Rapid function recommendation method, system and device for building industry

Abstract

The invention provides a building industry-oriented function rapid recommendation method, system and device, and the method comprises the steps: obtaining a plurality of candidate tasks from system building tasks based on the spatial position information of a current user; obtaining visual information of the current user, and determining task key elements based on the visual information; determining a target task from the plurality of candidate tasks based on the task key elements and the identity information of the current user; determining a user intention of the current user based on the target task; based on the user intention, determining a function recommended to the current user; the method, the system and the device provided by the invention have relatively high intelligence and automation levels, and can improve the operation convenience and the construction project management efficiency.

Description

Technical Field

[0001] The present invention relates to the field of construction, and in particular to a method, system and device for quickly recommending functions for the construction industry. Background Art

[0002] In the construction industry, work scenarios are complex and diverse, involving many people with clear responsibilities, and the tools and functions required vary depending on the specific tasks and scenarios. Traditional construction-related software or systems often have cumbersome function selections, and users need to search for the required functions in numerous menus, which is inefficient and prone to errors, seriously affecting work progress and quality.

[0003] In view of this, the present invention provides a method, system and device for quickly recommending functions for the construction industry to solve the above problems. Summary of the invention

[0004] One or more embodiments of the present invention provide a method for quickly recommending functions for the construction industry, the method comprising: obtaining multiple candidate tasks from system construction tasks based on spatial position information of a current user; obtaining visual information of the current user, and determining key elements of the task based on the visual information; determining a target task from the multiple candidate tasks based on the key elements of the task and identity information of the current user; determining the user intent of the current user based on the target task; and determining functions recommended to the current user based on the user intent.

[0005] One or more embodiments of the present invention also provide a rapid function recommendation system for the construction industry, the system comprising: an acquisition module, configured to acquire multiple candidate tasks from system construction tasks based on spatial position information of a current user; a first determination module, configured to acquire visual information of the current user, and determine key task elements based on the visual information; a second determination module, configured to determine a target task from the multiple candidate tasks based on the key task elements and identity information of the current user; a third determination module, configured to determine the user intent of the current user based on the target task; and a fourth determination module, configured to determine a function to be recommended to the current user based on the user intent.

[0006] One or more embodiments of the present invention also provide a device for rapid function recommendation for the construction industry, the device comprising: at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least part of the computer instructions to implement the method for rapid function recommendation for the construction industry as described in the above embodiments.

[0007] The method, system and device for rapid function recommendation for the construction industry provided in the present invention combine the user's visual information with the spatial position, system tasks and user identity to achieve rapid and accurate judgment of the current construction progress and user location, and at the same time combine the task management system with the construction tasks to achieve accurate inference of the operational intentions of various personnel. It has a high level of intelligence and automation, and can improve operational convenience and construction project management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The present invention will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0009] Figure 1 is a schematic diagram of an application scenario of a function rapid recommendation system for the construction industry according to some embodiments of the present invention;

[0010] Figure 2 is an exemplary flow chart of a method for rapid function recommendation for the construction industry according to some embodiments of the present invention;

[0011] Figure 3 is an exemplary flow chart of determining a target task according to some embodiments of the present invention;

[0012] Figure 4 is an exemplary flow chart of determining user intent according to other embodiments of the present invention;

[0013] Figure 5 is a schematic diagram of internal modules of a processor according to some embodiments of the present invention. DETAILED DESCRIPTION

[0014] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of the present invention. For ordinary technicians in this field, the present invention can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0015] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0016] As shown in the present invention and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular, but also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list, and the method or device may also include other steps or elements.

[0017] The present invention uses a flow chart to illustrate the operations performed by the system according to an embodiment of the present invention. It should be understood that the preceding or following operations are not necessarily performed precisely in order. On the contrary, the various steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0018] Figure 1 is a schematic diagram of an application scenario of a function rapid recommendation system for the construction industry according to some embodiments of the present invention. Figure 1 As shown, the function rapid recommendation system 100 for the construction industry includes a current user 110, a network 120, a terminal 130, a processor 140, and a storage device 150.

[0019] The current user 110 refers to a person who needs to be recommended a function or a client used by the person. For example, the current user 110 may include a construction worker, a team leader, a construction manager (such as a construction worker, an acceptance officer, etc.) or a client used by the person.

[0020] In some embodiments, personnel with different identities often correspond to different work tasks. Even personnel with the same identity may have different tasks to undertake in different areas of a construction project. For example, the work tasks corresponding to construction personnel include construction guidance, task drawing viewing, and tool query; the work tasks corresponding to team leaders include task splitting, task self-inspection, initiating acceptance and other management-related tasks; the work tasks corresponding to construction management personnel include project acceptance, construction inspection and other tasks. For another example, construction personnel in the electrical installation area correspond to work tasks related to electrical installation, and in the water supply and drainage installation area correspond to work related to pipeline installation and regulation. In some embodiments, function recommendations can be made based on the current user's location information, visual information, and identity information. For more information on related content, please refer to the relevant parts below, such as Figure 2 to Figure 4 .

