Construction log automatic generation method and system based on working condition keywords

Abstract

The invention discloses an automatic construction log generation method and system based on working condition keywords, and the method comprises the following steps: S1, creating project groups, and constructing a working condition keyword library K in the corresponding groups according to the project types; s2, group messages in the group are collected, working condition keywords are recognized in real time through an AI matching function, and the types of the group messages comprise at least one of texts, voices and images; when the keyword matching degree meets the condition that f (m, ki) is greater than or equal to theta, the log content is marked as effective log content, and an effective log information set is output; s3, performing context-aware optimization, structured analysis and standardization processing on the effective information set acquired in the step S2, and generating a corresponding construction log; and S4, distributing the construction logs obtained in the step S3 to related parties according to a preset list, and meanwhile, filing the construction logs to a cloud database according to dates, types and user identities. The invention aims to greatly improve the construction log recording efficiency.

Description

Technical Field

[0001] This invention relates to the field of engineering construction technology, and in particular to a method and system for automatically generating construction logs based on working condition keywords. Background Technology

[0002] Currently, construction logs are legally mandated engineering record documents filled out daily by construction units during the construction process, forming a core component of engineering technical data. In building construction, construction logs are crucial documents recording daily work content, progress, quality, and safety issues, playing a key role in project management, quality traceability, and final acceptance. However, traditional construction log recording methods rely primarily on manual entry, leading to the following problems: low information recording efficiency, as construction personnel must manually compile daily work content, easily resulting in omissions or non-standard descriptions, leading to incomplete log information; fragmented and difficult-to-integrate data, as construction information is often scattered across WeChat groups, verbal reports, or paper records, lacking unified structured storage and making it difficult to quickly aggregate and generate standardized logs; reliance on manual classification and analysis, requiring managers to spend significant time filtering effective information and manually categorizing it into different reports, resulting in low efficiency and a high risk of errors; and difficulty in data traceability, as modifications to paper logs or ordinary electronic documents make it impossible to trace the original record, and multiple versions of documents are stored separately, making it difficult to confirm the recorder and modification time in case of disputes.

[0003] Existing patent application CN118363933A discloses an automatic log generation system based on big data analysis, including the following steps: acquiring monitored image data; performing segmented analysis on the monitored image data to obtain multiple monitoring sub-images; pre-training a deconvolutional neural network model based on multiple image samples; inputting each monitoring sub-image into the deconvolutional neural network model for adaptive key feature extraction to obtain a key feature image corresponding to each monitoring sub-image; performing keyword log monitoring on the key feature image corresponding to each monitoring log sub-image to obtain the log monitoring result. While this invention can reduce the effectiveness of manual intervention, it still relies on manual verification, resulting in low log generation efficiency. Summary of the Invention

[0004] The main objective of this invention is to provide a method and system for automatically generating construction logs based on working condition keywords, aiming to solve the technical problems of low efficiency and non-standard information in existing traditional construction log recording.

[0005] To achieve the above objectives, this invention provides a method for automatically generating construction logs based on working condition keywords, the method comprising the following steps:

[0006] Step S1: Create project groups and build a working condition keyword library K in the corresponding groups according to project type;

[0007] Step S2: Collect group messages within the group and identify working condition keywords in real time using an AI matching function. Group message types include at least one of text, voice, and image. When the keyword matching degree satisfies f(m,ki)≥θ, it is marked as valid log content, and a set of valid log information is output, where m is the group message, ki is the i-th keyword in the working condition keyword library K, and θ is a preset threshold.

[0008] Step S3: Perform context-aware optimization, structured analysis and normalization on the effective information set collected in step S2 to generate the corresponding construction log.

[0009] Optionally, in step S1, the step of constructing the working condition keyword library K includes:

[0010] Define a set of keywords based on project type;

[0011] The intelligent matching algorithm associates groups with preset keyword types;

[0012] Users can add, delete, and modify keywords through the interactive interface.

[0013] Optionally, in step S2, if a single message matches multiple keywords, the matching keywords are dynamically prioritized according to the arbitration mechanism rules. The arbitration mechanism rules include dynamic weight ranking, conflict detection and resolution strategies, and the dynamic weight ranking priority is before the conflict detection and resolution strategies.

[0014] Optionally, in step S2, when keyword matching failure is detected or the arbitration mechanism rules cannot make an effective decision, a manual review process is triggered.

[0015] Optionally, in step S2, if the detected effective content is a risky keyword and the number of occurrences is higher than the preset frequency, an early warning mechanism is automatically triggered, and a Gantt chart deviation analysis report is generated.

