" corresponds to CET-4 words, "medium" corresponds to CET-6 words, and "high" corresponds to IELTS words.

[0060] Preferably, the sentence recognition and parsing module specifically includes the following submodules:

[0061] The first submodule of sentence recognition and parsing is used to segment and extract the main body of the article and the main body of the title based on the uploaded reading comprehension image or text, combined with OCR recognition technology;

[0062] The second submodule of sentence recognition and parsing is used to traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database. The matched words are highlighted in red in the article, and a link is added to open a window showing the analysis of the word when clicked;

[0063] The third submodule, sentence recognition and parsing, is used to traverse the English phrases in the article and match them one by one with the key phrases required for high school English exams that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the parsing of the phrase when clicked.

[0064] The fourth submodule, sentence recognition and parsing, is used to traverse the English words in the article and match them with a pre-set database of different difficulty levels according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article and a link is added. Clicking it will open a window showing the analysis of the word.

[0065] The fifth submodule of sentence recognition and parsing is used to match the extracted title content text with the sentence text of the article based on similarity, and underline the one or two sentences with the highest similarity;

[0066] The sixth submodule of sentence recognition and parsing is used to segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and highlight them in italics.

[0067] Preferably, the sentence recognition and parsing first submodule specifically includes the following submodules:

[0068] The first processing module is used to pre-process the uploaded reading comprehension picture to form a pre-processed reading comprehension picture;

[0069] A second processing module is used to input the pre-processed reading comprehension image into the neural network model, and separate the characters in the image through the neural network model to form separated characters;

[0070] The third processing module is used to convert the separated characters through a machine learning algorithm to form editable text obtained after the reading comprehension image recognition;

[0071] The fourth processing module is used to form complete text content based on the editable text obtained after the reading comprehension picture is recognized and the uploaded reading comprehension text.

[0072] Preferably, in the first processing module, the preprocessing includes performing operations such as denoising, binarization, and tilt correction on the image to improve the recognition rate.

[0073] Preferably, the fifth submodule of sentence recognition and parsing specifically includes the following submodules:

[0074] The fifth processing module is used to segment the text using the Jieba function library to obtain a vocabulary list for each text;

[0075] The sixth processing module is used to convert each word into a vector according to the value in the vocabulary list through the Word2Vec model;

[0076] The seventh processing module is used to perform weighted averaging on the vocabulary vectors of each text to form a semantic vector for each text;

[0077] The eighth processing module is used to calculate the semantic similarity between two texts based on the semantic vector of each text.

[0078] Preferably, in the fifth processing module, the vocabulary list of each text includes information such as the number of times the text appears and the context.

[0079] Preferably, in the eighth processing module, the similarity between the semantic vectors of two texts is calculated, and the specific formula is as follows:

[0080]

[0081] Among them, similarity semantic (text i ,text j) represents text i and text j The semantic similarity between texti Represents text i The semantic vector of ||v texti ||×||v textj || represents the semantic vector V texti and v textj The inner product between <v texti , V textj > represents the inner product of two semantic vectors.

[0082] Preferably, the sixth submodule of sentence recognition and parsing specifically includes the following submodules:

[0083] The ninth processing module is used to perform word segmentation and part-of-speech tagging on each input sentence through the Jieba function library;

[0084] The tenth processing module is used to extract the current word, part of speech and dependency data provided by the jieba function library from each sentence to form feature data;

[0085] an eleventh processing module, configured to construct a graph structure based on the feature data;

[0086] The twelfth processing module is used to calculate the score of each dependency relationship through the maximum margin Markov model and find the optimal dependency tree structure through the decoding algorithm;

[0087] The thirteenth processing module is used to convert the optimal dependency tree structure into a syntactic analysis result.

[0088] Preferably, in the eleventh processing module, the graph structure includes nodes and edges; wherein the nodes represent each word in the sentence, and the edges represent the dependency relationship between the words.