[0021] The network 120 can connect the components of the rapid function recommendation system 100 for the construction industry and / or connect other components outside the rapid function recommendation system 100 for the construction industry. In some embodiments, one or more components of the rapid function recommendation system 100 for the construction industry (e.g., the terminal 130, the processor 140, and the storage device 150, etc.) can be connected and / or communicate with each other through the network 120. For example, the terminal 130 can send the user's visual information to the processor 140, etc. through the network 120.

[0022] The terminal 130 can provide functional components related to user interaction and can implement user interaction functions (such as providing or displaying information and data to the user). The user can refer to the current user 110. As an example only, the terminal 130 can be one of a mobile device, a tablet computer, a laptop computer, a desktop computer, etc., or other devices with input and / or output functions, or any combination thereof.

[0023] In some embodiments, the terminal 130 may have a positioning function. For example, the terminal 130 may obtain the spatial location information of the current user 110 .

[0024] In some embodiments, the terminal 130 may have an image acquisition function. For example, the terminal 130 may acquire visual information of the current user 110 .

[0025] In some embodiments, the terminal 130 may also have a display function. For example, the terminal 130 may display recommended functions to the current user 110. The above description of the terminal 130 is only an example, and the terminal 130 may also have other functions, such as an input function, etc., which are not limited here.

[0026] The processor 140 is capable of processing information and / or data related to the rapid function recommendation system 100 for the construction industry to perform one or more functions described in the present invention. In some embodiments, the processor 140 can obtain multiple candidate tasks from the system construction tasks based on the spatial position information of the current user; obtain the visual information of the current user, and determine the key elements of the task based on the visual information; determine the target task from the multiple candidate tasks based on the key elements of the task and the identity information of the current user; determine the user intention of the current user based on the target task; and determine the function recommended to the current user based on the user intention. Detailed description of the relevant content can be found later (such as Figure 2 to Figure 4 etc.) related description.

[0027] In some embodiments, the processor 140 may include a central processing unit (CPU), a digital signal processor (DSP), a microcontroller unit (MCU), a computer, a user console, etc., or any combination thereof. In some embodiments, the processor 140 may include a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 140 may be local or remote. In some embodiments, the processor 140 may be implemented on a cloud platform. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, etc., or any combination thereof.

[0028] In some embodiments, processor 140 may be integrated within terminal 130 .

[0029] The storage device 150 is capable of storing data, instructions and / or any other information. In some embodiments, the storage device 150 may store data obtained from the terminal 130, the processor 140, etc., such as the spatial location information of the current user 110, etc. In some embodiments, the storage device 150 may include a large-capacity memory, a removable memory, a volatile read-write memory, a read-only memory (ROM), etc., or any combination thereof. In some embodiments, the storage device 150 may be executed on a cloud platform. In some embodiments, the storage device 150 may be connected to the network 120 to communicate with one or more other components of the function rapid recommendation system 100 for the construction industry (e.g., the terminal 130, the processor 140, etc.). In some embodiments, the storage device 150 may be part of the processor 140.

[0030] It is worth noting that the function rapid recommendation system 100 for the construction industry is provided for illustrative purposes only and is not intended to limit the scope of the present invention. For those of ordinary skill in the art, various changes and modifications can be made based on the description of the present invention. For example, the function rapid recommendation system 100 for the construction industry may also include a database, an information source, and the like. For another example, the function rapid recommendation system 100 for the construction industry may be implemented on other devices to achieve similar or different functions. However, these changes and modifications will not deviate from the scope of the present invention.

[0031] Figure 2 FIG. 2 is an exemplary flow chart of a method for quickly recommending functions for the construction industry according to some embodiments of the present invention. In some embodiments, process 200 may be executed by a processing device (eg, processor 140). Figure 2 As shown, process 200 includes the following steps.

[0032] Step 210, based on the spatial location information of the current user, a plurality of candidate tasks are obtained from the system building tasks.

[0033] Spatial location information refers to data / information related to the spatial location of the current user. For example, spatial location information may include three-dimensional positioning information of the current user's location. For more information about the current user, see Figure 1 and related instructions.

[0034] In some embodiments, the spatial location information can be obtained based on the positioning module of the terminal used by the user when the user logs in to the client. For example, when the current user logs in, the processor 140 can control the positioning module to obtain the spatial location information of the current user. In some embodiments, the positioning module can include a Beidou positioning system, a GPS positioning system, a Bluetooth positioning system, etc. In some embodiments, the positioning module can be integrated in the terminal 130.

[0035] In some embodiments, the spatial position information may also be obtained while obtaining the user's visual information. For example, when obtaining the visual information of the current user, the positioning module simultaneously obtains the spatial position information.

[0036] System construction tasks refer to tasks related to construction projects that are preset in the function rapid recommendation system for the construction industry. System construction tasks cover multiple stages of construction project planning, design, construction, and management. Exemplary system construction tasks include project site selection tasks, architectural design tasks (such as structural design, electrical system design), construction tasks (such as foundation treatment, concrete pouring), safety and quality management tasks, etc. In some embodiments, system construction tasks can be preset based on data of historical construction projects.

[0037] Candidate tasks refer to system building tasks that may be selected for recommendation to the current user.