[0016] Optionally, in step S3, the structured analysis and normalization steps include:

[0017] Generate a unique object ID for the original data OI;

[0018] Based on semantic analysis and matching of predefined template IDs, a standardized language is formed. This allows for tracing from template IDs to original records via forward indexing, while reverse indexing supports querying the processing trajectory of object IDs. Here, represents the i-th standardized log record generated; represents the original input data; represents the unique identifier; represents the template ID, referring to the predefined standardized language template in the system; and represents the context information.

[0019] Optionally, in step S3, context-aware optimization includes knowledge graph construction and context feature extraction.

[0020] Optionally, the method further includes: step S4, distributing the construction logs obtained in step S3 to relevant parties according to a preset list, and archiving the construction logs to a cloud database by date, type and user identity.

[0021] Furthermore, to achieve the above objectives, the present invention also provides an automatic construction log generation system based on working condition keywords, the system comprising:

[0022] Group Management Module: Used to create project groups and build a working condition keyword library K in the corresponding group according to the project type;

[0023] Multimodal parsing module: Communicatively connected to the group management module, used to collect group messages within the group and match them using the AI ​​matching function f(m,k). i = AI_Match(m,k) i )∈[0,1] Real-time identification of working condition keywords, group message type includes at least one of text, voice and image; when the keyword matching degree satisfies f(m,ki)≥θ, it is marked as valid log content and the valid log information set is output, where m is the group message, ki is the i-th keyword in the working condition keyword library K, and θ is the preset threshold;

[0024] Log generation engine: connected to the multimodal parsing module, including:

[0025] Context-aware units are used to optimize raw data using a construction domain knowledge graph.

[0026] The template matching unit calls standardized templates based on the log type selected by the user, transforming colloquial expressions into standard language.

[0027] The multi-format output unit is used to convert preprocessed datasets into construction logs according to user-specified chart types;

[0028] Distribution and Traceability Module: Connected to the output of the log generation engine, this module automatically distributes construction logs to relevant parties based on a preset list and project organizational structure; archives logs to a cloud database by date, type, and user identity, and records them using the operation log function R(O)={l1,l2,...,l N It enables full-cycle traceability, with each record containing a timestamp, operation type, and content.

[0029] In addition, to achieve the above objectives, the present invention also provides a project management efficiency evaluation system based on AI system interaction data. The system includes a memory, a processor, and an automatic construction log generation program based on working condition keywords stored in the memory and executable on the processor. When the automatic construction log generation program based on working condition keywords is executed by the processor, it implements the steps of the automatic construction log generation method based on working condition keywords as described above.

[0030] In addition, to achieve the above objectives, the present invention also provides a computer storage medium storing a construction log automatic generation program based on working condition keywords, wherein when the construction log automatic generation program based on working condition keywords is executed by a processor, it implements the steps of the construction log automatic generation method based on working condition keywords as described above.

[0031] Beneficial effects:

[0032] This invention provides a method and system for automatically generating construction logs based on working condition keywords. Driven by AI, it achieves full automation of the "collection → parsing → generation" process, eliminating manual processing and improving efficiency. The keyword library is dynamically adjusted to ensure terminology conforms to engineering specifications. A knowledge graph is used to convert spoken language into standard engineering language, and the risk prediction module automatically triggers warnings to reduce schedule deviations. Furthermore, the system, through modular collaborative work, forms an intelligent closed loop of "dynamic keyword binding → multimodal parsing → standardized generation → closed-loop traceability," ensuring data traceability and compliance while maintaining high efficiency and standardization, thus strengthening traceability. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating an embodiment of a method for automatically generating construction logs based on working condition keywords according to the present invention.

[0034] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0035] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0036] This invention provides an automatic construction log generation system based on working condition keywords. The system includes a memory, a processor, and an automatic construction log generation program based on working condition keywords stored in the memory and executable on the processor. When the processor executes the automatic construction log generation program based on working condition keywords, it implements the steps of the automatic construction log generation method based on working condition keywords as described below.

[0037] Furthermore, such as Figure 1 As shown, this invention provides a flowchart of a method for automatically generating construction logs based on working condition keywords. The method includes the following steps:

[0038] Step S1: Create project groups and build a working condition keyword library K in the corresponding group according to the type of project.

[0039] Specifically, group chats are created based on project type, and each group chat is bound to preset work condition keywords. The steps for building the work condition keyword library K include: defining a set of keywords based on project type; associating groups with preset keyword types through an intelligent matching algorithm; and adding, deleting, and modifying keywords through an interactive interface. Generally, administrators can expand or adjust the keyword library using AI tools to ensure that terminology is synchronized with industry standards or project requirements.