[0089] Preferably, in the twelfth processing module, the decoding algorithm includes a Viterbi algorithm, and the calculation formula of the Viterbi algorithm is as follows:

[0090] δ t (i) = max P(i t =i,i t-1 ,…,i1,o t ,…,o1|λ), i=1,2,…,N

[0091] Among them, δ t (i) represents the maximum probability value of all nodes in state i at time t (i.e., the shortest path from left to right in a similar graph structure), i t represents the shortest path, o trepresents the observation matrix parameters, o t is a constant, λ represents the probability matrix parameter, and λ is a constant.

[0092] Preferably, in the thirteenth processing module, the optimal dependency tree structure is converted into a syntactic analysis result, and the specific calculation formula is as follows:

[0093]

[0094] Among them, Φ(x, y) represents the feature vector of the syntactic tree y corresponding to x, w represents the feature weight, and G(x) represents the syntactic rule to which x belongs.

[0095] Preferably, the highlight processing module specifically includes the following submodules:

[0096] The first submodule of highlight processing is used to obtain the learning interest feature words, call the ChatGPT interface and send a query request to ChatGPT;

[0097] The second submodule of highlighting processing is used to process the query request after ChatGPT receives the query request, obtain high-quality articles related to the learning interest feature words and return them;

[0098] The third submodule of highlighting processing is used to obtain high-quality articles related to the learning interest feature words obtained through ChatGPT through an interface, and highlight the related articles.

[0099] Preferably, in the first submodule of highlighting processing, the query request includes information formed based on learning interest feature words for guiding ChatGPT to obtain high-quality articles related to the learning interest feature words.

[0100] Preferably, the third submodule of highlighting processing specifically includes the following submodules:

[0101] The first module of highlight processing is used to obtain the key words that must be tested in high school English and import them into the database. It traverses the English words in the article and matches them one by one with the words in the database. The matched words are highlighted in red in the article. Clicking on a link will open a window to display the analysis of this word.

[0102] The second module of highlight processing is used to obtain the key phrases that must be tested in high school English and import them into the database. It traverses the English phrases in the article and matches them one by one with the phrases in the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window to display the analysis of this phrase when clicked.

[0103] The third module of highlight processing is used to obtain CET-4 words, CET-6 words, and IELTS words and import them into different databases respectively. It traverses the English words in the article and matches them with different databases according to the selected difficulty mode. The matched words are highlighted in blue in the article. Click on the added link to open a window to display the analysis of this word.

[0104] Preferably, the content and analysis after the collection processing by the collection processing module can be viewed at any time.

[0105] In a third aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a high school English reading comprehension auxiliary learning method as described in any one of the first aspects of the present application.

[0106] In a fourth aspect, the present invention further provides an electronic device comprising: a memory storing a computer program; and a processor communicatively connected to the memory for executing a high school English reading comprehension assisted learning method described in any one of the first aspects of this application when the computer program is called.

[0107] Compared with the prior art, the present invention has the following obvious outstanding substantial features and significant advantages:

[0108] The present invention provides a high school English reading comprehension auxiliary learning method and device, which solves the problems existing in the prior art. It has the following advantages: (1) The present invention is an auxiliary learning software for high school English subject reading comprehension questions, which can effectively help students better improve their English learning ability. Using multiple difficulty modes, users can choose a learning method that suits them to learn English reading, which is more personalized. At the same time, the present invention does not provide actual answers. All methods are based on replacing the teacher's ideas to provide detailed article analysis for learning, and truly simulate the teacher's explanation in English class. (2) The present invention provides analysis for English reading by embedding the ChatGPT large model, and further customizes the relevant articles that users want to read. The multimodal analysis capability and real-time processing capability of the large model give unique advantages, allowing users to freely pursue topics of interest, thereby maximizing the release of personalized ideas and concepts. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] The accompanying drawings, which constitute part of the present invention, are provided to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0110] Figure 1 It is a flow chart of a high school English reading comprehension auxiliary learning method according to a preferred embodiment of the present invention.

[0111] Figure 2 The present invention is a flowchart of a high school English reading comprehension auxiliary learning software according to a preferred embodiment of the present invention.

