CN-121387281-B - Personalized virtual space automatic construction method based on user intention

CN121387281BCN 121387281 BCN121387281 BCN 121387281BCN-121387281-B

Abstract

The invention relates to the technical field of virtual scene construction and discloses a personalized virtual space automatic construction method based on user intention, which comprises the steps of receiving a natural language query text input by a user, calling a preset model context protocol platform and constructing a multi-element database corresponding to a virtual space construction intention vector; extracting key information entity corresponding to the virtual space construction intention vector from the multi-element database, calculating the semantic similarity between the virtual space construction intention vector and the key information entity, establishing a virtual space configuration set corresponding to the multi-element database to generate an interactable preview view corresponding to the virtual space construction intention vector, and generating a personalized virtual space through the virtual space configuration set and the interactable preview view based on the multi-element database. The invention can enable the depth of the virtual space construction process to be matched with the real-time intention of the user, and improves the user adaptation flexibility of the virtual space.

Inventors

Luo Sanbi

Assignees

北京中科院软件中心有限公司

Dates

Publication Date: 20260512
Application Date: 20251027

Claims (9)

1. An automatic construction method of a personalized virtual space based on user intention, which is characterized by comprising the following steps: Receiving a natural language query text input by a user to generate a virtual space construction intention vector of the user, calling a preset model context protocol platform to construct a multi-element database corresponding to the virtual space construction intention vector, wherein the multi-element database comprises an environment element library, an object instance library and an interaction logic library, and calling the preset model context protocol platform to construct the multi-element database corresponding to the virtual space construction intention vector, and the method comprises the following steps: Analyzing the space environment elements, the entity object elements and the interaction behavior elements in the virtual space construction intention vector; The data flow module based on the model context protocol platform respectively converts the space environment element, the entity object element and the interaction behavior element into standardized data units to obtain environment data units, object data units and behavior data units; Extracting an environment semantic tag in the environment data unit, a function semantic tag in the object data unit and a behavior semantic tag in the behavior data unit based on a tool calling module of the model context protocol platform; Constructing a cross-modal association relation set of the environment data unit, the object data unit and the behavior data unit according to the environment semantic tag, the functional semantic tag and the behavior semantic tag; Calculating the data call association weight of each data unit in the environment data unit, the object data unit and the behavior data unit based on the association strength value in the cross-modal association relation set; Generating a multi-source data scheduling priority strategy of the cross-modal association relation set according to the data call association weight; Constructing a multi-element database corresponding to the virtual space construction intention vector based on the multi-source data scheduling priority strategy and the cross-modal association relation set; Configuring a multidimensional attribute meta tag of the multi-element database to establish an inverted index and a feature vector index of the multi-element database, and extracting key information entities corresponding to the virtual space construction intention vector from the multi-element database according to the inverted index and the feature vector index; calculating the semantic similarity of the virtual space construction intention vector and the key information entity, and generating a semantic embedded vector and a space topology constraint condition corresponding to the virtual space construction intention vector based on the semantic similarity; Establishing a virtual space configuration set corresponding to the multi-element database according to the semantic embedded vector and the space topology constraint condition, and analyzing corresponding entity data in the multi-element database based on the virtual space configuration set; and generating an interactable preview view corresponding to the virtual space construction intention vector by combining the entity data, the semantic embedded vector and the space topology constraint condition, and generating a personalized virtual space of the user through the virtual space configuration set and the interactable preview view based on the multi-element database.
2. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the receiving the natural language query text input by the user to generate the virtual space construction intention vector of the user comprises: Determining the type of the virtual space required by the user through the natural language query text; Extracting a space entity object set, a space relation operator and a space attribute modifier in the natural language query text; constructing a space topological relation map of the required virtual space type according to the space entity object set, the space relation operator and the space attribute modifier; Based on the space topological relation map, analyzing a multi-granularity layout constraint condition of the required virtual space type; And generating a virtual space construction intention vector of the user according to the space topological relation graph and the multi-granularity layout constraint condition.
3. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the establishing the virtual space configuration set corresponding to the multi-element database according to the semantic embedded vector and the space topology constraint condition comprises: analyzing an environment element library, an object instance library and an interaction logic library in the multi-element database; extracting space style semantics and functional intention semantics in the semantic embedded vector; Screening candidate environmental elements in the environmental element library based on the spatial style semantics; According to the functional intention semantics, searching candidate object examples from the object example library; Decomposing the space topology constraint condition into an environment layout constraint condition and an object relation constraint condition; Generating an environment space layout scheme corresponding to the candidate environment elements and an environment element instance based on the environment layout constraint condition; outputting an object instance set corresponding to the candidate object instance and space pose parameters thereof according to the object relationship constraint condition and the environment space layout scheme; Based on the functional intention semantics, matching relevant candidate interaction logic from the interaction logic library; Calculating compatibility coefficients of the candidate interaction logic and the object relation constraint condition; screening target interaction logic from the candidate interaction logic according to the compatibility coefficient; And fusing the environment space layout scheme, the environment element examples, the object example set, the space pose parameters and the target interaction logic to establish a virtual space configuration set corresponding to the multi-element database.