[0040] In practical applications, the system establishes independent communication groups for different engineering projects and binds working condition keywords to each group. For working condition-related projects, the keyword set includes, but is not limited to, standard terms related to key construction stages such as "safety inspection," "project progress," "concrete pouring," and "material acceptance." The system has a pre-set standardized keyword library, categorized by construction professional fields, including civil construction, installation engineering, and quality inspection. The system uses an intelligent matching algorithm to associate each communication group with its corresponding pre-set keyword type. This matching process comprehensively considers factors such as project type attributes and the roles of group members. After matching, the system generates a preliminary keyword mapping table and pushes it to project managers for confirmation. The system also provides a keyword adjustment interface, allowing project managers to fine-tune the automatically matched keyword set according to actual needs. Adjustments include, but are not limited to, adding project-specific professional terms, deleting inapplicable general terms, and modifying the specific expressions of terms. The system records all adjustment operations and uses them to optimize subsequent automatic matching algorithms.

[0041] The confirmed set of keywords will be used as the standardized keyword library for this group. The system supports version management of the keyword library, which can record adjustments and changes at different times, ensuring the consistency and traceability of the construction log generation standards.

[0042] Step S2: Collect group messages within the group and match them using the AI ​​matching function f(m,k). i = AI_Match(m,k) i)∈[0,1] Real-time identification of working condition keywords, the group message type includes at least one of text, voice and image; when the keyword matching degree satisfies f(m,ki)≥θ, it is marked as valid content and the valid information set M is output, where m is the group message, ki is the i-th keyword in the working condition keyword library K, and θ is the preset threshold.

[0043] Specifically, staff are pre-assigned to corresponding groups, and multimodal work data is collected, including text reports, voice messages, and on-site construction images. AI is used to analyze group messages in real time and identify work-related keywords, intelligently filtering and extracting relevant content as required. Technically, the function expression f(m,k) is employed. i = AI_Match(m,k) i Let )∈[0,1] represent the group message m and the i-th predefined keyword k. i The degree of matching, when keyword k exists. i Make the matching degree f(m,k) i When a message m reaches or exceeds a threshold θ, the system marks it as valid log content. This determination process can be expressed using set theory. K = {k1,k2,...,k} n Let K be the set of keywords. Then, the system extracts key construction data from the messages and performs intelligent aggregation analysis on all keyword-related messages with a matching degree exceeding a threshold.

[0044] Preferably, each working condition keyword is configured with a dynamically adjusted matching threshold parameter θ, which is personalized according to the specific project characteristics, thereby improving the matching accuracy.

[0045] Furthermore, if a single message matches multiple keywords, the matching keywords are dynamically prioritized according to the arbitration mechanism rules. The arbitration mechanism rules include dynamic weighting, conflict detection and resolution strategies, and the dynamic weighting priority is higher than the conflict detection and resolution strategies.

[0046] Specifically, for dynamic keyword weight sorting, a preset keyword weight configuration table (stored in a cloud database) is queried. The weight value w(ki) is dynamically set based on the project impact. Keywords k1, k2, ..., kn are matched in real time and their weight values ​​are extracted and sorted in descending order of weight. j ∣w(k j )≥w(k j+1 ),j=1to n-1}, to represent an instance, message matching - device failure (w=0.9) and schedule delay (w=0.7) → sorting result: device failure > schedule delay.

[0047] Regarding the conflict detection and resolution strategy, the conflict type is first determined: is it a resource conflict (e.g., keywords requiring simultaneous use of the same equipment, such as concrete pouring and equipment maintenance) or a logical conflict (e.g., the contradiction between completed pouring and unqualified inspection)? The resolution strategy is then executed (based on a rule base): if a resource conflict is detected, the "nearest resource scheduling" strategy is executed (using Gantt chart data to allocate alternative equipment); if it is a logical conflict, the "timestamp priority" strategy is executed (adopting the keyword with the latest status). Example: A resource conflict exists between a pump truck malfunction requiring maintenance (equipment failure) and a pouring delay (progress delay) → the system schedules a backup pump truck, marking the equipment failure as a priority.