[0112] Figure 3 It is a structural diagram of a high school English reading comprehension auxiliary learning device according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0113] The present invention provides a method and device for assisting high school English reading comprehension learning. To make the objectives, technical solutions, and effects of the present invention more clear and explicit, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only intended to illustrate the present invention and are not intended to limit the present invention.

[0114] It should be noted that the terms "first," "second," and the like in the specification and claims of the present invention and the accompanying drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this manner are interchangeable under appropriate circumstances. In addition, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to these processes, methods, products, or apparatuses.

[0115] Example 1:

[0116] The embodiment of the present application provides a method for assisting learning of high school English reading comprehension, the execution subject of the method can be a smart terminal, and the smart terminal can be any user-end device with a learning app installed, specifically a smart phone, a tablet computer and a computer. The smart terminal can assist users in learning high school English subject reading comprehension questions based on the installed learning app. In the process of executing the above processing, the smart terminal can communicate with the background server of the learning app to obtain the relevant data required in the above process. Of course, in another feasible scenario, the above method can also be completed by the smart terminal and the background server of the learning app in collaboration, that is, the smart terminal can be used to interact with the user and transmit the interaction-related data to the background server for processing, and receive the processing results sent by the background server. The background server can be used to receive the interaction-related data transmitted by the smart terminal, and based on the preset data processing logic, generate and feedback the corresponding processing results. This embodiment is explained by taking the execution subject as the smart terminal as an example. The collaborative processing process of the smart terminal and the background server is similar to it, and this embodiment will not be repeated.

[0117] The following will be combined with specific implementation methods. Figure 1-Figure 2 The processing flow shown in the figure is described in detail. Figure 1-Figure 2 As shown, the high school English reading comprehension auxiliary learning method described in this embodiment specifically includes the following steps:

[0118] Step S1: Inquire about the user's specific learning level, obtain the evaluation score input by the user, divide the learning mode according to the evaluation score, and identify and analyze words and sentences of different difficulty levels.

[0119] The difficulty modes include "low", "medium" and "high", wherein "low" corresponds to CET-4 words, "medium" corresponds to CET-6 words, and "high" corresponds to IELTS words.

[0120] Optionally, step S1 specifically includes the following steps:

[0121] Step S101: Based on the uploaded reading comprehension image or text, the article body and the title body are segmented and extracted in combination with OCR recognition technology.

[0122] Optionally, step S101 specifically includes the following steps:

[0123] Step S101a: pre-process the uploaded reading comprehension picture to form a pre-processed reading comprehension picture.

[0124] The preprocessing includes performing operations such as denoising, binarization, and tilt correction on the image to improve the recognition rate.

[0125] Step S101b: input the pre-processed reading comprehension image into the neural network model, and separate the characters in the image through the neural network model to form separated characters.

[0126] Step S101c: convert the separated characters through a machine learning algorithm to form editable text obtained after the reading comprehension picture is recognized.

[0127] Step S101d: forming complete text content based on the editable text obtained after recognizing the reading comprehension image and the uploaded reading comprehension text.

[0128] Step S102: traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database, highlight the matched words in red in the article, and add a link to open a window to display the analysis of this word when clicked.

[0129] Step S103: traverse the English phrases in the article and match them one by one with the key phrases for high school English that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article, and a link is added to open a window showing the analysis of the phrase when clicked.

[0130] Step S104: traverse the English words in the article and match them with different pre-set databases according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article. Click on the added link to open a window showing the analysis of the word. Different databases are pre-imported into the database, which store words for CET-4, CET-6 and IELTS respectively to correspond to different difficulty levels. When the difficulty mode selected by the user is "low", the program queries the CET-4 database; when the difficulty mode selected by the user is "medium", the program queries the CET-6 database; when the difficulty mode selected by the user is "high", the program queries the IELTS database.

[0131] Step S105: perform similarity matching between the extracted title content text and the sentence text of the article, and underline one or two sentences with the highest similarity.

[0132] Optionally, step S105 specifically includes the following steps:

[0133] Step S105a: segment the text using the Jieba function library to obtain a vocabulary list for each text.

