An artificial intelligence english corpus quick matching system for teaching training purposes

Abstract

The application discloses an artificial intelligence English corpus rapid matching system for teaching training, which comprises, sequentially connected, a corpus preprocessing and multi-view index construction module, an input understanding and intention recognition module, a mixed recall and rearrangement module, a semantic clustering and knowledge unit generation module, and a correlation degree calculation and decomposition display module; the correlation degree calculation and decomposition display module is connected with an exercise generation and evaluation module and a learning portrait and recommendation module; and the corpus preprocessing and multi-view index construction module and the semantic clustering and knowledge unit generation module are connected with a subword alignment and word family aggregation module. The application realizes personalized adaptation based on a learning portrait, can provide targeted corpus, knowledge units and exercise recommendations according to the learning level, weak knowledge points and other characteristics of different users, adapts to the needs of different learning groups, and improves the efficiency and quality of English teaching training.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and English teaching technology, and in particular to an AI-based rapid matching system for English corpora used in teaching and training. Background Technology

[0002] In English teaching and training, corpora serve as core supporting resources, and their matching efficiency and adaptability directly affect teaching quality and learning outcomes. Existing English corpus systems suffer from the following shortcomings: First, inadequate corpus preprocessing and a lack of multi-dimensional indexing lead to slow matching speed and low accuracy. Second, they only support a single input format, failing to meet diverse user input needs. Third, their retrieval modes are limited, hindering the coordinated adaptation of basic, semantic, and personalized retrieval. Fourth, they lack a complete teaching loop design, failing to effectively combine corpus matching, knowledge learning, practice assessment, and personalized recommendations. Fifth, personalized adaptation is insufficient, failing to provide targeted corpus and learning resource recommendations based on users' learning levels, weak knowledge points, and other characteristics. Sixth, they lack support for sub-word alignment and word family aggregation, resulting in semantic ambiguity and incomplete knowledge unit extraction during corpus matching.

[0003] To address the aforementioned issues, there is an urgent need to design an AI-powered English corpus rapid matching system that can overcome the shortcomings of existing technologies, achieve rapid and accurate corpus matching, support multiple input formats, adapt to multi-layered retrieval modes, form a teaching closed loop, and realize personalized adaptation, so as to meet the intelligent needs of English teaching and training. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing an artificial intelligence-based rapid matching system for English corpora used in teaching and training.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A rapid matching system for an AI-powered English corpus for teaching and training purposes includes, in sequence, a corpus preprocessing and multi-view index construction module, an input understanding and intent recognition module, a hybrid recall and rearrangement module, a semantic clustering and knowledge unit generation module, and a correlation calculation and decomposition display module. The correlation calculation and decomposition display module is connected to a practice generation and evaluation module and a learning profile and recommendation module. The corpus preprocessing and multi-view index construction module and the semantic clustering and knowledge unit generation module are respectively connected to a sub-word alignment and word family aggregation module.

[0006] Preferably, the corpus preprocessing and multi-view index construction module is used to clean, deduplicatize, segment, tag parts of speech and standardize the original English corpus to generate standardized corpus; and to construct a multi-view index based on the standardized corpus, wherein the multi-view index includes at least a semantic view index, a grammatical view index and a teaching scenario view index, to support subsequent multi-dimensional fast retrieval.

[0007] Preferably, the input understanding and intent recognition module is used to receive multi-form input information from users, perform semantic parsing, format conversion and intent recognition on the input information, and determine the user's retrieval needs, learning needs and teaching and training needs; the multi-form input includes at least text input, voice input, image input and handwriting input, and the intent recognition includes at least retrieval intent, practice need intent, knowledge point learning intent and evaluation need intent.

[0008] Preferably, the hybrid recall and reordering module is used to obtain candidate English corpora based on user needs identified by the input understanding and intent recognition module, combined with the multi-view index constructed by the corpus preprocessing and multi-view index construction module, using a hybrid recall method that combines keyword recall, semantic recall, and knowledge recall; and reorders the candidate English corpora based on semantic similarity, teaching suitability, and user need matching degree, outputting a sorted candidate corpus set.

