113 results about "Template entry" patented technology

The invention discloses a multi-agent cooperative cue word engineering optimization method, system and device and a medium, and relates to the technical field of cue word engineering, the method comprises the following steps: after initially receiving a user input task, starting a cue word optimization process; an initial cue word template is created through a cue word generation agent, historical optimal templates are retrieved from a knowledge base for fusion, and candidate cue word templates are generated. And the analysis agent inputs the candidate template into the second large language model to generate a candidate answer. And the evaluation agent performs multi-dimensional evaluation on the answers and gives scores and defect analysis. And according to an analysis result, updating the cue word to generate a candidate template of the intelligent agent. And arranging the intelligent agent to cyclically execute the steps until the stop condition is met, and outputting the optimal cue word template. And finally, storing the optimal template into a knowledge base and establishing a semantic index to support subsequent retrieval reuse. According to the method, manual intervention is not needed, and a complete automatic process of generating, evaluating and optimizing the cue word is realized.

The invention relates to an intelligent bidding document compiling method, system and equipment based on a large language model. The method comprises the following steps: constructing a bid invitation knowledge graph; obtaining project demand information to determine demand fields, calling a knowledge graph association interface to match bid invitation file modules and terms for each demand field, and generating a demand field-association module mapping table; according to the mapping table and bid invitation industry specifications, a generation prompt template embedded with a dynamic nesting slot position and a knowledge graph rule is constructed, and a pre-trained bid invitation big language model is input to output a bid invitation clause draft; traversing the knowledge graph according to a demand change field to calculate a change influence radius, constructing an update prompt template in combination with a slot association rule and a constraint condition, inputting a large model to generate revised content only aiming at the demand field and an association module, replacing a corresponding position of a draft, and performing secondary verification on logic consistency through the knowledge graph, and obtaining the updated complete bid invitation file. By adopting the method, intelligence, precision and high efficiency of bid invitation file compiling can be realized.

The invention discloses a network threat knowledge automatic extraction method, electronic equipment and a storage medium, and the method comprises the following steps executed by a computer hardware system: collecting threat intelligence data related to an APT organization from a multi-source network security text, and processing the threat intelligence data to generate a standardized corpus; using the pre-training sentence vector model to generate semantic embedding for a corpus input text and a manual annotation example library text, and retrieving similar examples to construct an ICL prompt template; inputting a large language model subjected to LoRA fine tuning, and extracting structured triples of multiple types of entities and semantic relationships; generating standardized entity nodes and updated relation information by adopting semantic aggregation; and constructing an APT organization network threat intelligence knowledge graph and outputting a structured file. The method provides key technical support for APT attack tracing, threat situation awareness and automatic security policy generation.

The invention discloses a natural language database query method and system based on large-model two-channel context enhancement, and the method comprises the steps: carrying out the parallel execution of a business link and a mode link on a query statement, and respectively outputting business context information and mode link information; carrying out vectorization fusion on the two types of information to generate combined context information; generating a query intention vector based on the combined context information, and outputting a complexity label of the query statement through a complexity calculation model; and generating a prompt template according to complexity label differentiation, inputting a large model to generate an SQL statement, and executing query. Through a double-channel cooperation mechanism, the problems of semantic ambiguity, database structure complexity, poor field adaptability and the like in natural language query are solved, query accuracy and response efficiency are improved, and the method is suitable for multi-industry complex database scenes.

The invention discloses a steel ladle hoisting safety monitoring and early warning method, system and equipment and a medium. The method comprises the following steps: acquiring double-side images in a steel ladle hoisting process; inputting the bilateral images into a multi-target detection model, and outputting bounding boxes, category labels and confidence coefficients of the plurality of key targets; obtaining a local cutting image of each key target according to the bilateral images and the bounding box, the category label and the confidence of each key target, and generating a structured text in combination with the bounding box position, the relative pixel distance and the category label of each key target; inputting the local cutting image, the structured text and the prompt word template of each key target into a multi-modal big language model, and outputting a behavior label, a risk level, a reason, a recommendation action and an instruction state of the current steel ladle hoisting scene through multi-modal reasoning; early warning is conducted according to the recommended action, and the steel ladle hoisting equipment is controlled according to the instruction state. According to the invention, intelligent monitoring and linkage early warning in the steel ladle hoisting process can be realized.