[0134] The vocabulary list of each text includes information such as the number of times the text appears and the context.

[0135] Step S105b: Convert each word into a vector according to the value in the vocabulary list through the Word2Vec model.

[0136] Step S105c: perform weighted averaging on the vocabulary vectors of each text to form a semantic vector for each text.

[0137] Step S105d: Calculate the semantic similarity between two texts based on the semantic vector of each text.

[0138] The similarity between the semantic vectors of two texts is calculated using the following formula:

[0139]

[0140] Among them, similarity semantic (text i ,text j ) represents text i and text j The semantic similarity between texti Represents text i The semantic vector of ||v texti ||×||v textj || represents the semantic vector V texti and vtextj The inner product between <v texti , V textj > represents the inner product of two semantic vectors.

[0141] Step S106: Segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and change them to italics to highlight them.

[0142] Optionally, syntactic analysis is performed based on a maximum margin Markov network, specifically including the following steps:

[0143] Step S106a: perform word segmentation and part-of-speech tagging on each input sentence using the Jieba function library.

[0144] Step S106b: extract the current word, part of speech and dependency data provided by the jieba function library from each sentence to form feature data.

[0145] Step S106c: construct a graph structure according to the feature data.

[0146] The graph structure includes nodes and edges; the nodes represent each word in the sentence, and the edges represent the dependency relationship between words.

[0147] Step S106d: Calculate the score of each dependency relationship through the maximum margin Markov model, and find the optimal dependency tree structure through the decoding algorithm.

[0148] The decoding algorithm includes the Viterbi algorithm, and the calculation formula of the Viterbi algorithm is as follows:

[0149] δ t (i) = maxP(i t =i,i t-1 ,…,i1,o t ,…,o1|λ), i=1,2,…,N

[0150] Among them, δ t (i) represents the maximum probability value of all nodes in state i at time t (i.e., the shortest path from left to right in a similar graph structure), i t represents the shortest path, o t represents the observation matrix parameters, o t is a constant, λ represents the probability matrix parameter, and λ is a constant.

[0151] Step S106e: Convert the optimal dependency tree structure into a syntactic analysis result.

[0152] The optimal dependency tree structure is converted into a syntactic analysis result. The specific calculation formula is as follows:

[0153]

[0154] Among them, Φ(x, y) represents the feature vector of the syntactic tree y corresponding to x, w represents the feature weight, and G(x) represents the syntactic rule to which x belongs.

[0155] Step S2: further inquire about the user's learning interests, obtain the learning interest feature words input by the user, obtain high-quality related articles through the interface of ChatGPT using the learning interest feature words, and highlight the related articles.

[0156] Optionally, step S2 specifically includes the following steps:

[0157] Step S201: After acquiring the learning interest feature words, call the ChatGPT interface and send a query request to ChatGPT.

[0158] The query request includes information formed based on learning interest feature words and used to guide ChatGPT to obtain high-quality articles related to the learning interest feature words.

[0159] Step S202: After receiving the query request, ChatGPT processes the query request, obtains high-quality articles related to the learning interest feature words, and returns them.

[0160] Step S203: obtaining high-quality articles related to the learning interest feature words obtained through ChatGPT through an interface, and highlighting the related articles.

[0161] Optionally, step S203 specifically includes the following steps:

[0162] Step S203a, traverse the English words in the article, and match them one by one with the key high school English test words pre-imported into the database, highlight the matched words in red in the article, and add a link to open a window showing the analysis of this word when clicked.

[0163] Step S203b: traverse the English phrases in the article and match them one by one with the key phrases for high school English that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article, and a link is added to open a window showing the analysis of the phrase when clicked.

[0164] Step S203c, traverse the English words in the article, and match them with different pre-set databases according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article, and a link is added and clicked to open a window showing the analysis of the word. Among them, different databases are pre-imported into the database, which store words for CET-4, CET-6 and IELTS respectively to correspond to different difficulty levels. When the difficulty mode selected by the user is "low", the program queries the CET-4 database; when the difficulty mode selected by the user is "medium", the program queries the CET-6 database; when the difficulty mode selected by the user is "high", the program queries the IELTS database.