[0009] Preferably, the semantic clustering and knowledge unit generation module is used to perform semantic clustering on the candidate corpus set output by the hybrid recall and rearrangement module, and divide the corpus groups with similar semantics; and combine the sub-word alignment results and word family aggregation results output by the sub-word alignment and word family aggregation module to extract core knowledge points from the corpus groups and generate standardized English knowledge units, wherein the knowledge units include at least word units, phrase units, sentence structure units, discourse units and grammar units.

[0010] Preferably, the sub-word alignment and word family aggregation module is used to perform cross-corpus and cross-knowledge unit alignment processing on sub-words in standardized corpora and knowledge units to eliminate sub-word ambiguity; and based on the semantic, word form and part-of-speech features of sub-words, it performs word family aggregation on English vocabulary to form a set of related word families, which is used to optimize the corpus matching accuracy and the completeness of knowledge units.

[0011] Preferably, the correlation calculation and decomposition display module is used to calculate the correlation between user needs and each candidate corpus and each knowledge unit, and uses quantitative values ​​and hierarchical display methods to decompose and display the correlation results, corpus content and knowledge units to the user; the hierarchical display includes at least the display of core matching content, the display of related knowledge points and the display of extended corpus.

[0012] Preferably, the exercise generation and evaluation module is used to automatically generate suitable English teaching and training exercises based on the correlation results and knowledge units output by the correlation calculation and decomposition display module, combined with the user's learning needs. The exercise types include at least vocabulary exercises, sentence pattern exercises, translation exercises, reading exercises, and writing exercises. At the same time, it automatically evaluates the exercises completed by the user, outputs evaluation results, error analysis, and improvement suggestions, and realizes a closed loop of exercise-evaluation-feedback.

[0013] Preferably, the learning profile and recommendation module is used to construct a personalized learning profile for the user based on the evaluation results output by the practice generation and evaluation module, the user's search history, learning history and needs preferences. The learning profile includes at least the learning level, weak knowledge points, learning pace and learning preferences. Based on the learning profile and combined with the correlation calculation results, the module recommends suitable corpora, knowledge units and practice content for the user to achieve personalized adaptation.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In this invention, the corpus standardization and multi-dimensional index construction are achieved through the corpus preprocessing and multi-view index construction module. Combined with the hybrid recall and reordering module, the speed and accuracy of corpus matching are greatly improved. 2. In this invention, a complete teaching loop is formed by combining the practice generation and evaluation module, the learning profile and recommendation module, and the correlation calculation and decomposition display module, thereby realizing integrated teaching support of "learning-practice-evaluation-feedback-recommendation". 3. In this invention, by setting up the sub-word alignment and word family aggregation modules, semantic ambiguity of sub-words is effectively eliminated, the integrity of knowledge units is improved, and the accuracy of corpus matching is enhanced; 4. In this invention, personalized adaptation is achieved based on learning profiles. It can provide targeted corpora, knowledge units and practice recommendations according to the learning level, weak knowledge points and other characteristics of different users, adapt to the needs of different learning groups and improve the efficiency and quality of English teaching and training. Attached Figure Description

[0015] Figure 1 This is a logic block diagram of an artificial intelligence-based English corpus fast matching system for teaching and training purposes proposed in this invention.

[0016] The diagram shows: 1. Corpus preprocessing and multi-view index construction module; 2. Input understanding and intent recognition module; 3. Hybrid recall and rearrangement module; 4. Semantic clustering and knowledge unit generation module; 5. Relevance calculation and decomposition display module; 6. Exercise generation and evaluation module; 7. Learning profile and recommendation module; 8. Sub-word alignment and word family aggregation module. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Reference Figure 1 A rapid matching system for an AI-powered English corpus for teaching and training purposes includes, in sequence, a corpus preprocessing and multi-view index construction module 1, an input understanding and intent recognition module 2, a hybrid recall and rearrangement module 3, a semantic clustering and knowledge unit generation module 4, and a correlation calculation and decomposition display module 5. The correlation calculation and decomposition display module 5 is connected to a practice generation and evaluation module 6 and a learning profile and recommendation module 7. The corpus preprocessing and multi-view index construction module 1 and the semantic clustering and knowledge unit generation module 4 are respectively connected to a sub-word alignment and word family aggregation module 8.