[0038] In some embodiments, the candidate tasks are correlated with the spatial location information of the current user. That is, the current user is in different locations, and the corresponding candidate tasks are different. The processor 140 can obtain the candidate tasks from the system construction tasks based on the spatial location information of the current user. For example, when the spatial location information indicates that the current user is in the foundation position, the processor 140 can select tasks related to the foundation area from the system construction tasks as candidate tasks (for example, tasks such as foundation treatment, foundation excavation, and concrete pouring are selected as candidate tasks).

[0039] In some embodiments, the candidate tasks may be represented by a set. For example, the candidate task set R may be represented as R = {r1, r2, r3, ..., r m}. Among them, r1, r2, r3, ..., r m They represent each candidate task respectively, and m is the number of candidate tasks.

[0040] In some embodiments of the present invention, multiple candidate tasks are determined by the spatial location information of the current user, and tasks related to the user's area can be screened out from a large number of system building tasks, thereby narrowing the search scope of the target task and improving processing efficiency.

[0041] Step 220, obtaining visual information of the current user, and determining the key elements of the task based on the visual information.

[0042] Visual information refers to information / data obtained by analyzing images and / or video data acquired by image acquisition devices (such as cameras, webcams, etc.). For example, visual information may include objects (such as construction materials, construction tools, etc.), people, scenes (such as kitchens, bathrooms, etc.), etc.

[0043] In some embodiments, the image acquisition device can automatically or manually acquire image information of the space where the current user is located. For example, the current user can manually control the image acquisition device to acquire image information of the space where the current user is located. For another example, after the current user logs in, the processor 140 can automatically control the image acquisition device to regularly or periodically acquire image information of the space where the current user is located. The image acquisition device can be integrated into the terminal 130.

[0044] In some embodiments, the processor 140 may also process the acquired image information to acquire visual information. Exemplary image processing methods may include image recognition technology, machine vision technology, and the like.

[0045] The key elements of a task refer to the key data and / or information related to determining the target task. For example, the key elements of a task may include construction equipment, construction raw materials, etc. The key elements of a task can reflect information such as the type of task and the execution area. Among them, the target task refers to the task that the current user is most likely to perform.

[0046] In some embodiments, the processor 140 may select data and / or information related to the determination of the target task from the visual information as the mission-critical elements. For example, the processor 140 may select building components (such as columns, beams, walls, floors, windows, doors, etc.), construction equipment (such as scaffolding, cranes), construction raw materials (such as bricks, steel bars, concrete), etc. from the visual information as mission-critical elements. In some embodiments, the mission-critical elements may be a set of multiple different elements. For example, different building components, construction equipment, and construction raw materials may be used as multiple different elements to form mission-critical elements.

[0047] Step 230 , determining a target task from a plurality of candidate tasks based on the key elements of the task and the identity information of the current user.

[0048] User identity information refers to data related to the identity of the current user. The current user's identity information includes name, age, responsibilities (such as construction personnel, team leader, quality inspection personnel, etc.), contact information, etc.

[0049] In some embodiments, the processor 140 may determine whether the task key element is related to the task information of the candidate task, and determine the target task according to the matching degree between the candidate task and the identity information of the current user. Figure 3 and its related parts.

[0050] Step 240 , determining the user intention of the current user based on the target task.

[0051] User intent refers to the current user's goals or needs when using construction-related software / systems. For example, user intent can include obtaining construction instructions, viewing drawings, team management, etc.

[0052] In some embodiments, the processor 140 can determine one or more task states corresponding to the target task and their corresponding task events; determine the state vector corresponding to each task state based on the task state and its corresponding task event; determine the task vector based on the state vector and the target task; determine the event vector corresponding to each task event; determine the task event matrix based on the event vector corresponding to each task event and the task vector; and determine the user intention of the current user based on the task event matrix.

[0053] The task status refers to the execution status of the target task. In some embodiments, the task status includes not started, in progress, pending acceptance, completed, etc.

[0054] Task events refer to the operations that need to be performed when executing the target task.

[0055] In some embodiments, different task states may correspond to different task events. For example, Z1 indicates that the task state is not started, and the corresponding task events include: planning S1, task allocation S2, material preparation S3; Z2 indicates that the task state is in progress, and the corresponding task events include: progress tracking S4, task self-check S5, quality acceptance S6; Z3 indicates that the task state is in progress, and the corresponding task events include: acceptance application S7, problem feedback S8, and rectification initiation S9; Z4 indicates that the task state is completed, and the corresponding task events include: construction summary S 10 , Achievements Display 11 The above correspondence between task status and task event is only for illustrative purposes and does not constitute a limitation. Task status may also correspond to other task events (for example, the task event corresponding to the task status of "in progress" may also include a specific construction process, etc.).

[0056] The state vector refers to a vectorized representation of a task state. In some embodiments, the state vector may represent task events corresponding to different task states.

[0057] In some embodiments, the processor 140 may combine the task state and the task event to obtain a state vector. For example, the processor 140 may combine the task state Z1 (indicating that the task state is not started) with the corresponding task events S1, S2, and S3 to obtain a state vector Z1=[S1, S2, S3] indicating that the task state is not started. Similarly, the processor 140 may also obtain state vectors Z2=[S4, S5, S6], Z3=[S7, S8, S9], and Z4=[S1, S2, S3] corresponding to the task states in progress, pending acceptance, and completed, respectively. 10 , S 11 ]wait.