[0165] Step S3: Collect important words, article contents and their analyses.

[0166] In an embodiment, a collection function is provided to users to enhance their learning and review experience. Users can choose to collect necessary words and articles and view the collected content and analysis in the collection folder at any time.

[0167] For example, when a user encounters a word that requires special attention during browsing, they can directly click on the word. Immediately, the program will pop up a detailed analysis window for this word at the top of the page or at a specified location. In the upper right corner of the analysis window, there is a prominent five-pointed star icon as a favorite button. After the user clicks this five-pointed star icon, the word and its analysis information are automatically recorded in a special favorite word database for subsequent reference.

[0168] Similarly, to save an article, users click on the article title or a designated area, and the program displays the full article content. A five-pointed star-shaped "Save" button is located in the upper right corner of the article. Once clicked, the entire article and its associated information are added to a dedicated database of saved articles.

[0169] To facilitate user management and access to saved content, Favorites are divided into two sections: Word Favorites and Article Favorites, allowing users to view them separately. For example, in the Word Favorites, the saved words themselves can be displayed in list format for quick identification; in the Article Favorites, the first five words of each article are displayed as a preview to help users quickly recall the article content. To view the complete information of a specific word or article, users only need to click the corresponding item in the Favorites, and the program will call and display the saved content from the corresponding database.

[0170] Example 2:

[0171] Based on the same technical concept, this embodiment also provides a high school English reading comprehension auxiliary learning device, such as Figure 3 As shown, the device specifically includes the following modules:

[0172] The sentence recognition and parsing module is used to obtain evaluation scores, divide the learning modes according to the evaluation scores, and then recognize and parse words and sentences of different difficulty modes.

[0173] The difficulty modes include "low", "medium" and "high", wherein "low" corresponds to CET-4 words, "medium" corresponds to CET-6 words, and "high" corresponds to IELTS words.

[0174] The sentence recognition and parsing module specifically includes the following submodules:

[0175] The first submodule of sentence recognition and parsing is used to segment and extract the article body and title body based on the uploaded reading comprehension pictures or texts in combination with OCR recognition technology.

[0176] Optionally, the first submodule of sentence recognition and parsing specifically includes the following submodules:

[0177] The first processing module is used to pre-process the uploaded reading comprehension picture to form a pre-processed reading comprehension picture.

[0178] The preprocessing includes performing operations such as denoising, binarization, and tilt correction on the image to improve the recognition rate.

[0179] The second processing module is used to input the pre-processed reading comprehension image into the neural network model, and separate the characters in the image through the neural network model to form separated characters.

[0180] The third processing module is used to convert the separated characters through a machine learning algorithm to form an editable text obtained after the reading comprehension picture recognition.

[0181] The fourth processing module is used to form complete text content based on the editable text obtained after the reading comprehension picture is recognized and the uploaded reading comprehension text.

[0182] The second submodule of sentence recognition and parsing is used to traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database. The matched words are highlighted in red in the article, and a link is added to open a window to display the analysis of the word when clicked.

[0183] The third submodule of sentence recognition and parsing is used to traverse the English phrases in the article and match them one by one with the key phrases required for high school English exams that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article, and a link is added to open a window showing the analysis of the phrase when clicked.

[0184] The fourth submodule of sentence recognition and analysis is used to obtain CET-4 words, CET-6 words, and IELTS words and import them into different databases respectively, traverse the English words in the article, and match them with different pre-set databases according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article, and a link is added to open a window to display the analysis of this word when clicked; among them, different databases are pre-imported into the database, which store words for CET-4, CET-6 and IELTS respectively to correspond to different difficulty levels. When the difficulty mode selected by the user is "low", the program queries the CET-4 database; when the difficulty mode selected by the user is "medium", the program queries the CET-6 database; when the difficulty mode selected by the user is "high", the program queries the IELTS database.

[0185] The fifth submodule of sentence recognition and parsing is used to perform similarity matching between the extracted title content text and the sentence text of the article, and underline the one or two sentences with the highest similarity.

