# US 2025/0156643 A1 · System for semantic analysis and automatic solution of mathematical application problem

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0156643 A1
- Document ID: US20250156643A1
- Status: Published application
- Kind: A1
- Application number: 18/638,592
- Filed: 2024-04-17
- Published: 2025-05-15
- Earliest claimed: 2023-11-10
- Assignees: Bei Jing Bda Network & Information Co., Ltd
- Applicants: Bei Jing Bda Network &Information Co.,Ltd
- Inventors: Ping ZHU; Pohua LV; Xuetao JIANG
- Web page: https://patentgrep.com/patent/US20250156643A1

## CPC

- G06F 40/30: Semantic analysis
- G06F 40/242: Dictionaries
- G06F 40/247: Thesauruses; Synonyms
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present disclosure provides a system for semantic analysis and automatic solution of a mathematical application problem. The system includes: a semantic framework mode matching module, configured to: perform mode matching on a clause vocabulary sequence with a semantic framework, to form local semantic information of the mathematical application problem; a scenario semantic analysis module, configured to: form, based on a local feature vocabulary string of a scenario, a global semantic feature vocabulary string of the scenario; a data meta-variable naming module, configured to: generate a global semantic name of a data meta-variable based on a local name of a variable; a module for explicitly expressing a computation relationship, configured to: explicitly express an explicit computation rule and an implicit computation rule between data meta-variables, to construct a dynamic semantic circle; and a module for implementing a machine thinking mechanism, configured to solve the questionable data meta-variable.

## Claims

### Claim 1

A system for semantic analysis and automatic solution of a mathematical application problem, comprising:
a semantic framework mode matching module, configured to: generate a clause vocabulary sequence of a mathematical application problem based on an input text, perform mode matching on the clause vocabulary sequence with a semantic framework, to form local semantic information of the mathematical application problem, wherein the local semantic information represents local semantics between clause vocabularies; and a semantic framework matched with the clause vocabulary sequence expresses the local semantics;
a scenario semantic analysis module, configured to: form a local feature vocabulary string of each scenario based on a feature vocabulary that represents scenario semantics and that is in a labeled clause vocabulary sequence, and form a global semantic feature vocabulary string of each scenario through semantic inheritance and semantic overloading of the local feature vocabulary string of each scenario, wherein the scenario semantics comprises global semantics and local semantics;
a data meta-variable naming module, configured to: perform global semantic naming on a data meta-variable of a clause based on the local feature vocabulary string of each scenario through a synonym and a technology of inheritance and overloading of global semantics between clauses;
a module for explicitly expressing a computation relationship, configured to: explicitly express an explicit computation rule and an implicit computation rule between data meta-variables, to construct a dynamic semantic circle, wherein the data meta-variable comprises an explicit known-value data meta-variable, a questionable data meta-variable, and an implicit intermediate data meta-variable; and the dynamic semantic circle comprises an explicitly-expressed computation rule, an implicitly-expressed computation rule, and a solution formula; and
a module for implementing a machine thinking mechanism, configured to solve the questionable data meta-variable based on the explicit known-value data meta-variable, the implicit intermediate data meta-variable, and the dynamic semantic circle of the computation rule.

### Claim 2 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, further comprising a general knowledge base, wherein the general knowledge base comprises a dictionary for a thinking machine, a library of semantic frameworks, a library of synonyms and near-synonyms, a knowledge map of subordinate relationships of concepts, a library of classifications of attribute relationships of concepts, a library of commonly used formulas, a library of computation rules of formulas, and a library of semantically constrained relationships.

### Claim 3 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, further comprising a semantic labeling module, wherein
the semantic labeling module is configured to: perform, based on word segmentation and part of speech tagging, semantic labeling on a clause vocabulary sequence that is not successfully matched with the semantic framework, and add a semantic framework of the clause vocabulary sequence that is not successfully matched with the semantic framework to a library of semantic frameworks in a general knowledge base.