[0048] Furthermore, when the system detects that the keyword priority arbitration engine cannot make an effective decision, it automatically triggers a manual review process. The arbitration failure message is automatically marked as "pending processing" and pushed to the terminal of the designated responsible person. Through manual annotation of processing results and recording of manual decision-making results, the system updates the weight configuration through reinforcement learning. For example, if a person manually increases the crack detection weight from 0.9 to 1.0 multiple times, the system automatically increases the priority of that keyword. Thus, the arbitration mechanism's priority order—dynamic weight sorting → intelligent conflict resolution → human-machine collaborative optimization—solves efficiency issues (reducing manual intervention), accuracy issues (conflict resolution strategies reduce misjudgments), and self-evolutionary capabilities (a weight learning mechanism based on human feedback, adapting to different engineering scenario requirements) in multi-keyword scenarios.

[0049] At the same time, when the system detects that keyword matching fails, it also automatically triggers a manual review process.

[0050] Furthermore, if the detected content is a risk keyword and its frequency exceeds a preset range, an early warning mechanism is automatically triggered, and a Gantt chart deviation analysis report is generated. Specifically, the system is configured with a risk prediction module to monitor the frequency of risk keywords in engineering communication data in real time. For example, when preset risk keywords such as "delay" and "material shortage" appear frequently, an early warning mechanism is automatically activated. The data association engine intelligently matches the risk warning signals with the progress data of the engineering management system to extract the affected project nodes. The system also includes a visualization analysis unit that generates a dynamic Gantt chart report based on the deviation analysis algorithm, visually presenting the scope and degree of risk impact through color-coded annotations and progress deviation curves.

[0051] Furthermore, a collaborative work group architecture involving multiple users can be established. This architecture integrates all management, technical, and construction personnel related to the construction project into a unified instant messaging platform, forming a complete project communication network. The system supports the creation of multiple specialized groups, which can be flexibly configured according to construction specialties, work areas, or division of responsibilities, ensuring the orderly flow of various construction information.

[0052] Step S3: Perform context-aware optimization, structured analysis and normalization on the effective information set collected in step S2 to generate the corresponding construction log.

[0053] Specifically, the AI ​​performs context-aware optimization, structured analysis, and standardization on the collected chat messages. Based on the log type selected by the user, it activates the corresponding template, organizes the content as required, and generates construction logs according to the output format selected by the user, ensuring that the logs do not contain any XML tags.

[0054] The purpose of context-aware optimization is to enhance AI's deep understanding of construction data and identify implicit information (such as schedule deviations or resource conflicts). Specifically, this includes knowledge graph construction and context feature extraction.

[0055] Among these features, a knowledge graph for the construction field was constructed, endowing the AI ​​system with deep semantic understanding and logical reasoning capabilities. This knowledge graph systematically integrates structured industry knowledge such as construction specifications, schedule standards, resource allocation, and technological logic, and establishes a multi-dimensional semantic association network.

[0056] Contextual feature extraction analyzes the contextual information of group chat records, including timestamps (such as message sending time), user roles (such as construction workers and supervising engineers), and spatial locations. This allows for the deduction of implicit states: that is, predicting potential problems through logical relationships in the knowledge graph (such as "pump truck malfunction → potential delays in progress"). Specifically, by using context-aware algorithms to deeply correlate real-time collected work data with the knowledge graph, it can not only identify explicit information in data such as construction logs and progress reports, but also automatically discover potential problems such as construction progress deviations and resource conflicts based on the logical relationships and constraints in the knowledge graph through an inference engine.

[0057] Structured analysis involves cleaning unstructured data and identifying core operational elements. It filters out irrelevant characters (such as emojis) while retaining project-related data. It also generates intermediate datasets and labels key fields (such as task type and risk level).

[0058] The specific process is as follows: (1) Unique identifier allocation, that is, generating a unique object identifier for each original record in the effective information set obtained in step 2, and supporting full-cycle traceability. (2) Through natural language processing technology, the system performs multi-dimensional intelligent analysis on the original records, including working condition keyword extraction (matching the preset keyword library), semantic similarity calculation (using NLP models to compare the similarity between messages and standard terms (e.g., “finished” → “completed construction”) and context feature analysis (accurately identifying core elements such as construction procedures, equipment status, and quality inspection). Based on the analysis results, the system accurately matches colloquial expressions with standardized templates in the knowledge base to achieve intelligent conversion from unstructured language to standard engineering terms. Example: Original message: “Concrete in Area A was finished today” Analysis result: {Procedure: “Concrete pouring”, Status: “Completed”, Area: “Area A”}.

[0059] Standardization: Its purpose is to transform vague expressions into standardized engineering language to ensure log compliance.