[0186] Optionally, the fifth submodule of sentence recognition and parsing specifically includes the following submodules:

[0187] The fifth processing module is used to segment the text through the jieba function library to obtain a vocabulary list for each text.

[0188] The vocabulary list of each text includes information such as the number of times the text appears and the context.

[0189] The sixth processing module is used to convert each word into a vector according to the value in the vocabulary list through the Word2Vec model.

[0190] The seventh processing module is used to perform weighted averaging on the vocabulary vectors of each text to form a semantic vector for each text.

[0191] The eighth processing module is used to calculate the semantic similarity between two texts based on the semantic vector of each text.

[0192] The similarity between the semantic vectors of two texts is calculated using the following formula:

[0193]

[0194] Among them, similarity semantic (text i ,text j ) represents text i and text j The semantic similarity between texti Represents text i The semantic vector of ||v texti||×||v textj || represents the semantic vector V texti and v textj The inner product between <v texti , V textj > represents the inner product of two semantic vectors.

[0195] The sixth submodule of sentence recognition and parsing is used to segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and highlight them in italics.

[0196] Optionally, the sixth submodule of sentence recognition and parsing specifically includes the following submodules:

[0197] The ninth processing module is used to perform word segmentation and part-of-speech tagging on each input sentence through the jieba function library.

[0198] The tenth processing module is used to extract the current word, part of speech and dependency data provided by the jieba function library from each sentence to form feature data.

[0199] The eleventh processing module is used to construct a graph structure according to the feature data.

[0200] Preferably, in the eleventh processing module, the graph structure includes nodes and edges; wherein the nodes represent each word in the sentence, and the edges represent the dependency relationship between the words.

[0201] The twelfth processing module is used to calculate the score of each dependency relationship through the maximum margin Markov model and find the optimal dependency tree structure through the decoding algorithm.

[0202] The decoding algorithm includes the Viterbi algorithm, and the calculation formula of the Viterbi algorithm is as follows:

[0203] δ t (i) = max P(i t =i,i t-1 ,…,i1,o t ,…,o1|λ), i=1,2,…,N

[0204] Among them, δ t (i) represents the maximum probability value of all nodes in state i at time t (i.e., the shortest path from left to right in a similar graph structure), i t represents the shortest path, o t represents the observation matrix parameters, o t is a constant, λ represents the probability matrix parameter, and λ is a constant.

[0205] The thirteenth processing module is used to convert the optimal dependency tree structure into a syntactic analysis result.

[0206] The optimal dependency tree structure is converted into a syntactic analysis result. The specific calculation formula is as follows:

[0207]

[0208] Among them, Φ(x,y) represents the feature vector of the syntactic tree y corresponding to x, w represents the feature weight, and G(x) represents the syntactic rule to which x belongs.

[0209] The highlighting processing module is used to obtain learning interest feature words, obtain high-quality related articles by accessing the ChatGPT interface of the learning interest feature words, and highlight the related articles.

[0210] Optionally, the highlight processing module specifically includes the following submodules:

[0211] The first submodule of highlight processing is used to obtain learning interest feature words, call the ChatGPT interface and send a query request to ChatGPT.

[0212] The query request includes information formed based on learning interest feature words and used to guide ChatGPT to obtain high-quality articles related to the learning interest feature words.

[0213] The second submodule of highlighting processing is used to process the query request after ChatGPT receives the query request, obtain high-quality articles related to the learning interest feature words and return them.

[0214] The third submodule of highlighting processing is used to obtain high-quality articles related to the learning interest feature words obtained through ChatGPT through an interface, and highlight the related articles.

[0215] Optionally, the third submodule of highlighting processing specifically includes the following submodules:

[0216] The first module of highlight processing is used to obtain the key words that must be tested in high school English and import them into the database. It traverses the English words in the article and matches them one by one with the words in the database. The matched words are highlighted in red in the article. Clicking on a link will open a window to display the analysis of this word.