### Claim 4 (depends on 3)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 3, wherein the semantic labeling module specifically comprises:
a semantic dimension determining unit, configured to: perform, based on word segmentation and part of speech tagging, semantic labeling on the clause vocabulary sequence that is not successfully matched with the semantic framework, and determine a to-be-labeled semantic dimension for the mathematical application problem, wherein the to-be-labeled semantic dimension comprises a subordinate relationship of concepts, a conditional computation relationship, a computation relationship, an inequality relationship, an attribute relationship of concepts, a constant, and a variable that are implied in a text;
a definition unit, configured to: based on the to-be-labeled semantic dimension, define a semantic labeling grammar, and determine various labels, system-reserved keywords, and system-reserved global variable names; and
an updating unit, configured to label a semantic framework of the clause through the semantic labeling grammar, and store a semantic framework of the clause that passes through correctness validation into the library of semantic frameworks.

### Claim 5 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, wherein the scenario semantic analysis module specifically comprises a scenario division and labeling submodule; and
the scenario division and labeling submodule comprises:
a sorting unit, configured to: check input clause vocabulary sequences in sequence, and sort scenario semantic identifiers of each clause from large to small based on a category;
an analysis unit for a first clause scenario, configured to: if a first clause does not have an explicit scenario semantic identifier that belongs to a subordinate vocabulary of a knowledge map of subordinate relationships of concepts and does not have any relative participle of a text for problems, based on a scenario semantic identifier sequence, supplement the first clause with a default initial scenario semantic identifier, start a new scenario, and continuously analyze a next clause, wherein the any relationship comprises a parallel relationship or a subordinate relationship;
a first analysis unit for a current clause scenario, configured to: if a locally explicated semantic identifier string of the current clause scenario is NULL, or if a local semantic identifier is the same as a global semantic identifier of a last clause scenario, assume that the current clause scenario continues the last clause scenario, and continuously analyze a next clause;
a second analysis unit for a current clause scenario, configured to: if an explicated semantic identifier string of the current clause scenario comprises a subordinate vocabulary of a knowledge map of subordinate relationships of concepts, start a new scenario, and continuously analyze a next clause;
a third analysis unit for a current clause scenario, configured to: if a global semantic identifier of a last clause scenario comprises a subordinate vocabulary of a knowledge map of subordinate relationships of concepts, continue a last scenario, and continuously analyze a next clause;
a fourth analysis unit for a current clause scenario, configured to: if a current clause comprises a data meta-variable whose local name is the same as a local name of a data meta-variable of a last clause, and a global semantic identifier of a data meta-variable of the current clause is the same as a global semantic identifier of the data meta-variable of the last clause, supplement a default scenario semantic identifier, and assume that the current clause starts a new scenario, and continuously analyze a next clause;
a fifth analysis unit for a current clause scenario, configured to: if a global semantic identifier of the current clause scenario is different from a global semantic identifier of a last clause scenario, assume that a current clause starts a new scenario, and continuously analyze a next clause; and
a sixth analysis unit for a current clause scenario, configured to: continue the last clause scenario, and continuously analyze a next clause.

### Claim 6 (depends on 5)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 5, wherein the scenario semantic analysis module specifically comprises: a submodule for inheritance and overloading of a scenario semantic identifier; and
the submodule for inheritance and overloading of a scenario semantic identifier comprises:
a global semantic analysis unit for a first clause scenario, configured to: take a local semantic identifier of the first clause scenario as a global semantic identifier of the first clause scenario, and continuously analyze a next clause;
a global semantic analysis unit for a current clause scenario, configured to: if the current clause scenario continues the last clause scenario, determine that global semantics of the current clause scenario is the same as global semantics of the last clause scenario, and continuously analyze a next clause; and
a unit for reverse comparison between identifiers, configured to: perform, in sequence, reverse comparison between a first identifier of local semantics of a current scenario and a global semantic identifier string of the last scenario, constitute a global semantic identifier of the current scenario, and continuously analyze a next clause.