4. The method for automatically constructing a personalized virtual space based on user intention according to claim 3, wherein the calculating compatibility coefficient of the candidate interaction logic and the object relation constraint condition comprises: Analyzing the interaction space requirement of the candidate interaction logic, and calculating the space matching degree between the interaction space requirement and the object relation constraint condition; Identifying the strictness grade of the object relation constraint condition, and determining a constraint intensity coefficient corresponding to the strictness grade; Analyzing the functional intention labels of the candidate interaction logic, and calculating the context semantic matching degree of the functional intention labels and the current virtual space scene; Identifying real-time demand urgency corresponding to the computation complexity characteristics corresponding to the candidate interaction logic and the object relation constraint conditions; Comprehensively extracting system load influence factors of the candidate interaction logic and the object relation constraint condition according to the calculation complexity characteristics and the real-time demand urgency; And combining the space matching degree, the constraint intensity coefficient, the context semantic matching degree and the system load influence factor, wherein the compatibility coefficient of the candidate interaction logic and the object relation constraint condition is calculated by the following formula: Wherein, the The coefficient of compatibility is represented by a coefficient of compatibility, Representing the coefficient of the intensity of the constraint, A weight balance coefficient representing the spatial matching degree and the context semantic matching degree, The degree of spatial matching is represented by the degree of spatial matching, Representing the degree of semantic matching of the context, Representing the system load influencing factor(s), And the weight balance coefficient of the system load influence factor is represented.
5. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the configuring the multi-dimensional attribute meta tag of the multi-element database comprises: extracting core data characteristics of an environment element library, an object instance library and an interaction logic library in the multi-element database; defining a general dimension tag set of the multi-element database based on the core data characteristics; Analyzing the spatial position attribute, the physical quantity type and the time granularity in the environment element library to generate an environment class exclusive meta-tag of the multi-element database; Extracting entity types, functional uses and geometric form features in the object instance library to determine object class exclusive meta tags of the multi-element database; analyzing a logic triggering mode, response timeliness and a scope of the interaction logic library to generate an interaction class exclusive meta tag of the multi-element database; integrating the universal dimension tag set, the environment class exclusive meta tag, the object class exclusive meta tag and the interaction class exclusive meta tag to construct the multidimensional attribute meta tag of the multi-element database.
6. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein configuring the multidimensional attribute meta-tag of the multivariate database to establish the inverted index and the feature vector index of the multivariate database comprises: according to the multidimensional attribute meta-tag, a meta-tag key value pair of the multi-element database is analyzed; Extracting meta tag semantic features of the meta tag key value pairs; Generating a meta tag feature vector set of the multi-element database based on the meta tag semantic features; establishing a feature vector index of the multi-element database through the meta tag feature vector set; Identifying high-frequency keywords and rare keywords in the meta tag key value pair to calculate keyword weights of the meta tag key value pair; Determining the inverted index item weight of the multi-element database through the keyword weight; generating a weighted inverted list of the multi-element database according to the inverted index item weight; And establishing an inverted index of the multi-element database based on the weighted inverted list.
7. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the generating the interactable preview view corresponding to the virtual space construction intention vector by combining the entity data, the semantic embedded vector and the space topology constraint condition comprises: mapping the entity data to a preset Unity engine, and generating an environment element instance, an object instance and an interaction logic instance corresponding to the entity data through an instantiation interface of the Unity engine; Based on the semantic embedded vector, analyzing semantic style preference and functional interaction requirements of the virtual space construction intention vector to adjust multidimensional environment configuration parameters of the environment element examples; generating an object pose configuration scheme of the object instance according to the space topology constraint condition; extracting triggering conditions and response scripts of the interaction logic instance, and carrying out dynamic binding processing on the triggering conditions, the response scripts and specific events of the object instance through an event delegation system of the Unity engine to obtain an interaction behavior logic chain; Integrating the multidimensional environment configuration parameters, the object pose configuration scheme and the interactive behavior logic chain to construct a Unity lightweight real-time rendering scene corresponding to the virtual space construction intention vector; Defining adaptive rendering optimization parameters of the Unity lightweight real-time rendering scene based on the cluster center distance of the semantic embedded vector; And generating an interactable preview view corresponding to the virtual space construction intention vector according to the lightweight real-time rendering scene and the self-adaptive rendering optimization parameter.
8. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the generating a semantic embedded vector corresponding to the virtual space construction intention vector based on the semantic similarity comprises: Screening out candidate key information entities corresponding to the construction intention vectors of the virtual space based on the semantic similarity; constructing a multi-dimensional incidence matrix of the candidate key information entity and the virtual space construction intention vector; extracting principal component feature vectors and corresponding importance weights thereof in the multi-dimensional association matrix; based on the importance weight, carrying out weighted fusion processing on the principal component feature vector to obtain a weighted fusion feature vector; Mapping the weighted fusion feature vector to a low-dimensional semantic space to obtain a low-dimensional semantic embedded vector corresponding to the candidate key information entity; calculating the cluster center distance and the distribution density of the low-dimensional semantic embedded vector in the low-dimensional semantic space; and according to the cluster center distance and the distribution density, executing normalization processing of the low-dimensional semantic embedding vector to obtain the semantic embedding vector corresponding to the virtual space construction intention vector.
9. The method for automatically constructing a personalized virtual space based on user intention according to claim 1, wherein the generating a spatial topology constraint condition corresponding to the virtual space construction intention vector based on the semantic similarity comprises: Extracting a spatial relationship predicate in the virtual space construction intention vector, and determining a spatial constraint intensity coefficient corresponding to the spatial relationship predicate; Calculating semantic-feature coupling weights among different spatial relationship predicates based on the semantic similarity; Establishing a priority ordering rule of the spatial relationship predicates according to the spatial constraint intensity coefficient and the semantic-feature coupling weight; based on the priority ordering rule, determining a constraint logic combination mode corresponding to the virtual space construction intention vector and an execution sequence thereof; defining a constraint tolerance interval corresponding to the virtual space construction intention vector according to the space constraint intensity coefficient; and integrating the constraint logic combination mode, the execution sequence and the constraint tolerance interval to generate a space topology constraint condition corresponding to the virtual space construction intention vector.