[0217] The second module of highlight processing is used to obtain the key phrases that must be tested in high school English and import them into the database. It traverses the English phrases in the article and matches them one by one with the phrases in the database. The matched phrases are highlighted in yellow in the article, and a link is added. Clicking it will open a window to display the analysis of this phrase.

[0218] The third module of highlight processing is used to obtain CET-4 words, CET-6 words, and IELTS words and import them into different databases respectively. It traverses the English words in the article and matches them with different databases according to the selected difficulty mode. The matched words are highlighted in blue in the article. Click on the added link to open a window to display the analysis of this word.

[0219] The collection processing module is used to collect important words and article content and their analysis, and the content and analysis after the collection processing by the collection processing module can be viewed at any time.

[0220] An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the high school English reading comprehension auxiliary learning method described in Example 1 of the present application.

[0221] An embodiment of the present application also provides an electronic device, which includes: a memory storing a computer program; and a processor communicating with the memory, which executes a high school English reading comprehension assisted learning method described in Example 1 of the present application when the computer program is called.

[0222] While the specific embodiments of the present invention have been described in detail above, these are merely exemplary and the present invention is not limited thereto. For those skilled in the art, any equivalent modifications and substitutions to the present invention are also within the scope of the present invention. Therefore, any equivalent changes and modifications made without departing from the spirit and scope of the present invention are intended to be encompassed within the scope of the present invention.

Claims

1. A method for assisting high school English reading comprehension learning, characterized in that: The specific steps include: Step S1: obtaining an evaluation score, dividing the learning mode according to the evaluation score, and identifying and parsing words and sentences of different difficulty modes; Step S2: Obtain learning interest feature words, obtain high-quality related articles using the learning interest feature words through the interface of ChatGPT, and highlight the related articles; Step S3: Collect important words and article contents and their analyses; In step S1, the evaluation scores are obtained, and after the learning modes are divided according to the evaluation scores, the words and sentences of different difficulty levels are identified and parsed, which specifically includes the following steps: Step S101: Segment and extract the article body and title body based on the uploaded reading comprehension image or text using OCR recognition technology; Step S102: traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database, highlight the matched words in red in the article, and add a link to open a window showing the analysis of the word when clicked; Step S103: traverse the English phrases in the article and match them one by one with the key phrases required for high school English exams that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the analysis of the phrase when clicked. Step S104: traverse the English words in the article and match them with the pre-set database of different difficulty levels according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article, and a link is added to open a window showing the analysis of the word when clicked; Step S105: perform similarity matching between the extracted title content text and the sentence text of the article, and underline one or two sentences with the highest similarity; Step S106: Segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and change them to italics to highlight them.

2. The method for assisting high school English reading comprehension learning according to claim 1, characterized in that: In step S101, based on the uploaded reading comprehension image or text, the article body and the title body are segmented and extracted in combination with OCR recognition technology, which specifically includes the following steps: Step S101a: pre-process the uploaded reading comprehension picture to form a pre-processed reading comprehension picture; Step S101b: inputting the pre-processed reading comprehension image into a neural network model, and separating the characters in the image through the neural network model to form separated characters; Step S101c: converting the separated characters through a machine learning algorithm to form editable text obtained after the reading comprehension image is recognized; Step S101d: forming complete text content based on the editable text obtained after recognizing the reading comprehension image and the uploaded reading comprehension text.

3. The method for assisting high school English reading comprehension learning according to claim 1, characterized in that: In step S105, similarity matching is performed between the extracted title content text and the sentence text of the article, and one or two sentences with the highest similarity are underlined and highlighted; The similarity matching is achieved by using an NLP algorithm to calculate the semantic similarity between two texts based on semantic vectors, specifically including the following steps: Step S105a: segment the text using the Jieba library to obtain a vocabulary list for each text; Step S105b: Convert each word into a vector according to the value in the vocabulary list through the Word2Vec model; Step S105c: performing weighted averaging on the vocabulary vectors of each text to form a semantic vector for each text; Step S105d: Calculate the semantic similarity between two texts based on the semantic vector of each text.