[0058] The task vector is a vectorized representation of the task state of the target task. In some embodiments, the task vector can represent different task states of the target task. For example, the task vector R = [Z1, Z2, Z3, Z4]. In some embodiments, when there are multiple target tasks, the multiple target tasks can also be represented by the task vector. For example, R1 = [Z 1-1 , Z 1-2 , 0, Z 1-4 ], R2=[Z 2-1 , 0, 0, Z 2-4 ], R3=[Z 3-1 , Z 3-2 , Z 3-3 , Z 3-4 ], etc. Among them, the first number in the vector value represents the target task number, the second number represents the task status number, and the Z 2-1 Taking as an example, the number 2 represents the sequence number of the target task, 1 represents the first state of the target task (corresponding to state Z1); the vector value 0 represents that the target task does not exist in this state.

[0059] The event vector is a vectorized representation of a task event. For example, an event vector can be represented as [S 1-1 , S 1-2 , S 1-3 , S 1-4 , S 1-5 ], etc. The first number in the vector value represents the target task number, and the second number represents the task event number, with S 1-2 For example, the number 1 represents the sequence number of the target task, and 2 represents the sequence number of the task event.

[0060] The task event matrix is ​​a matrix representing all the task events that need to be executed to complete the target task. In some embodiments, the processor 140 can obtain the task event matrix by replacing the state vector in the task vector with the event vector. In some embodiments, for a task vector R = [Z1, Z2, Z3, Z4] of a target task, the processor 140 can replace the state vectors Z1, Z2, Z3, Z4 with the corresponding event vectors, respectively, to obtain the task event matrix M = [S1, S2, S3, S4, S5, S6, 0, 0, 0, S 10 , S 11 ]. The vector value 0 indicates that there is no task event that needs to be executed in the task state.

[0061] In some embodiments, for multiple task vectors (eg, R1 = [Z 1-1 , Z 1-2 , 0, Z 1-4 ], R2=[Z 2-1 , 0, 0, Z 2-4 ], R3=[Z 3-1 , Z 3-2 , Z 3-3 , Z 3-4 ], etc.), the processor 140 can replace the state vector with the event vector of the task event corresponding to each task state, obtain multiple sub-matrices, and combine the multiple sub-matrices to obtain the task event matrix. Exemplarily, formula (1) is an exemplary representation of the task event matrix M' obtained by combining multiple sub-matrices corresponding to multiple target tasks.

[0062] In some embodiments, the processor 140 can assign a value to each element in the task event matrix based on the event completion status corresponding to each task event to obtain a task event status matrix; based on a preset task screening method, filter out the target event from the task event status matrix; based on the target event, determine the user intention of the current user.

[0063] The event completion status includes completed, uncompleted, etc. In some embodiments, the processor 140 may assign a value of 1 to a task event whose event completion status is completed in the task event matrix, and assign a value of 0 to a task event whose event completion status is uncompleted. 1-1 If the completion status is completed, the task event S 1-1 The value is 1 if it is not completed, and 0 if it is not completed.

[0064] The task event state matrix is ​​a matrix representation of the completion state of the task event. In some embodiments, after the processor 140 completes the assignment of values ​​to the elements in the task event matrix, the matrix obtained is the task event state matrix.

[0065] The target event refers to the task event that needs to be completed with the highest priority in the target task. It is understandable that the task events in the target task are often continuous, that is, the subsequent tasks cannot be completed without completing the predecessor tasks. Therefore, setting the task that needs to be completed with the highest priority as the target event can ensure the smooth execution of the target task, and the subsequent recommended functions for the current user can better meet their needs.

[0066] In some embodiments, the processor 140 may filter out the target event from the task event state matrix based on a preset task screening method. An exemplary preset task screening method includes that the processor 140 filters out the task events assigned a value of 0 from the task event state matrix, and selects the task event with the smallest sequence number as the target event. For example, for the task event state matrix P = [S1 = 1, S2 = 1, S3 = 1, S4 = 0, S5 = 0, S6 = 0, 0, 0, 0, S 10 =0, S 11 =0], indicating that the task events S1, S2, and S3 in the target task have been completed, and the processor 140 selects the uncompleted task events S4, S5, S6, and S7 with a value of 0. 10 , S 11 , and select the task event S4 with the smallest task event number as the target event that needs to be completed most at present.

[0067] In some embodiments, the preset task screening method also includes other methods. For example, in the task event matrix, the priority of the task event can be arranged from left to right in the order of high priority to low priority. After the processor 140 screens out the unfinished events assigned a value of 0, the unfinished event on the far left is selected as the target event. For another example, it can also be arranged in order of low priority to high priority, and the unfinished event on the far right is selected as the target event, which is not limited here.

[0068] In some embodiments, the processor 140 may use the relevant operations corresponding to the target event as the user intention of the current user. For example, if the target event is concrete pouring, the user intention of the current user may include construction personnel allocation, construction guidance, construction drawing viewing, etc.

[0069] In some embodiments, the processor 140 may also determine the user intent of the current user through a machine learning model. For more information on the related content, see Figure 4 and related instructions.