Description

Personalized virtual space automatic construction method based on user intention Technical Field The invention relates to an automatic personalized virtual space construction method based on user intention, and belongs to the technical field of virtual scene construction. Background The personalized virtual space is a dynamic digital scene system built around the intention of a user by means of a Unity engine technology, is used for precisely matching core requirements of the user in different scenes such as entertainment, work and study, and aims to provide highly customized, immersive and interactive smooth virtual experience for the user. However, the conventional virtual space is mainly constructed based on a predefined script and closed static resources, for example, the position and interaction rule of the scene element are set through fixed code logic, and although the method can ensure the stability of the initial operation of the scene, the interaction mechanism is completely dependent on a preset flow, so that the real-time intention of the user cannot be understood, the real-time intention of the user is difficult to integrate with external dynamic data (such as ecological monitoring information), and finally the flexibility of the user operation and the adaptability of the scene are seriously insufficient. Therefore, a solution is needed to enable the depth of the virtual space construction process to be matched with the real-time intention of the user, so that the user adaptation flexibility of the virtual space is improved. Disclosure of Invention The invention provides an automatic personalized virtual space construction method based on user intention, which mainly aims to enable the depth of a virtual space construction process to be matched with the real-time intention of a user and improve the user adaptation flexibility of the virtual space. In order to achieve the above object, the present invention provides a method for automatically constructing a personalized virtual space based on user intention, comprising: Receiving a natural language query text input by a user, generating a virtual space construction intention vector of the user, calling a preset model context protocol platform, and constructing a multi-element database corresponding to the virtual space construction intention vector, wherein the multi-element database comprises an environment element library, an object instance library and an interaction logic library; Configuring a multidimensional attribute meta tag of the multi-element database to establish an inverted index and a feature vector index of the multi-element database, and extracting key information entities corresponding to the virtual space construction intention vector from the multi-element database according to the inverted index and the feature vector index; calculating the semantic similarity of the virtual space construction intention vector and the key information entity, and generating a semantic embedded vector and a space topology constraint condition corresponding to the virtual space construction intention vector based on the semantic similarity; Establishing a virtual space configuration set corresponding to the multi-element database according to the semantic embedded vector and the space topology constraint condition, and analyzing corresponding entity data in the multi-element database based on the virtual space configuration set; and generating an interactable preview view corresponding to the virtual space construction intention vector by combining the entity data, the semantic embedded vector and the space topology constraint condition, and generating a personalized virtual space of the user through the virtual space configuration set and the interactable preview view based on the multi-element database. Optionally, the receiving the natural language query text input by the user to generate the virtual space construction intent vector of the user includes: Determining the type of the virtual space required by the user through the natural language query text; Extracting a space entity object set, a space relation operator and a space attribute modifier in the natural language query text; constructing a space topological relation map of the required virtual space type according to the space entity object set, the space relation operator and the space attribute modifier; Based on the space topological relation map, analyzing a multi-granularity layout constraint condition of the required virtual space type; And generating a virtual space construction intention vector of the user according to the space topological relation graph and the multi-granularity layout constraint condition. Optionally, the calling a preset model context protocol platform to construct a multivariate database corresponding to the virtual space construction intention vector includes: Analyzing the space environment elements, the entity object elements and the interaction beh