The invention discloses a public opinion home address standardization method and system, and the method comprises the following steps: S1, obtaining and processing multi-source address original data to obtain an address data block, processing the data block based on a pre-trained text embedding model to obtain a corresponding semantic vector, and storing the address data block and the corresponding semantic vector, establishing a multi-source address knowledge base; s2, processing public opinion text data through the text embedding model to obtain a query vector of the public opinion text data; s3, querying the multi-source address knowledge base based on the query vector to obtain a plurality of address data blocks related to the query vector; and S4, constructing a cue word template based on the public opinion text data and the address data blocks, and inputting the cue word template into a large language model to perform public opinion attribution address standardization analysis.

The invention provides a knowledge editing method based on multilingual semantic retrieval, and belongs to the technical field of natural language process.The method achieves cross-lingual knowledge updating through the two stages of multilingual knowledge retrieval and context editing and comprises the steps that firstly, a multilingual retrieval model based on XLM-R is used for mapping queries and knowledge base entries to a shared semantic space, and then the shared semantic space is used for conducting context editing; semantic correlation is judged through a classifier, and target language knowledge is retrieved; and splicing the retrieval result and the query in a zero sample or small sample mode to generate a prompt template, and inputting the prompt template into a large language model to complete editing. The method is characterized in that monolingual limitation is broken through, collaborative updating of 12 languages is supported, a retrieval-editing decoupling architecture is adopted to achieve model irrelevant adaptation, and retrieval and context learning are fused to improve editing accuracy. The method can efficiently correct multi-language large model fact errors, is suitable for scenes such as search engines and intelligent customer services, avoids the cost of full-model retraining, and remarkably improves the efficiency and accuracy of cross-language knowledge updating.

The invention discloses a model robustness automatic evaluation method based on large model driving and related equipment. The method comprises the following steps: inputting a task template of a tested model into a dynamic Prompt engine module to obtain a sample generation instruction; generating an adversarial sample according to the sample generation instruction through a sample generation and verification module, and performing model evaluation on the adversarial sample; if the confrontation sample passes the model verification, extracting a high-dimensional multi-modal feedback vector of the tested model through a multi-modal feedback analysis module; a reusable semantic instruction is generated through a strategy control module according to the high-dimensional multi-modal feedback vector so as to optimize a generation strategy; and if the confrontation sample does not pass the model verification, performing effect evaluation on the confrontation sample through an evaluation and interpretation generation module and generating an interpretability evaluation report. According to the method, progressive generation of the adversarial sample and exponential-level improvement of the model robustness evaluation efficiency can be realized, a clear model improvement direction is provided for developers, and the method can be widely applied to the technical field of computers.

The invention relates to the technical field of knowledge maps, and provides a screening type dynamic classification self-adaptive relation extraction method, device and equipment, and the method comprises the following steps: defining an entity type list, inputting a text by a user according to a cue word template in a labeling format, judging the length of the text, and extracting the entity type list according to the length of the text; selecting a corresponding strategy according to the text length to generate a candidate entity pair; the association strength of each candidate entity pair is calculated by combining static corpus statistics and a dynamic semantic hybrid scoring model, and the candidate entity pairs with low association are filtered; constructing a dynamic prompt request LLM to generate a natural language abstract describing the subject entity and the object entity, and classifying the abstract into a predefined relationship type to obtain a structured triple; and importing the triple into a graph database to construct the knowledge graph. According to the method, the knowledge extraction efficiency is improved, and the problems of calculation redundancy and high cost are solved.