[0070] In some embodiments, the processor 140 may also verify the user's intention based on the current user's identity information. For example, if the current user's identity information is a construction worker and the determined user intention is project acceptance, it means that the verification fails, the user's identity information does not match the user's intention, and the user's intention needs to be re-determined.

[0071] In some embodiments of the present invention, verifying the user intention based on the identity information of the current user can further improve the accuracy of the user intention and obtain better function recommendation effect.

[0072] Step 250, based on the user's intention, determine the functions recommended to the current user.

[0073] The functions recommended to the current user refer to the functions in the building-related software / system that the current user is most likely to want to use.

[0074] In some embodiments, the processor 140 may recommend the function corresponding to the user's intention to the current user. For example, when the user's intention includes construction guidance, construction drawing viewing, etc., the processor 140 may recommend the functions of construction guidance and construction drawing viewing to the user.

[0075] In some embodiments of the present invention, by combining the user's visual information with multimodal information such as spatial position, system task, and user identity, it is possible to comprehensively analyze and quickly select the most relevant and applicable function options for the user, eliminating the tedious and complicated function search process in the construction software / system, reducing the risk of incorrect function selection, and improving operational convenience and project management efficiency.

[0076] Figure 3 FIG. 300 is an exemplary flow chart of determining a target task according to some embodiments of the present invention. In some embodiments, process 300 may be executed by a processing device (eg, processor 140). Figure 3 As shown, process 300 includes the following steps.

[0077] Step 310 : for each of the plurality of candidate tasks, determine whether the candidate task is related to the task key element based on its corresponding task information.

[0078] Task information refers to the information / data required to execute the candidate task. For example, task information includes construction task items, construction areas, construction schedules, task collaborators, construction phases corresponding to construction tasks, construction drawings of construction areas, etc. For more information about key elements of tasks, see Figure 2 and its related parts.

[0079] In some embodiments, for each of the multiple candidate tasks, the processor 140 may determine whether the corresponding task information and the task key element include the same content, and if so, the processor 140 may determine that the candidate task is related to the task key element. For example, if the task key element includes specific construction equipment and / or construction materials, and the task information of a candidate task includes content related to the use or maintenance of the equipment / materials, the processor 140 may determine that the candidate task is related to the task key element.

[0080] In some embodiments, the task information and the task key elements may also be represented by vectors. The processor 140 may determine whether the candidate task is related to the task key element based on the vector distance between the task information and the task key element corresponding vector. For example, if the vector distance between the task information and the task key element corresponding to the candidate task is less than a preset threshold, the processor 140 may determine that the candidate task is related to the task key element. The vector distance may include one or more of cosine distance, Euclidean distance, Manhattan distance, Chebyshev distance, etc.

[0081] Step 320 : In response to the candidate task being associated with the task key element, determining a degree of match between the candidate task and the identity information of the current user.

[0082] The matching degree is a parameter that evaluates the relevance of the candidate task to the current user. Different users have different relevance to different tasks. The higher the matching degree, the higher the relevance between the current user and the candidate task, and the more likely the current user is to want to perform the task.

[0083] In some embodiments, the processor 140 can determine the task focus of the current user based on the identity information of the current user; determine the weight of each element in the task key elements based on the task focus; and perform weighted calculation on the correlation between the candidate tasks and each element in the task key elements based on the weight to determine the degree of matching.

[0084] Task focus refers to the aspects that the current user pays attention to, values, or prioritizes for the candidate task. For the same candidate task, different users may have different task focus. Taking tasks related to construction operations (such as concrete pouring) as an example, the task focus of the team leader may include task splitting and acceptance, while the task focus of the construction manager may include task progress monitoring, quality control, task acceptance, etc.

[0085] In some embodiments, there is a preset correspondence between the task focus and the user's identity information. For example, the storage device 150 may pre-store a preset comparison table of user identity information and task focus in different system building tasks, and the processor 140 may obtain the task focus of the current user by reading the preset comparison table from the storage device 150 according to the user's identity information.

[0086] The weight is a parameter that measures the current user's attention to each element in the mission-critical element. The larger the weight, the more attention the current user pays to the element in the mission-critical element. In some embodiments, the weight of each element in the mission-critical element corresponding to the identity information of different users can be preset.

[0087] The correlation between the candidate task and each element in the task-critical element refers to the degree of correlation between the candidate task and the task-critical element. In some embodiments, the processor 140 may determine the correlation between the candidate task and each element in the task-critical element based on the task information of the candidate task and the vector distance between the vectors corresponding to each element in the task-critical element. For example, the smaller the vector distance, the greater the correlation between the candidate task and the element in the task-critical element. Further, the processor 140 may perform data processing such as standardization and normalization on the vector distance and use it as the correlation between the candidate task and each element in the task-critical element.

[0088] In some embodiments, the processor may perform weighted calculation on the relevance between the candidate task and each element in the task key element based on the weight to determine the matching degree. Exemplarily, the matching degree may be determined based on the following formula (2): S(r i )=p1×ω1+p2×ω2+…+p n ×ω n (2) Among them, S(r i ) represents the matching degree between the candidate task and the identity information of the current user; p1, p2, …, p n Respectively represent the correlation between the candidate task and each element in the task key element; ω1, ω2, …, ω n They represent the weight of each element in the mission-critical elements.