[0060] The specific process includes:

[0061] (1) Template ID matching: The system calls the log template library, where each template corresponds to standard fields. Based on semantic similarity calculation, the optimal template is selected. A standardized language template library containing standard terminology in the construction field is pre-established, and each language template is assigned a unique template ID, forming a structured and standardized knowledge system. Simultaneously, a hierarchical coding rule is used to manage template IDs, ensuring the scalability and systematic nature of the templates. Each template contains a standardized field structure and strict validation rules, ensuring that the generated logs conform to industry standards and supporting dynamic hot updates of templates to maintain synchronization with the latest construction specifications. When log generation is required, the system matches the most suitable template ID through similarity calculation and converts unstructured information into standardized log entries according to predefined conversion rules.

[0062] (2) Language standardization conversion: converting fuzzy expressions into standard language, the expression of which is l i =log(oi,id) i ,t i ,ci), where l i The generated normalized log entry; oi represents the original input data; id i t is a unique identifier; i Template ID refers to a predefined standardized language template in the system; ci is context information, including timestamps, user roles, and other data.

[0063] (3) Two-way indexing mechanism: To ensure the traceability of the entire data processing process, the system has established a comprehensive two-way indexing mechanism. The forward index enables fast retrieval from the template ID to all related original records, while the reverse index supports querying the complete processing process trajectory by object ID.

[0064] To ensure the traceability of data processing, the system establishes a two-way indexing mechanism of "object ID-template ID". This allows for quick retrieval of all original records of similar events by template ID, and also allows querying of the corresponding standardized processing results by object ID.

[0065] For example, in a group chat that is bound to keywords related to construction progress and conditions, construction worker Zhang reports today's construction progress in text form. His report is: "The concrete work in Area A is finished today. The pump truck had a small problem, but it won't delay things. We'll continue tomorrow." The AI ​​will generate an object ID for this message, extract keywords, calculate semantic similarity, and analyze contextual features based on the context information. It will match a standardized template and generate a template ID, replace "finished" with "construction completed," associate the pump truck number and add fault details, and replace "won't delay things" with "construction schedule meets plan requirements."

[0066] Meanwhile, the system continuously monitors the processing, automatically verifies the integrity of the conversion results, and records the complete processing chain to meet the compliance requirements of the engineering quality management system. Through machine learning technology, the system can continuously optimize the template matching accuracy and improve the intelligence level of log generation.

[0067] This method ensures the standardization and traceability of data processing through a unique identifier system, supports multi-scenario expansion through a hierarchical template management mechanism, and achieves accurate terminology conversion by combining professional knowledge in the construction field. Ultimately, it outputs structured construction logs that conform to industry standards. The system-generated logs contain complete construction parameter dimensions, retaining detailed information from the original records while meeting standardization requirements, providing reliable data support for quality management and problem tracing during the construction process.

[0068] After context-aware optimization, structured analysis, and normalization, the process also includes data format adaptation. This involves generating various output formats according to user needs, enabling multi-format output generation. The log output format adaptation function is achieved through an intelligent multimodal conversion engine, which can automatically generate construction logs that conform to industry standards and can be flexibly presented based on user requirements. The system can generate various output formats according to user needs, and users can select the desired log presentation format through the system interface, including common engineering file formats such as images and tables. The system will intelligently adjust the content layout and visual style based on user selection, ensuring that the output logs not only meet industry standard requirements but also satisfy the actual needs of different usage scenarios.

[0069] Specific procedures:

[0070] Output format selection: User-specified format (such as table, image, PDF), system adapts layout.

[0071] Data preprocessing: Before generating charts or tables, the AI ​​system first preprocesses the log data to construct a standardized work log dataset, whose expression is D. i =(t i ,task_typei,valuei), where D i For data points; t i It is the timestamp of the log item (used for positioning on the timeline); task_type i It is the task type (e.g., "meeting", "development", etc.); value i This refers to the numerical values ​​of the tasks (such as quantity, completion rate, etc.). During the layout process, the system automatically identifies key information elements in the log content, such as construction progress and quality inspection results. Simultaneously, the system supports user-defined output style parameters, including font size, color theme, company logo, etc., to meet the brand image requirements of different projects. Aggregated dataset: C = {D1, D2, ..., D...} n}

[0072] Format conversion engine: Based on the preprocessed dataset C = {D1, D2, ..., D} n The AI ​​system calls a chart transformation function to automatically generate a visual chart according to the user-specified chart type (such as bar chart, line chart, pie chart, etc.). The expression is G(C,chart_type) = Graph_Transform(C,chart_type), where G is the generated chart; C is the input work log dataset, i.e., the aforementioned D. i The function is composed of C; chart_type is the selected chart type (e.g., bar chart, line chart, pie chart, etc.); Graph_Transform(C, chart_type) is the function that transforms the log data into a chart.