The invention relates to the technical field of artificial intelligence, and discloses an AI dialogue system and method based on multi-modal data fusion, and the system comprises a processing module which is used for obtaining multi-modal user dialogue data, carrying out the preprocessing of the multi-modal user dialogue data, and obtaining a user emotion feature set; the type module is used for determining a user emotion type according to the user emotion feature set and setting a prompt word template according to the user emotion type; the model module is used for filling the multimodal user dialogue data into a cue word template, and inputting the cue word template filled with the data into a large language model to obtain a model answer result; and the correction module is used for acquiring the preset number of multi-modal user dialogue data, determining a user emotion trend according to the preset number of multi-modal user dialogue data, inputting the user emotion trend into the cue word template, and generating a model correction answer result. According to the method, the multi-modal dialogue data of the user can be accurately recognized, and the accuracy and efficiency of multi-modal intention understanding of the user are effectively improved.

The invention discloses a network threat behavior reasoning method and system based on large model retrieval enhancement, and the method comprises the following steps: a computer system executes the following steps: determining a query target and an associated scene or organization by receiving a network threat analysis query request of a user; retrieving related nodes and relationships from the threat intelligence knowledge graph, and generating an initial reasoning sub-graph; calling a GraphRAG framework, and carrying out semantic matching retrieval on a subgraph node context and an external threat intelligence knowledge base to obtain supplementary evidence; constructing a comprehensive prompt template, inputting the large language model subjected to LoRA fine tuning, and generating a conformity ATTamp; reasoning a conclusion of the CK framework; and finally, supplementing the newly added nodes and the relationship to the knowledge graph, and outputting a structured file or a visual result. According to the method, retrieval enhancement and large model reasoning technologies are combined, the problems of cross-document attack chain restoration and complex threat behavior logical reasoning are effectively solved, and the method has remarkable application value in the aspects of threat intelligence tracing, behavior prediction and defense decision assistance.

The invention provides a data processing method and device, equipment, a storage medium and a program product, and relates to the technical field of computers. The method comprises the steps that first question information input by a user is subjected to word segmentation processing and then subjected to semantic matching with a pre-configured lexicon, and second question information is determined; the second question information is normalized and vectorized and then subjected to similarity calculation with a pre-configured table library, an entity meeting a similarity threshold value and a standardized query template are determined, the pre-configured table library comprises a plurality of tables, and the standardized query template comprises a structure query language and a skeleton question; inputting the first question information, the entity and the standardized query template into a first large language model, and determining multi-table query information; and inputting the first question information and the multi-table query information into a second large language model, and determining response information. Based on the large language model, the cross-multi-table question can be effectively answered by deeply analyzing the user question and combining with the table library generated by multi-table preprocessing, and the problem that cross-multi-table answers cannot be effectively processed due to the fact that table question answering driven by the large language model mainly aims at a single table is solved.

1. The name of the design product: electronic device research and development experiment management graphical user interface. 2. The use of the design product: an electronic device. 3. The design points of the design product: the interface content in the graphical user interface. 4. The picture or photo that best shows the design points: front view. 5. The product carrier is the conventional design, and the rear view, left view, right view, top view, and bottom view are omitted. 6. The use of the graphical user interface: graphical user interface for research and development experiment management system. 7. Other circumstances that need to be explained: the front view is the task list interface of the research and development experiment management system. Clicking "+" experiment design" in the interface enters the change state diagram 1 task addition interface. The interface pops up the task information setting pop-up window. Fill in the relevant information in the pop-up window. Click the "experiment template" addition icon (the icon with a magnifying glass on the right side of the experiment template input box) to enter the change state diagram 2 task template addition interface. The interface pops up the experiment template selection pop-up window. The left side of the pop-up window is the template display area, and the middle is the process node display area. After selecting the required template in the template display area, the corresponding task node is generated in the process node display area. Click any node in the process node display area to enter the change state diagram 3 node information interface. The right side of the interface is the node editing area. Click "add group" in the area to enter the change state diagram 4 group setting interface. The interface pops up the "edit group" pop-up window. After setting the relevant parameters in the pop-up window, click "OK" to enter the change state diagram 5 complete group setting. Click "group design" in the change state diagram 5 under the three small dot icons on the right side of the group design bar to enter the change state diagram 6. The interface pops up the "material calculation" pop-up window. After setting the relevant parameters in the pop-up window, click "generate data" to enter the change state diagram 7. Click "preference settings" in the change state diagram 7 to enter the change state diagram 8. The interface pops up the execution mode setting pop-up window. Click "OK" after selecting "detailed view" in the pop-up window to enter the change state diagram 9 detailed mode experiment execution.