[0019] The corpus preprocessing and multi-view index construction module 1 is used to clean, deduplicatize, segment, tag parts of speech and standardize the original English corpus to generate standardized corpus; and to build a multi-view index based on the standardized corpus. The multi-view index includes at least a semantic view index, a grammatical view index and a teaching scenario view index to support subsequent multi-dimensional fast retrieval.

[0020] The input understanding and intent recognition module 2 is used to receive multi-form input information from users, perform semantic parsing, format conversion and intent recognition on the input information, and determine the user's retrieval needs, learning needs and teaching and training needs; multi-form input includes at least text input, voice input, image input and handwriting input, and intent recognition includes at least retrieval intent, practice need intent, knowledge point learning intent and evaluation need intent.

[0021] The hybrid recall and reordering module 3 is used to obtain candidate English corpora based on the user needs identified by the input understanding and intent recognition module 2, combined with the multi-view index constructed by the corpus preprocessing and multi-view index construction module 1. It adopts a hybrid recall method that combines keyword recall, semantic recall and knowledge recall. Then, it reorders the candidate English corpora based on semantic similarity, teaching suitability and user need matching degree, and outputs the sorted candidate corpus set.

[0022] The semantic clustering and knowledge unit generation module 4 is used to perform semantic clustering on the candidate corpus set output by the hybrid recall and rearrangement module 3, and divide the corpus groups with similar semantics; and combined with the sub-word alignment and word family aggregation results output by the sub-word alignment and word family aggregation module 8, core knowledge points are extracted from the corpus groups to generate standardized English knowledge units. The knowledge units include at least word units, phrase units, sentence structure units, discourse units and grammar units.

[0023] The sub-word alignment and word family aggregation module 8 is used to perform cross-corpus and cross-knowledge unit alignment processing on sub-words in standardized corpora and knowledge units to eliminate sub-word ambiguity; and based on the semantic, word form and part-of-speech features of sub-words, it performs word family aggregation on English words to form a set of related word families, which is used to optimize the corpus matching accuracy and the completeness of knowledge units.

[0024] The correlation calculation and decomposition display module 5 is used to calculate the correlation between user needs and each candidate corpus and each knowledge unit. It uses quantitative values ​​and hierarchical display methods to decompose and display the correlation results, corpus content and knowledge units to the user. The hierarchical display includes at least the display of core matching content, the display of related knowledge points and the display of extended corpus.

[0025] The practice generation and evaluation module 6 is used to automatically generate suitable English teaching training exercises based on the correlation results and knowledge units output by the correlation calculation and decomposition display module 5, combined with the user's learning needs. The exercise types include at least vocabulary exercises, sentence pattern exercises, translation exercises, reading exercises, and writing exercises. At the same time, it automatically evaluates the exercises completed by the user, outputs evaluation results, error analysis, and improvement suggestions, and realizes a closed loop of practice-evaluation-feedback.

[0026] The learning profile and recommendation module 7 is used to construct a personalized learning profile for users based on the evaluation results output by the practice generation and evaluation module 6, the user's search history, learning history and needs preferences. The learning profile includes at least the learning level, weak knowledge points, learning pace and learning preferences. Based on the learning profile and combined with the relevance calculation results, the module recommends suitable corpora, knowledge units and practice content for users to achieve personalized adaptation.

[0027] Working principle: Through corpus preprocessing and multi-view index construction module 1, corpus standardization and multi-dimensional index construction are achieved. Combined with hybrid recall and reordering module 3, the speed and accuracy of corpus matching are greatly improved. Through practice generation and evaluation module 6, learning profile and recommendation module 7, combined with relevance calculation and decomposition display module 5, a complete teaching loop is formed, realizing integrated teaching support of "learning-practice-evaluation-feedback-recommendation". The setting of sub-word alignment and word family aggregation module 8 effectively eliminates semantic ambiguity of sub-words, improves the integrity of knowledge units, and enhances the accuracy of corpus matching. Based on learning profile, personalized adaptation is achieved, which can provide targeted corpus, knowledge units and practice recommendations according to the learning level, weak knowledge points and other characteristics of different users, adapt to the needs of different learning groups, and improve the efficiency and quality of English teaching and training.