[0073] Output generation:

[0074] Table format: Directly generate structured tables (fields include task, progress, and person in charge).

[0075] Image format: Convert the dataset into a chart (such as a Gantt chart to display progress deviations).

[0076] Technical Implementation: Visual Optimization: Automatically adjust fonts and colors to conform to engineering specifications (e.g., highlighting delayed tasks in red).

[0077] This feature enables one-click conversion of construction logs from raw data to various standardized formats, significantly improving the efficiency and convenience of engineering document management. All output content maintains its correlation with the original data and retains complete engineering parameter information, providing a flexible solution for recording and reporting the construction process.

[0078] In addition, it includes log content organization and template filling, enabling structured log content according to user needs and adapting to different log types (such as daily and weekly reports). The AI ​​system performs semantic analysis on group chat records, identifies work-related keywords, and automatically matches the most suitable log type template based on the user-specified log type. The template contains standardized log structures, such as task descriptions, progress status, and problem records, ensuring that the generated logs conform to industry standards.

[0079] Specific procedures:

[0080] Log type selection: Users select the log type (such as "Daily Construction Report" or "Quality Inspection Report") through the interface, and the system automatically calls up the corresponding template.

[0081] Content population rules: Field mapping: Populate template fields with analysis results. Logical validation: Ensure data consistency (e.g., "Completion progress ≤ 100%"). Automated population is achieved; AI prioritizes key information.

[0082] For example, if Mr. Wang from the company needs to fill out a daily construction log form, he can select the output format on the system. The AI ​​will automatically preprocess the data according to the requirements and then automatically generate a daily construction log form that meets the requirements and industry standards based on the set time period.

[0083] Step S4: The generated construction logs are sent to relevant parties according to the preset list. At the same time, the system archives the logs to the cloud database by date, type, user identity, etc., supporting version rollback and permission management to ensure that the data is traceable and complies with the engineering file management specifications.

[0084] The system maintains a dynamically updated database of project personnel, which records the roles, permissions, and receiving preferences of each relevant party.

[0085] Once standardized construction logs are generated, the AI ​​engine will automatically identify the construction specialty, responsible area, and related personnel involved in the log content. Based on preset distribution rules, the logs will be accurately pushed to the work terminals of relevant responsible persons such as project managers, supervising engineers, and construction team leaders through in-system message notifications.

[0086] In terms of data archiving and traceability, the system has established a complete operation audit mechanism. All user operations on the platform are recorded in real time and converted into structured log information.

[0087] User operations on the platform are O = {o1, o2, ..., o n}, each o i This represents a user action, each action o i This will be recorded and fed back to the log system L. The specific implementation uses the operation recording function R(O) = {l1, l2, ..., l...} N}, where R(O) is the operation logging function, which transforms operation O into log records L, each record L i It includes information such as the timestamp of the operation, the type of operation, and the content of the operation.

[0088] Each audit log entry contains the following key information fields: a unique identifier for the operation, a timestamp for the operation, the operation type, the user's identity information (including user ID and role type), a description of the operation content (and the status of the operation result). These audit logs are stored in a distributed cloud database in chronological order, using a multi-replica mechanism to ensure data security.

[0089] The system offers powerful log retrieval capabilities, supporting queries based on multiple criteria such as operation ID, time range, user identity, and operation type. The retrieval process employs an optimized indexing algorithm to ensure rapid location of target records even within massive amounts of log data.

[0090] In terms of implementation, the system creates an inverted index for each operation. The index key includes key fields such as operation time, user identity, and operation type, enabling query expressions to be executed efficiently.

[0091] Each operation has a unique identifier (ID) id i And record the retrieval using this identifier, the expression being l i =log(o i ,id i ,t i ,c i ), where id i It is the unique identifier for the operation; t i It is the timestamp of the operation; c i It refers to the content or description of the operation.

[0092] Specifically, based on preset distribution rules, logs are precisely pushed to the work terminals of relevant responsible persons such as project managers, supervising engineers, and construction team leaders via in-system message notifications.

[0093] Suppose that construction worker Wang needs to send daily construction log reports to relevant responsible persons such as the project manager, quality inspector, and supervising engineer. AI will automatically identify the responsible persons based on the construction log content, and Wang can also adjust the identified responsible persons. Wang can select the sending time on the system, and according to preset distribution rules, the AI ​​can automatically send the daily construction log reports to the selected responsible persons. After sending, the relevant responsible persons will immediately receive a message notification and the daily construction log report in the system.