The invention provides a leading-edge technology maturity level interval discriminant analysis method and device, and the method comprises the steps: obtaining a technology text which is pre-collected for a to-be-judged technology, and constructing an original data set; preprocessing the original data set to obtain an updated data set; inputting each piece of data in the updated data set into a preset data extraction model, wherein the data extraction model extracts triple data consisting of a technical direction, a technical dynamic state and a technical index in each piece of data in the updated data set; determining a transition level of the to-be-judged technology based on matching of a preset transition keyword library and the updated data set; technical information including the triple data and the transition level is supplemented into a preset structured JSON cue word template, the filled structured JSON cue word template is input into a preset large language model, and the large language model outputs a maturity judgment level.

The invention provides an interaction method, electronic equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring an input interaction context; the interaction context and a preset system cue word template are input into a large language model, at least one candidate PEFT module is obtained, and the system cue word template is used for indicating the large language model to select and output the candidate PEFT module from a plurality of preset PEFT modules according to the interaction context; determining a target PEFT module according to the candidate PEFT module, and loading the target PEFT module to the large language model to obtain a large language model after the module is loaded; and inputting the interaction context into the large language model loaded with the module to obtain an interaction response. According to the method and the device, automatic selection and loading of the PEFT module can be realized, and the interaction performance and expression of a large language model for the current interaction context can be improved.

The embodiment of the invention provides a firewall policy generation method and device, a medium and a program product, and relates to the technical field of policy generation. The method comprises the following steps: acquiring current business demand description information; extracting at least one target document fragment related to the business demand description information from a preset document knowledge base; extracting at least one target configuration sample matched with the business demand description information from a preset configuration sample library; constructing a prompt template based on the business demand description information, the target document fragment and the target configuration sample; and inputting the prompt template into a preset strategy generation model to obtain a strategy configuration scheme output by the strategy generation model. According to the embodiment of the invention, the normative document and the configuration sample are matched according to the business demand, the prompt template is constructed based on the business demand information, the normative document and the configuration sample, and the model is input to generate the strategy configuration scheme, so that the firewall strategy configuration efficiency and accuracy can be effectively improved.

The invention relates to a conference summary generation method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a conference summary generation request for a target conference, the conference summary generation request comprising conference data of the target conference and summary generation demand information, and the summary generation demand information comprising a preset scoring rule; according to the conference data and a preset scoring rule, determining at least one candidate conference summary generation model from a plurality of preset conference summary generation models; determining a first conference summary generation model from the at least one candidate conference summary generation model; and inputting the conference data and the conference summary template corresponding to the conference data into a first conference summary generation model to obtain a target conference summary based on the first conference summary generation model. According to the method, the optimal model matched with the conference data can be quickly and accurately determined for the conference data of various scenes, manual debugging is not needed in the process, the labor cost can be reduced, and the processing efficiency is improved.