[0089] Step 330: determine the target task based on the matching degree.

[0090] In some embodiments, the processor 140 may determine the candidate task with the highest matching degree as the target task. In some embodiments, the processor 140 may also determine one or more candidate tasks with a matching degree greater than a matching degree threshold as the target task.

[0091] In some embodiments of the present invention, the relevance of the task to the user identity and the user visual information (i.e., the key elements of the task) is comprehensively considered, and the task that the user is most likely to perform is determined based on the degree of matching. This can ensure that the task recommended to the user best suits his or her current situation and needs, thereby improving the accuracy of subsequent recommended functions for the user.

[0092] Figure 4 FIG. 4 is an exemplary flow chart of determining user intent according to other embodiments of the present invention. In some embodiments, process 400 may be executed by a processing device (eg, processor 140). Figure 4 As shown, process 400 includes the following steps.

[0093] Step 410, obtaining the task content corresponding to the target task.

[0094] Task content refers to the specific operations and work requirements that need to be performed in the target task. Different target tasks may have different task contents. For example, the task content of foundation construction tasks includes earth excavation and measurement, foundation reinforcement and concrete pouring, etc.; the task content of safety management tasks includes the formulation of safety training plans, regular safety inspections and accident reporting, etc.

[0095] In some embodiments, the processor 140 may obtain the task content corresponding to the target task in a variety of ways. For example, the task content may be obtained by manual input; for another example, the processor 140 may also process the target task through methods such as data mining and text analysis, machine learning models, etc. to obtain the corresponding task content.

[0096] Step 420: extract features based on the task content to obtain a first feature vector.

[0097] The first feature vector is a vector that represents the features of the task content.

[0098] In some embodiments, the first feature vector can be obtained based on a variety of methods. For example, the first feature vector can be extracted by processing the task content based on statistical methods, transformation methods (such as Fourier transform), dimensionality reduction methods (such as principal component analysis PCA), and pre-trained machine learning models.

[0099] Step 430: extract features based on the identity information of the current user to obtain a second feature vector.

[0100] The second feature vector is a vector representing features of the identity information of the current user.

[0101] In some embodiments, the second feature vector can be obtained based on a variety of methods. For example, the second feature vector can be obtained by processing the identity information of the current user based on statistical methods, transformation methods (such as Fourier transform), dimensionality reduction methods (such as principal component analysis PCA), and pre-trained machine learning models to extract the second feature vector.

[0102] Step 440: fuse the first feature vector and the second feature vector to obtain a fused feature vector.

[0103] The fused feature vector is the result of fusing multiple feature vectors.

[0104] In some embodiments, the processor 140 may fuse the first feature vector and the second feature vector by various methods. For example, the processor 140 may fuse the first feature vector and the second feature vector by linear fusion (e.g., vector concatenation), weighted fusion, nonlinear fusion (e.g., deep learning), etc. to obtain a fused feature vector.

[0105] In some embodiments, the processor 140 can also obtain the historical user intentions of the current user; determine the intention tendency vector based on the identity information of the current user and the historical user intentions; obtain a fused feature vector based on the intention tendency vector, the first feature vector and the second feature vector through a feature fusion model; the feature fusion model is a machine learning model.

[0106] Historical user intent refers to the goals or needs of current users when using building-related software / systems in historical data.

[0107] In some embodiments, the historical user intentions may be acquired based on historical data. For example, the storage device 150 may store a large amount of historical data including historical user intentions, and the processor 140 may read the historical user intentions of the current user from the storage device 150 .

[0108] The intention tendency vector is a vector that represents the target user's tendency towards different task events in the target task. In some embodiments, the intention tendency vector can be represented by letters + numbers. For example, the intention tendency vector Q can be expressed as Q = [ABC + numbers], where the letters represent the identity information of the current user (such as quality inspectors, managers, construction workers, etc.), and the numbers represent the current user's tendency towards the task content, which can be represented by 0 or 1, 1 represents intention tendency, and 0 represents no intention tendency.

[0109] The feature fusion model is a machine learning model. For example, the feature fusion model may be one or more of a Transformer model, a neural network model, a deep neural network model, etc. In some embodiments, the processor 140 may input the intention tendency vector, the first feature vector, and the second feature vector into the feature fusion model, and the feature fusion model outputs a fused feature vector.

[0110] In some embodiments, the feature fusion model can be acquired based on training. In some embodiments, the first training sample of the feature fusion model may include a large number of historical intention tendency vectors, historical first feature vectors, and historical second feature vectors. The first label may include a feature fusion vector obtained by manual annotation. In some embodiments, when annotating to obtain the first label, the intention tendency vector can be used as the weighted weight of the first feature vector and the second feature vector. For example, if the tendency of a certain task is 1, it is more inclined to retain the vector value corresponding to the task content (corresponding to the vector value of the first feature vector) during fusion. The tendency of a certain task is 0, and it is more inclined not to retain the vector value corresponding to the task during fusion, and retain the vector value of the user identity (corresponding to retaining the vector value of the second feature vector). In some embodiments, when annotating to obtain the first label, the tendency weight can also be set according to the value of the intention tendency vector, and the values ​​of the first feature vector and the second feature vector can be weighted and fused.