[0094] Furthermore, to achieve the above objectives, the present invention also provides an automatic construction log generation system based on working condition keywords, the system comprising:

[0095] Group Management Module: Used to create project groups and build a working condition keyword library K in the corresponding group according to the project type;

[0096] Multimodal parsing module: Communicatively connected to the group management module, used to collect group messages within the group and match them using the AI ​​matching function f(m,k). i = AI_Match(m,k) i )∈[0,1] Real-time identification of working condition keywords, group message type includes at least one of text, voice and image; when the keyword matching degree satisfies f(m,ki)≥θ, it is marked as valid log content and the valid log information set is output, where m is the group message, ki is the i-th keyword in the working condition keyword library K, and θ is the preset threshold;

[0097] Log generation engine: connected to the multimodal parsing module, including:

[0098] Context-aware units are used to optimize raw data using a construction domain knowledge graph.

[0099] The template matching unit calls standardized templates based on the log type selected by the user, transforming colloquial expressions into standard language.

[0100] The multi-format output unit is used to convert preprocessed datasets into construction logs according to user-specified chart types;

[0101] Distribution and Traceability Module: Connected to the output of the log generation engine, this module automatically distributes construction logs to relevant parties based on a preset list and project organizational structure; archives logs to a cloud database by date, type, and user identity, and records them using the operation log function R(O)={l1,l2,...,l N It enables full-cycle traceability, with each record containing a timestamp, operation type, and content.

[0102] Furthermore, the group management module specifically includes:

[0103] The keyword configuration unit is used to define a set of keywords that match the project type;

[0104] The intelligent matching unit uses algorithms to associate groups with keyword types;

[0105] The interface can be dynamically adjusted, allowing administrators to add, delete, and modify keywords and record version history.

[0106] Furthermore, the multimodal analysis module includes a risk prediction submodule, which performs the following operations:

[0107] Real-time statistics on the frequency of occurrence of risky keywords such as "delay" and "material shortage";

[0108] When the frequency exceeds the dynamic threshold, a Gantt chart deviation report is generated based on the associated project progress data;

[0109] The scope of risk impact is visualized through color scale annotations and deviation curves.

[0110] Furthermore, the template matching unit of the log generation engine includes:

[0111] A two-way indexing mechanism establishes a two-way traceability path through a unique object ID and a template ID;

[0112] Semantic converters transform ambiguous expressions into standard engineering terms;

[0113] The dynamic updater keeps the template library synchronized with the latest construction specifications in real time.

[0114] Furthermore, the multi-format output unit uses a graph conversion function.

[0115] G(C,chart_type) = Graph_Transform(C,chart_type) implements log visualization, where:

[0116] C represents the preprocessed dataset; chart_type supports bar charts, line charts, and Gantt charts.

[0117] Furthermore, the cloud database of the distribution traceability module employs a multi-dimensional classification index:

[0118] First-level directory: Project ID;

[0119] Second-level directory: Log type (progress / quality / security);

[0120] Third-level directory: Date (YYYY-MM-DD);

[0121] Metadata encapsulation, digital signatures, and blockchain storage are used to prevent tampering.

[0122] Furthermore, the operation recording function R(O) = {l1,l2,...,l...} N The traceability mechanism includes:

[0123] The inverted index engine uses operation time, user ID, and log version number as index keys.

[0124] The visual traceability interface displays the operation process along a timeline and highlights the discrepancies.

[0125] Furthermore, the risk prediction submodule works in conjunction with the log generation engine; when high-risk keywords are detected:

[0126] Freeze the regular log generation process;

[0127] Prioritize triggering the alert log template (T_Risk_Alert) to generate a report;

[0128] The warning report is distributed to the terminal of the person in charge of safety and displayed at the top.

[0129] Furthermore, the system integrates reinforcement learning algorithms and dynamically optimizes them based on human review feedback:

[0130] Keyword weight configuration w(ki):

[0131] w(ki) new =w(ki) old +α·(Artificial weight baseline value - w(ki)) old )

[0132] The conflict resolution strategy library has added new rules to support the automatic handling of similar conflicts.