### Claim 7 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, wherein the data meta-variable naming module specifically comprises:
a first feature comparison unit, configured to: perform, in sequence, reverse comparison between a first feature of a local name of a variable of a data meta-variable of a backbone statement and a global semantic identifier of a current scenario based on the local semantic feature vocabulary string;
a first naming unit for a global name of a variable, configured to: if a first feature is the same as a global semantic identifier of any current scenario, remove a first feature in a local name of a variable, splice a global semantic identifier string of the current scenario with a local name feature string of the variable, constitute a global name of the variable, and continuously analyze a next clause;
a second naming unit for a global name of a variable, configured to: if a first feature and a global semantic identifier of any current scenario are in a parallel relationship in a knowledge map of subordinate relationships of concepts, remove the global semantic identifier of the current scenario and a global semantic identifier after the global semantic identifier of the current scenario, splice a global semantic identifier string of the current scenario with a local name feature string of a variable, constitute a global name of the variable, and continuously analyze a next clause;
a third naming unit of a global name of a variable, configured to: if a first feature and a global semantic identifier of any current scenario are in a membership in a knowledge map of subordinate relationships of concepts, remove a global semantic identifier after the global semantic identifier of the current scenario, splice a global semantic identifier string of the current scenario with a local name feature string of a variable, constitute a global name of the variable, and continuously analyze a next clause;
a fourth naming unit of a global name of a variable, configured to: if a first feature and a global semantic identifier of a first current scenario are in a membership in a knowledge map of subordinate relationships of concepts through comparison, determine a local name feature string of a variable as a global name of the variable, and continuously analyze a next clause; and
a fifth naming unit of a global name of a variable, if a first feature and a global semantic identifier of a first current scenario are not in a same relationship, a parallel relationship, or a membership through comparison, splice a global semantic identifier string of the scenario with a local name feature string of a variable, constitute a global name of the variable, and continuously analyze a next clause.

### Claim 8 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, wherein the module for explicitly expressing a computation relationship specifically comprises:
a mathematical object recognition unit, configured to: recognize a mathematical object implied in scenario semantics in a semantic framework sequence of the problem clause, wherein the mathematical object comprises a set, a proportion, an interval, a number axis, and coordinates;
a dynamic semantic circle output unit, configured to: output dynamic semantic circles of a computation relationship of data meta-variables and a conditional computation relationship that are default by a mathematical object implied in context of problems;
a first input unit, configured to: input, into the dynamic semantic circle, a computation relationship of data meta-variables that is explicitly expressed by local semantics of a semantic framework sequence;
a second input unit, configured to: input, into the dynamic semantic circle, a computation relationship of data meta-variables that is explicitly expressed in a global semantic framework;
a third input unit, configured to: input, into the dynamic semantic circle, a mathematical general knowledge formula that is matched with the problem;
a fourth input unit, configured to: input, into the dynamic semantic circle, a common mathematical attribute formula that is matched with the problem; and
a derivation unit, configured to: output a mathematical computation relationship that corresponds to an equivalent correlation relationship and that is matched with the problem, and supplement a computation relationship that is derived from a mathematical computation relationship in the dynamic semantic circle.

### Claim 9 (depends on 1)

The system for semantic analysis and automatic solution of a mathematical application problem according to claim 1, wherein the module for implementing a machine thinking mechanism is specifically configured to:
derive and compute, in the dynamic semantic circle, the intermediate data meta-variable and the questionable data meta-variable based on a known-value data meta-variable in a table of data meta-variables, determine whether solving for the problem is completed, and if solving for the questionable data meta-variable is completed, return successfully; and
if solving for all questionable data meta-variables is not completed, set a next variable in the table of data meta-variables as x, derive and compute a quadratic equation in the dynamic semantic circle, solve for x, continue to derive the dynamic semantic circle, determine whether solving for the problem is completed, and if solving for the questionable data meta-variable is completed, return successfully;
if solving for all questionable data meta-variables is not completed, set any two variables in the table of data meta-variables as x and y, perform elimination based on a binary equation, and if elimination is successful, continue to solve for the questionable data meta-variable, or if solving for the questionable data meta-variable is not completed, return successfully; or
if solving for all questionable data meta-variables is not completed and the problem satisfies a constraint condition for a variable feature, change the questionable data meta-variable into a known-value data meta-variable, and continue to solve for the questionable data meta-variable until the constraint condition for the variable feature is satisfied.