[0028] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A rapid matching system for an artificial intelligence English corpus for teaching and training purposes, characterized in that, The system includes a corpus preprocessing and multi-view index construction module (1), an input understanding and intent recognition module (2), a hybrid recall and rearrangement module (3), a semantic clustering and knowledge unit generation module (4), and a correlation calculation and decomposition display module (5), which are connected in sequence. The correlation calculation and decomposition display module (5) is connected to a practice generation and evaluation module (6) and a learning profile and recommendation module (7). The corpus preprocessing and multi-view index construction module (1) and the semantic clustering and knowledge unit generation module (4) are respectively connected to a sub-word alignment and word family aggregation module (8).

2. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 1, characterized in that, The corpus preprocessing and multi-view index construction module (1) is used to clean, deduplicatize, segment, tag parts of speech and standardize the original English corpus to generate standardized corpus; and to construct a multi-view index based on the standardized corpus. The multi-view index includes at least a semantic view index, a grammatical view index and a teaching scenario view index to support subsequent multi-dimensional fast retrieval.

3. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 2, characterized in that, The input understanding and intent recognition module (2) is used to receive multi-form input information from users, perform semantic parsing, format conversion and intent recognition on the input information, and determine the user's retrieval needs, learning needs and teaching and training needs; the multi-form input includes at least text input, voice input, image input and handwriting input, and the intent recognition includes at least retrieval intent, practice need intent, knowledge point learning intent and evaluation need intent.

4. The AI-powered rapid matching system for English corpus used in teaching and training, as described in claim 3, is characterized in that... The hybrid recall and rearrangement module (3) is used to obtain candidate English corpora based on the user needs identified by the input understanding and intent recognition module (2) and the multi-view index constructed by the corpus preprocessing and multi-view index construction module (1). It adopts a hybrid recall method that combines keyword recall, semantic recall and knowledge recall. Based on semantic similarity, teaching suitability and user needs matching degree, the candidate English corpora are rearranged and the sorted candidate corpus set is output.

5. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 4, characterized in that, The semantic clustering and knowledge unit generation module (4) is used to perform semantic clustering on the candidate corpus set output by the hybrid recall and rearrangement module (3) and divide the corpus groups with similar semantics; and combined with the sub-word alignment and word family aggregation results output by the sub-word alignment and word family aggregation module (8), core knowledge points are extracted from the corpus groups to generate standardized English knowledge units, which include at least word units, phrase units, sentence structure units, discourse units and grammar units.

6. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 5, characterized in that, The sub-word alignment and word family aggregation module (8) is used to perform cross-corpus and cross-knowledge unit alignment processing on sub-words in standardized corpus and knowledge units to eliminate sub-word ambiguity; and based on the semantic, word form and part-of-speech features of sub-words, it performs word family aggregation on English words to form a set of related word families, which is used to optimize the corpus matching accuracy and the completeness of knowledge units.

7. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 6, characterized in that, The correlation calculation and decomposition display module (5) is used to calculate the correlation between user needs and each candidate corpus and each knowledge unit. It uses quantitative values ​​and hierarchical display methods to decompose and display the correlation results, corpus content and knowledge units to the user. The hierarchical display includes at least the display of core matching content, the display of related knowledge points and the display of extended corpus.

8. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 7, characterized in that, The exercise generation and evaluation module (6) is used to automatically generate suitable English teaching training exercises based on the correlation results and knowledge units output by the correlation calculation and decomposition display module (5) and the user's learning needs. The types of exercises include at least vocabulary exercises, sentence pattern exercises, translation exercises, reading exercises and writing exercises. At the same time, the exercises completed by the user are automatically evaluated, and evaluation results, error analysis and improvement suggestions are output to realize the closed loop of exercise-evaluation-feedback.

9. The AI-powered English corpus rapid matching system for teaching and training purposes according to claim 8, characterized in that, The learning profile and recommendation module (7) is used to construct a personalized learning profile of a user based on the evaluation results output by the practice generation and evaluation module (6), the user's search history, learning history and needs preferences. The learning profile includes at least the learning level, weak knowledge points, learning pace and learning preferences. Based on the learning profile and the results of correlation calculation, the system recommends suitable corpora, knowledge units and practice content to users to achieve personalized adaptation.

Images