[0133] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0134] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0135] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A construction log automatic generation method based on a work condition keyword, characterized by, The method comprises the following steps: Step S1, creating a project group and constructing a working condition keyword library K in the corresponding group according to the project type; Step S2, collecting group messages in the group, and performing AI matching on the collected messages f(m, k i ) = AI_Match(m, k i ) ∈ [0, 1] real-time identification of working condition keywords, the group message type includes at least one of text, voice and image; when the keyword matching degree satisfies f(m, ki) ≥ θ, it is marked as valid log content, and the valid log information set is output, wherein m is a group message, ki is the i th keyword in the working condition keyword library K, and θ is a preset threshold; Step S3, performing context-aware optimization, structured analysis and standardization processing on the effective information set collected in step S2 to generate a corresponding construction log; Step S4, distributing the construction log obtained in step S3 to the relevant parties according to the pre-set list, and archiving the construction log to the cloud database according to the date, type and user identity.

2. The construction log automatic generation method based on a work condition keyword according to claim 1, characterized by, In step S1, the step of constructing the working condition keyword library K comprises: Defining a keyword set according to the project type; Associating the group with the pre-set keyword type through an intelligent matching algorithm; Adding, deleting or modifying keywords through an interactive interface.

3. The construction log automatic generation method based on a work condition keyword according to claim 1, characterized by, In step S2, if a single message matches multiple keywords, the matching keywords are dynamically prioritized through an arbitration mechanism rule, and the arbitration mechanism rule includes weight dynamic prioritization, conflict detection and resolution strategy, and the weight dynamic prioritization priority is before the conflict detection and resolution strategy.

4. The construction log automatic generation method based on a work condition keyword according to claim 3, characterized by, In the step S2, when it is detected that the keyword matching fails or the arbitration mechanism rule cannot make an effective decision, an artificial review process is triggered.

5. The construction log automatic generation method based on a work condition keyword according to claim 4, characterized by, In step S2, if the effective content is a risk keyword and the occurrence frequency is higher than the pre-set frequency, an early warning mechanism is automatically triggered, and a Gantt chart deviation analysis report is generated.

6. The construction log automatic generation method based on a work condition keyword according to claim 1, characterized by, In step S3, the structured analysis and standardization processing step comprises: Generating a unique object ID for the original data oi; Based on semantic analysis, match the pre-defined template ID, form the standard language l i = log(o i , id i , t i , c i ) to realize the traceability of template ID to original record through forward index, and reverse index supports the processing track query of object ID, wherein l i is the i th standardized log generated; oi is the original input data; id i is the unique identifier; t i is the template ID, which refers to the pre-defined standardized language template in the system; ci is the context information.

7. The construction log automatic generation method based on a work condition keyword according to claim 1, wherein, In step S3, the context-aware optimization includes knowledge graph construction and context feature extraction.

8. A construction log automatic generation system based on a work condition keyword, characterized by, The system comprises: A group management module for creating a project group and constructing a working condition keyword library K in the corresponding group according to the project type; A multi-modal analysis module in communication connection with the group management module is configured to collect group messages in the group, identify key words of working conditions in real time through an AI matching function f(m, k i ) = AI_Match(m, k i ) ∈ [0, 1], and the group message types include at least one of text, voice and image; when the key word matching degree satisfies f(m, ki) ≥ θ, the key word is marked as valid log content, and an effective log information set is output, where m is a group message, ki is the i-th key word in a key word library K of working conditions, and θ is a preset threshold. A log generation engine connected with the multi-modal analysis module, comprising: A context-aware unit for optimizing the original data through a construction field knowledge graph; A template matching unit for converting colloquial expressions into standard language based on the user-selected log type and calling a standardized template; A multi-format output unit for converting the pre-processed data set into a construction log according to the user-specified chart type; Distribution traceability module: connected with the log generation engine output, for automatically distributing construction logs to relevant parties according to a preset list and project organization structure; archiving logs to a cloud database according to date, type and user identity, and realizing whole-cycle traceability through an operation record function R(O)={l1, l2,..., l N} each record containing a timestamp, operation type and content.

9. A construction log automatic generation system based on a work condition keyword, characterized by, The system comprises a memory, a processor and a working condition keyword-based construction log automatic generation program stored on the memory and executable on the processor, and the working condition keyword-based construction log automatic generation program is executed by the processor to realize the steps of the working condition keyword-based construction log automatic generation method according to any one of claims 1 to 8.

10. A computer storage medium, characterized in that, The computer storage medium stores a working condition keyword-based construction log automatic generation program, and the working condition keyword-based construction log automatic generation program is executed by the processor to realize the steps of the working condition keyword-based construction log automatic generation method according to any one of claims 1 to 8.

Images