Search

CN-121981105-A - Text reading cognition analysis method and system

CN121981105ACN 121981105 ACN121981105 ACN 121981105ACN-121981105-A

Abstract

The invention discloses a text reading cognition analysis method and a system, and belongs to the technical field of artificial intelligence education. The method realizes the technical breakthrough from grammar surface layer analysis to cognitive deep analysis by constructing a macroscopic-mesoscopic-microscopic three-layer analysis architecture. The method comprises the steps of identifying text types based on a static text model library, carrying out structural analysis by adopting a three-layer architecture, wherein a macroscopic layer extracts a chapter logic framework and generates a memory anchoring thinking body, a mesoscopic layer analyzes paragraph functions and writing methods, a microscopic layer quantitatively analyzes logic relations among sentences, and generates a logic chain visual result and provides interactive training guidance. Compared with the prior art, the invention breaks through the technical limitation that the traditional text analysis can only process the surface language features, and solves the technical problems of deep analysis of a text logic structure, recognition of cultural thinking differences and personalized training generation through the innovative three-layer analysis architecture and memory anchoring thinking body technology.

Inventors

  • WEI PENGCHEN

Assignees

  • 北京市优谛科技有限公司

Dates

Publication Date
20260505
Application Date
20251129

Claims (13)

  1. 1. A text reading cognitive analysis method, comprising the steps of: -receiving input text; -identifying text types based on a static corpus of text models; -structural parsing of text using a macro-mesoscopic-microscopic three-layer analysis architecture; -generating a logical chain and a visual result of the memory anchored mind body; -providing an interactive training guidance based on the analysis result.
  2. 2. The method of claim 1, wherein the macro layer analysis comprises: -parsing the chapter global logic structure; -generating a memory anchored mind body at chapter level.
  3. 3. The method of claim 1, wherein the mesoscopic layer analysis comprises: -identifying a paragraph function type; -analyzing logical join relationships between paragraphs.
  4. 4. The method of claim 1, wherein the microlayer analysis comprises: -quantitatively analyzing the strength of logical relations between sentences; -identifying the method of the repair and its logical function.
  5. 5. The method of claim 1, wherein the static corpus of cultural relics model comprises a structural feature matrix and a thought pattern feature vector of a plurality of cultural relics.
  6. 6. The method of claim 1, wherein the logical chain visualization results support a dual mode display of an analysis mode and a training mode.
  7. 7. The method of claim 1, wherein the interactive training guidance comprises a personalized training regimen generation based on logical weak link identification.
  8. 8. The method according to claim 1, wherein the construction of the memory anchored mind body comprises: -analyzing the chapter-to-fix relation based on the fix structure theory, extracting a logical framework; -identifying core concepts in the text and associations between the concepts; -creating a corresponding visual symbology according to the logical relationship type.
  9. 9. The method of claim 1, wherein the evaluation of the strength of the logical relationship is calculated by the following feature weights: -intensity weights of logical connectives; -semantic similarity assessment; Context consistency analysis.
  10. 10. The method of claim 1, wherein the visual rendering employs a hierarchical optimization strategy to ensure smooth presentation of complex logical structures on a common computing device.
  11. 11. A text reading cognitive analysis system implementing the method of any of claims 1-10, comprising: -a text input and pre-processing module for receiving, washing and structuring input text; -a genre recognition engine for invoking a static genre model library to recognize text types; -a multi-layer analysis engine for performing macro-mesoscopic-microscopic three-layer analysis; -a visualization generation module for generating a visualization of the logic chain and the memory anchored mind body; -a training guidance module for providing interactive training guidance.
  12. 12. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any of claims 1-10.
  13. 13. A text reading cognitive analysis device, comprising: -a memory for storing a computer program according to claim 12; -a processor for executing the computer program to implement the method according to any of claims 1-10.

Description

Text reading cognition analysis method and system 1. Technical field The invention belongs to the technical field of intersection of artificial intelligence education technology and natural language processing, in particular relates to a text reading cognition analysis method and a system, and especially relates to an intelligent text cognition analysis scheme which is based on a macroscopic-mesoscopic-microscopic three-layer analysis architecture and can realize logic chain structural analysis and memory anchoring visualization. The invention specifically covers the following technical directions: 1. Intelligent text analysis, which relates to deep learning and application of graphic neural network in text structure analysis 2. Cognitive calculation, including logic relation quantification, thinking mode modeling, cognitive load assessment and other techniques 3. Educational artificial intelligence covering personalized training generation, adaptive learning path planning and other applications 4. Visual man-machine interaction, including technical fields such as logic chain visualization and interactive exploration interfaces The technical scheme of the invention mainly solves the technical defects of the traditional text analysis system in the aspects of logic deep analysis, thinking pattern recognition and cognitive training generation, and belongs to the innovative application of artificial intelligence technology in the field of education and cognition. 2. Background art The prior text analysis technology has the following technical bottlenecks in the education application scene: 1. Defects of insufficient analysis depth Current mainstream text analysis tools (e.g., stanford parsers, SPACY, etc.) are based primarily on dependency and component syntactic analysis techniques, the technical limitations of which are presented in: Syntax structure capable of handling only single sentence level, lack of logical relationship recognition capability across sentences The inability to build a macroscopic logical framework of text, which makes it difficult for the learner to grasp the overall arguments of the article-the analysis results stay in superficial language features, failing to reveal deep thinking organization patterns of the text 2. Analysis blank of cultural thinking difference In a cross-cultural language learning scenario, the prior art has obvious shortcomings: -lack of a model for quantitative analysis of differences in the patterns of the chinese thinking (e.g. chinese "spiral" logic vs english "rectilinear" logic) -inability to identify cultural thinking features embodied in text, resulting in difficulty for learners to understand the expressed logic in different cultural contexts-structural identification method of existing systems (e.g. CN 113657125A) without considering the influence of cultural factors, limited applicability 3. Problem of disjoint training and analysis Text analysis systems of mainstream learning platforms (e.g., ape coaching, work groups, etc.) on the market suffer from training deficiencies: The generated training content is disjointed with the deep logic structure of the text, so that the effective improvement of the thinking ability cannot be realized, the training scheme stays in the shallow stage (memory and understanding) of the Brucella cognitive classification, and high-order thinking training such as analysis, evaluation, creation and the like is lacked Failure to dynamically adjust the training difficulty based on the text analysis results, violation of the "recent development" education theory 4. Inherent limitations of technical architecture The technical implementation mode of the existing system has fundamental constraint: Rule-based methods have limited coverage and are difficult to handle complex linguistic phenomena The poor interpretability of statistical learning methods (such as LDA models), the inability to provide clear logic analysis paths-the "black box" nature of deep learning models (BERT, GPT, etc.) results in non-traceable, unexplainable analysis processes 5. Lack of visual expression Existing visualization schemes (e.g., dendrograms, mind maps) have expression limitations: Dynamic building process without clear presentation of logic chains It is difficult to express the association of multi-level text structures (macroscopic-mesoscopic-microscopic) Lack of interactive logic exploration functionality, limiting the deep learning experience Aiming at the technical defects, the invention provides a text reading cognition analysis scheme based on a three-layer analysis framework, and realizes deep analysis breakthrough from a grammar level to a thinking level. "Technical problem": "more specifically defined", The logic level is clearer and tighter, "Professional depth": "more technical direction", "Criticizing pertinence" more accurate and powerful " 3. Summary of the invention First, the technical problem to be solved The invention aims to solve the