4. The method for assisting high school English reading comprehension learning according to claim 1, characterized in that: In step S106, all sentences in the article are segmented by periods, and long and difficult sentences are matched according to grammatical analysis and word count, and are changed to italics to highlight them; Among them, the syntactic analysis based on the maximum margin Markov network includes the following steps: Step S106a: perform word segmentation and part-of-speech tagging on each input sentence using the Jieba function library; Step S106b: extract the current word, part of speech and dependency data provided by the jieba function library from each sentence to form feature data; Step S106c: constructing a graph structure according to the feature data; Step S106d: Calculate the score of each dependency relationship through the maximum margin Markov model, and find the optimal dependency tree structure through the decoding algorithm; Step S106e: Convert the optimal dependency tree structure into a syntactic analysis result.

5. The method for assisting high school English reading comprehension learning according to claim 1, characterized in that: In step S2, learning interest feature words are obtained, high-quality related articles are obtained by accessing the ChatGPT interface for the learning interest feature words, and the related articles are highlighted, which specifically includes the following steps: Step S201: After acquiring the learning interest feature words, call the ChatGPT interface and send a query request to ChatGPT; Step S202: After receiving the query request, ChatGPT processes the query request, obtains high-quality articles related to the learning interest feature words, and returns them; Step S203: obtaining high-quality articles related to the learning interest feature words obtained through ChatGPT through an interface, and highlighting the related articles.

6. The method for assisting high school English reading comprehension learning according to claim 5, characterized in that: In step S203, the relevant articles are highlighted, which specifically includes the following steps: Step S203a: Obtain key words that must be tested in high school English and import them into the database. Then, traverse the English words in the article and match them one by one with the words in the database. The matched words are highlighted in red in the article. A link is added and clicked to open a window showing the analysis of the word. Step S203b: Obtain the key phrases required for high school English exams and import them into the database. Then, traverse the English phrases in the article and match them one by one with the phrases in the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the analysis of the phrase when clicked. Step S203c: Obtain CET-4 words, CET-6 words, and IELTS words and import them into different databases respectively. Traverse the English words in the article and match them with different databases according to the selected difficulty mode. Highlight the matched words in blue in the article. Add a link and click it to open a window showing the analysis of this word.

7. A device for assisting high school English reading comprehension learning, characterized in that: Specifically, it includes the following modules: A sentence recognition and parsing module is used to obtain evaluation scores, classify learning modes according to the evaluation scores, and then recognize and parse words and sentences of different difficulty modes; A highlighting processing module is used to obtain learning interest feature words, obtain high-quality related articles by accessing the ChatGPT interface of the learning interest feature words, and highlight the related articles; The collection processing module is used to collect important words and article content and their analysis; The sentence recognition and parsing module specifically includes the following submodules: The first submodule of sentence recognition and parsing is used to segment and extract the main body of the article and the main body of the title based on the uploaded reading comprehension image or text, combined with OCR recognition technology; The second submodule of sentence recognition and parsing is used to traverse the English words in the article and match them one by one with the key high school English test words pre-imported into the database. The matched words are highlighted in red in the article, and a link is added to open a window showing the analysis of the word when clicked; The third submodule, sentence recognition and parsing, is used to traverse the English phrases in the article and match them one by one with the key phrases required for high school English exams that have been pre-imported into the database. The matched phrases are highlighted in yellow in the article and a link is added to open a window showing the parsing of the phrase when clicked. The fourth submodule, sentence recognition and parsing, is used to traverse the English words in the article and match them with a pre-set database of different difficulty levels according to the difficulty mode selected by the user. The matched words are highlighted in blue in the article and a link is added. Clicking it will open a window showing the analysis of the word. The fifth submodule of sentence recognition and parsing is used to match the extracted title content text with the sentence text of the article based on similarity, and underline the one or two sentences with the highest similarity; The sixth submodule of sentence recognition and parsing is used to segment all sentences in the article by periods, match long and difficult sentences based on grammatical analysis and word count, and highlight them in italics.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for assisting high school English reading comprehension learning is implemented as described in any one of claims 1 to 6.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for assisting high school English reading comprehension learning as described in any one of claims 1 to 6 is implemented.

Images