[0111] In some embodiments of the present invention, by fusion of features and input of more feature vectors (such as intent tendency vectors, etc.), various types of information input into the user intent determination model can be increased, thereby improving the accuracy of model prediction.

[0112] Step 450 , based on the fused feature vector, the user intention of the current user is determined through a user intention determination model.

[0113] The user intention determination model is a machine learning model. For example, the user intention determination model can be one or more of a neural network model, a deep neural network model, and the like.

[0114] In some embodiments, the processor 140 may input the fused feature vector into a user intention determination model, and the user intention determination model outputs the user intention of the current user. For more information about user intention, see Figure 2 and its related parts.

[0115] In some embodiments, the input to the user intent determination model also includes a third eigenvector and / or a fourth eigenvector.

[0116] The third eigenvector is a vector that represents the eigenvalues ​​of the task event matrix. In some embodiments, the third eigenvector can be extracted by processing the task event matrix based on statistical methods, transformation methods (such as Fourier transform), dimensionality reduction methods (such as principal component analysis PCA), and pre-trained machine learning models. For more information about the task event matrix, see Figure 2 and its related parts.

[0117] The fourth eigenvector is a vector that represents the eigenvalues ​​of the task event state matrix. In some embodiments, the fourth eigenvector can be extracted by processing the task event state matrix based on statistical methods, transformation methods (such as Fourier transform), dimensionality reduction methods (such as principal component analysis PCA), and pre-trained machine learning models. For more information about the task event state matrix, see Figure 2 and its related parts.

[0118] In some embodiments of the present invention, the input of the user intention determination model includes the third eigenvector and / or the fourth eigenvector, which can more accurately determine the user's true intention based on the specific task event and the corresponding status, thereby improving the accuracy of the subsequent functions recommended to the user.

[0119] In some embodiments, the user intent determination model can be obtained through training. In some embodiments, the second training sample of the user intent determination model can include a large amount of historical data (such as historical first feature vectors and historical second feature vectors, etc.). The second label can include the user intent obtained by manual annotation or the user intent inferred from the function selected by the current user in history.

[0120] In some embodiments of the present invention, determining the user intention of the current user by means of a machine learning model can reduce the influence of subjective factors and make the identification of user intention more accurate.

[0121] It should be noted that the above descriptions of processes 200, 300, and 400 are only for illustration and description, and do not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to processes 200, 300, and 400 under the guidance of this specification. However, these modifications and changes are still within the scope of this specification.

[0122] Figure 5 is a schematic diagram of the internal modules of a processor according to some embodiments of the present invention. Figure 5 As shown, the processor 140 includes an acquisition module 510 , a first determination module 520 , a second determination module 530 , a third determination module 540 and a fourth determination module 550 .

[0123] The acquisition module 510 is configured to acquire a plurality of candidate tasks from the system building tasks based on the spatial location information of the current user.

[0124] The first determination module 520 is configured to obtain visual information of the current user and determine the key elements of the task based on the visual information.

[0125] The second determination module 530 is configured to determine a target task from a plurality of candidate tasks based on the key elements of the task and the identity information of the current user.

[0126] In some embodiments, the second determination module 530 is further configured to determine, for each of a plurality of candidate tasks, whether the candidate task is related to a task key element based on its corresponding task information; in response to the candidate task being related to the task key element, determine a degree of match between the candidate task and the identity information of the current user; and determine a target task based on the degree of match.

[0127] In some embodiments, the second determination module 530 is further configured to determine the task focus of the current user based on the identity information of the current user; determine the weight of each element in the task key elements based on the task focus; and perform weighted calculation on the correlation between the candidate tasks and each element in the task key elements based on the weight to determine the degree of matching.

[0128] The third determination module 540 is configured to determine the user intention of the current user based on the target task.

[0129] In some embodiments, the third determination module 540 is further configured to determine one or more task states corresponding to the target task and their corresponding task events; determine the state vector corresponding to each task state based on the task state and its corresponding task event; determine the task vector based on the state vector and the target task; determine the event vector corresponding to each task event; determine the task event matrix based on the event vector corresponding to each task event and the task vector; and determine the user intention of the current user based on the task event matrix.

[0130] In some embodiments, the third determination module 540 is further configured to assign a value to each element in the task event matrix based on the event completion status corresponding to each task event to obtain a task event status matrix; based on a preset task screening method, filter out the target event from the task event status matrix; based on the target event, determine the user intention of the current user.

[0131] In some embodiments, the third determination module 540 is further configured to obtain the task content corresponding to the target task; perform feature extraction based on the task content to obtain a first feature vector; perform feature extraction based on the identity information of the current user to obtain a second feature vector; fuse the first feature vector and the second feature vector to obtain a fused feature vector; based on the fused feature vector, determine the user intent of the current user through a user intent determination model; the user intent determination model is a machine learning model.