The application provides a procurement document generation and evaluation method based on process mining analysis, relates to the technical field of enterprise management, and comprises the following steps: collecting event log data in an enterprise procurement system, constructing a procurement process hypergraph model through a process hypergraph discovery algorithm, and generating a variation correlation topological structure; constructing a self-adaptive rule mapping network according to the procurement process hypergraph model and the variation correlation topological structure, inputting specific procurement demand data into the evolved template by utilizing the network to dynamically evolve the procurement document template, filling the file content by applying a semantic intervention filling mechanism, and generating a complete procurement document; the process hypergraph discovery algorithm is introduced to construct the procurement process hypergraph model, so that complex correlations such as multi-party interaction, group decision and high-order dependence in the procurement activity can be effectively captured, and the expression capacity of the process model is significantly improved compared with the traditional Petri net or simple directed graph model.

The objective is to provide technology that allows for easy input and management of ESG-related information. [Solution] An information processing system comprising: a storage unit for storing ESG-related information of business entities constituting the upstream or downstream of a supply chain; a setting unit for setting a template into which a predetermined business entity can input ESG-related information; wherein the template has multiple items set; the setting unit can set necessary items from the set multiple items for which information input from a predetermined business entity is required; and a presentation unit that can present the set necessary items in a manner that allows the predetermined business entity to input them.

The invention discloses a CAD question and answer method, system and device based on a large model and a medium, and relates to the field of computer aided design, and the method comprises the steps: receiving a target file uploaded by a user and a corresponding first question; inputting the first question and a preset structured prompt template into a preset first large model, and generating a tool calling sequence in combination with a preset calculation tool set; wherein the structured prompt template is used for constraining a decomposition process of the first big model to the first problem; according to the tool calling sequence, corresponding tool functions are selected and called from the calculation tool set in sequence, the target file is processed, and an operation result and corresponding structured evidence are obtained; and through the first large model, performing answer generation according to the operation result and the structured evidence to obtain a first answer of the first question. By implementing the method and the device, the accuracy of providing answers by a large model in a CAD (Computer Aided Design) workflow can be improved.

Artificial Intelligence-based System and Method for Generating and Deploying Applications via a Cognitive Environment System and method for generating and deploying applications via cognitive environment is provided. Digital PODs comprising digital AI personas are triggered in response to system prompt for evaluating first set of features from input data using LLMs based on a predefined function to generate first outcome. First outcome is generated based on external data retrieved from external tools. First outcome is compared with input data and comparison is inputted to LLMs as prompt template. LLMs evaluate whether to proceed with next step in a series of steps associated with subsequent predefined functions to arrive at final outcome. Next step involves generating second outcome by extracting second set of features from first outcome and applying subsequent predefined function over first outcome. Series of steps are implemented till LLMs iteratively determine that the final outcome is comparable to desired outcome.

The application discloses a scientific research review portrait construction method based on a large language model and related equipment, which can be applied to the technical field of natural language processing and artificial intelligence. The application divides the preprocessed structured data set based on a preset label system to obtain data subsets corresponding to different labels; then inputs the data subsets and hierarchical prompt word templates into a large language model to obtain candidate labels of each dimension, labels support data associated with the candidate labels in the data subsets, and then performs confidence verification on the candidate labels labeled with the support data to obtain target portrait labels, so that high-quality, structured and multi-dimensional portraits can be obtained, the accuracy of the assignment of scientific research review experts is improved, the target portrait labels and the support data are stored in a traceable manner, and reliable portrait data can be provided in an actual application scenario, and the accuracy of the assignment of scientific research review experts is further improved.

The present disclosure provides a whole-process electrical construction process monitoring method and system, which comprises the following steps: formulating a data entry template, importing and storing equipment data with a position number and material data without a position number by using the data entry template, editing the equipment data with a position number, and encoding the material data without a position number to be recognized and collected to the corresponding process; building the engineering quantity process of all electrical construction, and generating an electrical system diagram; uploading a single file or a compressed package, filling in the corresponding figure number when uploading the process template entry, quickly opening the drawing based on the figure number, and displaying the engineering progress.