[0132] In some embodiments, the third determination module 540 is further configured to obtain the historical user intention of the current user; determine the intention tendency vector based on the identity information of the current user and the historical user intention; obtain a fused feature vector based on the intention tendency vector, the first feature vector and the second feature vector through a feature fusion model; the feature fusion model is a machine learning model.

[0133] The fourth determination module 550 is configured to determine a function recommended to the current user based on the user intention.

[0134] It should be noted that the above description of the processor 140 and its modules is only for convenience of description and does not limit the present invention to the scope of the embodiments. Figure 5 The acquisition module 510, the first determination module 520, the second determination module 530, the third determination module 540 and the fourth determination module 550 of the processor 140 disclosed in the disclosure may be different modules in a system, or one module may implement the functions of two or more of the above modules. For example, each module may share a storage module, or each module may have its own storage module.

[0135] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of the present invention. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements and corrections to the present invention. Such modifications, improvements and corrections are suggested in the present invention, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of the present invention.

[0136] It should be noted that in order to simplify the description of the present invention and thus facilitate the understanding of one or more embodiments, in the above description of the present invention, multiple features are sometimes grouped into one embodiment, drawing or description thereof. However, this disclosure method does not mean that the subject matter of the present invention requires more features than those mentioned in the claims. In fact, the features of an embodiment are less than all the features of a single embodiment disclosed above.

Claims

1. A method for rapid function recommendation for the construction industry, characterized in that: The method comprises: Based on the spatial location information of the current user, multiple candidate tasks are obtained from the system building tasks; Acquiring visual information of the current user, and determining key elements of the task based on the visual information; Determining a target task from the plurality of candidate tasks based on the key elements of the task and the identity information of the current user; Based on the target task, determining the user intention of the current user; Based on the user intention, a function recommended to the current user is determined.

2. The method according to claim 1, characterized in that The determining a target task from the plurality of candidate tasks based on the key task elements and the identity information of the current user includes: For each of the plurality of candidate tasks, determining whether the candidate task is related to the task key element based on the corresponding task information; In response to the candidate task being related to the task key element, determining a degree of match between the candidate task and the identity information of the current user; The target task is determined based on the matching degree.

3. The method according to claim 2, characterized in that The determining the matching degree between the candidate task and the identity information of the current user includes: Determining a task focus of the current user based on the identity information of the current user; Based on the task focus, determining the weight of each element in the task key elements; The relevance between the candidate task and each element in the task key element is weightedly calculated based on the weight to determine the matching degree.

4. The method according to claim 1, characterized in that The determining the user intention of the current user based on the target task includes: Determine one or more task states corresponding to the target task and their corresponding task events; Based on the task states and the corresponding task events, determining a state vector corresponding to each task state; Determining a task vector based on the state vector and the target task; Determine an event vector corresponding to each of the task events; Determine a task event matrix based on the event vector and the task vector corresponding to each of the task events; Based on the task event matrix, a user intention of the current user is determined.

5. The method according to claim 4, characterized in that The determining the user intention of the current user based on the task event matrix includes: Based on the event completion status corresponding to each of the task events, each element in the task event matrix is ​​assigned a value to obtain a task event status matrix; Based on a preset task screening method, screening out target events from the task event state matrix; Based on the target event, a user intent of the current user is determined.

6. The method according to claim 1, characterized in that The determining the user intention of the current user based on the target task includes: Obtaining the task content corresponding to the target task; Perform feature extraction based on the task content to obtain a first feature vector; Perform feature extraction based on the identity information of the current user to obtain a second feature vector; fusing the first feature vector and the second feature vector to obtain a fused feature vector; Based on the fused feature vector, the user intention of the current user is determined through a user intention determination model; the user intention determination model is a machine learning model.

7. The method according to claim 6, characterized in that The step of fusing the first feature vector and the second feature vector to obtain a fused feature vector includes: Obtaining the historical user intention of the current user; Determining an intention tendency vector based on the identity information of the current user and the historical user intentions; Based on the intention tendency vector, the first feature vector and the second feature vector, the fused feature vector is obtained through a feature fusion model; the feature fusion model is a machine learning model.

8. A function rapid recommendation system for the construction industry, characterized in that: The system comprises: an acquisition module, configured to acquire a plurality of candidate tasks from the system building tasks based on the spatial location information of the current user; A first determination module is configured to obtain visual information of the current user and determine a key task element based on the visual information; A second determination module is configured to determine a target task from the plurality of candidate tasks based on the key elements of the task and the identity information of the current user; A third determination module is configured to determine the user intention of the current user based on the target task; The fourth determination module is configured to determine the function recommended to the current user based on the user intention.

9. The system according to claim 8, characterized in that The third determination module is further configured to: Determine one or more task states corresponding to the target task and their corresponding task events; Based on the task states and the corresponding task events, determining a state vector corresponding to each task state; Determining a task vector based on the state vector and the target task; Determine an event vector corresponding to each of the task events; Determine a task event matrix based on the event vector and the task vector corresponding to each of the task events; Based on the task event matrix, a user intention of the current user is determined.

10. A function rapid recommendation device for the construction industry, characterized in that: The apparatus comprises: at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is used to execute at least part of the computer instructions to implement the method for rapid function recommendation for the construction industry as described in any one of claims 1 to 7.

Images