This invention provides a method, apparatus, device, and medium for expanding e-commerce multi-category product data. The method includes: S10, acquiring product category information; S20, constructing a prompt word template based on the product category information; S30, inputting the prompt word template into a Qwen-Image model to generate product images; S40, performing multi-dimensional scoring on the generated product images and outputting a comprehensive score; S50, filtering based on the comprehensive score to obtain high-quality images as expansion data; through a multi-category adaptation mechanism and a multi-dimensional scoring mechanism, high-quality and diversified e-commerce product image expansion is achieved.

The invention provides a business process simulation method and device, equipment and a storage medium, and the method comprises the steps: receiving a simulation condition sent by a user side, and obtaining the log data of a business process; the simulation conditions are input into a preset embedded model for vectorization processing, and simulation condition codes are obtained; screening case feature vectors of one or more target business cases matched with the simulation condition codes from the reference case knowledge base according to a preset similarity threshold value and a preset similarity calculation method; according to the simulation condition, selecting a corresponding simulation cue word template from a preset cue word template library; inputting the case feature vector of each target business case and the simulation cue word template into a preset large language model for structured processing to obtain simulation structure data; and modifying the log data of the business process according to the simulation structure data to obtain a final simulation result corresponding to the business process, so that the deviation between the final simulation result and an actual result is reduced.

The invention relates to the technical field of data space, and discloses a multi-agent-based data space adaptive data delivery method and device, and the method comprises the steps: inputting an unstructured data product API interface document and a data APP interface document into a large language model agent, and obtaining a structured data product API calling template and a data APP receiving template; inputting the data product API calling template and the data APP receiving template into a large language model agent to obtain a mapping conversion file; inputting the data into a large language model agent by using a contract to obtain a constraint rule file; inputting the mapping conversion file and the constraint rule file into a large language model agent to obtain a delivery process control chart; and acquiring data from the data product API according to the delivery process control chart by utilizing a large language model agent, and injecting the acquired data into the data APP. According to the invention, the automatic adaptation capability of the data delivery process in the data space can be improved, and data circulation and utilization are promoted.

The invention relates to the technical field of database retrieval, in particular to a patent retrieval type intelligent construction method and device and a storage medium. The method provided by the invention comprises the following steps: acquiring a natural language of a patent queried by a user; inputting the natural language into a large language model for intention understanding to obtain intention information; the intention information at least comprises a retrieval technology intention, a time range intention, a region intention and an application subject intention; converting the intention information into an sql query statement; executing the sql query statement in a pre-established data table to obtain key information; generating structured information according to the key information; inputting the structured information and the mode information of the patent database into a cue word template, and inputting the cue word template into a large language model to obtain a retrieval formula output by the large language model; and performing retrieval in a patent database based on the retrieval formula. According to the method, an accurate retrieval formula is constructed by utilizing a large language model.

The invention discloses a nuclear power log information processing method and device.The method comprises the steps that a first interface is displayed, and the first interface comprises a template control; in response to a trigger operation for the template control, displaying a pre-configured log entry template, the log entry template being obtained according to historical nuclear power log information; in response to a first selection operation, displaying a target log entry template, the target log entry template being one of the log entry templates; obtaining handover information input based on the target log entry template; the nuclear power log information is generated according to the handover information, in the steps, the pre-configured log entry template is adopted, a user can select the target log entry template and input the handover information based on the target log entry template, the obtained nuclear power log information is more standard, and information description is more accurate.

The application provides a large model-based multi-user mixed interaction data processing method and system. The method comprises the following steps: acquiring a natural language request set input by multiple users concurrently; performing high-dimensional semantic vector coding on the user request by using a Transform architecture to generate a user input semantic vector, obtaining a theme center set and a semantic mapping matrix constructed in terms of semantic similarity; constructing a behavior unit set; selecting a historical context fragment of each user based on the behavior unit set and a state vector set to obtain a selected context fragment set and a corresponding fragment source mapping table; clustering and matching a template based on the selected context fragment set in combination with the user request, inputting a large model, and outputting a personalized response of multiple users. The application establishes a processing framework with a user-theme behavior pair as the minimum decision granularity.