A mobile terminal-based content remote updating system

Abstract

The application relates to the technical field of electronic commerce, and discloses a content remote updating system based on a mobile terminal, which comprises a behavior graph module, an intention drift module, a content pre-fetching module, a dynamic adaptation module, a dependency arrangement module and a mounting rendering module, extracts a page jump time sequence transfer relationship of a mobile commerce terminal user behavior record, generates an electronic commerce browsing behavior topology graph, encodes a commodity category space interest migration direction gradient based on a topology graph node aggregation area and an edge density distribution, obtains an intention drift vector, extracts a matching candidate content component from a preset resource set according to the intention drift vector, generates a loading condition expression, reads terminal available memory and a downlink bandwidth, substitutes the expression into the condition expression, judges, marks a to-be-loaded component, arranges a to-be-loaded component dependency level, obtains a loading sequence and a priority identifier, mounts the to-be-loaded component to a display framework placeholder node according to the loading sequence and the priority, and obtains terminal final presentation content; and the application can improve content remote updating efficiency.

Description

Technical Field

[0001] This invention relates to the field of e-commerce technology, and in particular to a remote content update system based on a mobile terminal. Background Technology

[0002] Currently, remote content updates on mobile business terminals mostly adopt a passive request or full push operation mode. This lacks proactive prediction of user browsing behavior and interest shifts, resulting in low matching rates between pre-fetched content components and actual user interests. This easily leads to inefficient consumption of terminal memory and network bandwidth resources. Existing update mechanisms lack hierarchical arrangement of content component dependencies and unified planning of loading sequences, easily causing component loading order errors. This, in turn, causes rendering stutters in the terminal display framework, increasing the overall time required for content updates.

[0003] Existing content update solutions struggle to dynamically adjust loading strategies based on the real-time operating status of mobile terminals. They also lack the ability to set loading trigger conditions tailored to the terminal's resource status for individual content components, and their adaptation to available RAM and downlink bandwidth is coarse-grained. Mobile terminals with different hardware configurations cannot receive update content that matches their own resource capacity. Resource scheduling during the update process is not sufficiently rational, and the terminal's workload is difficult to balance, resulting in significant limitations in the overall efficiency of content updates. Therefore, improving the efficiency of remote content updates has become an urgent problem to be solved. Summary of the Invention

[0004] To achieve the above objectives, this invention provides a remote content update system based on a mobile terminal. The system includes a behavior graph module, an intent drift module, a content prefetching module, a dynamic adaptation module, a dependency orchestration module, and a mounting rendering module, wherein: The behavior graph module is used to extract the temporal transition relationship of page jump path in the user behavior record in the mobile business terminal, and generate the e-commerce browsing behavior topology map of the user in the mobile business terminal based on the temporal transition relationship. The intent drift module is used to perform directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering area and the density distribution of the transfer directed edge in the e-commerce browsing behavior topology graph, so as to obtain the intent drift vector of the user's corresponding interest change trend. The content prefetching module is used to extract candidate content components corresponding to the direction of the intent drift vector from a preset content resource set, and generate the loading condition expression for each candidate content component; The dynamic adaptation module is used to read the available running memory value and the current downlink bandwidth value of the mobile business terminal, substitute the available running memory value and the current downlink bandwidth value into the loading condition expression corresponding to the candidate content component for logical judgment, and mark the candidate content component with the judgment result as the component to be loaded. The dependency orchestration module is used to perform dependency hierarchy orchestration on the component dependency relationship identifiers of the components to be loaded, so as to obtain the loading order and loading priority identifiers of the components to be loaded. The mounting rendering module is used to mount the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier, so as to obtain the final display content of the mobile business terminal.

[0005] In a preferred embodiment, when the behavior mapping module extracts the temporal transition relationships of page jump paths from user behavior records within the mobile commerce terminal and generates an e-commerce browsing behavior topology map of the user in the mobile commerce terminal based on the temporal transition relationships, it is specifically used for: Read user behavior records from the local storage of the mobile business terminal. The user behavior records include e-commerce page access events, page identifiers and access timestamps of the e-commerce page access events. Based on the time sequence of the access timestamps, the page identifiers of two adjacent e-commerce page access events are paired to obtain the initial jump path sequence of the e-commerce page access events. The initial jump path sequence contains jump path pairs arranged in time sequence, and the jump path pairs are composed of source e-commerce page identifiers and destination e-commerce page identifiers. The number of times the same jump path pair appears in the initial jump path sequence is counted, and the number of occurrences is used as the transition weight value of the jump path pair; The jump path pairs with the same source e-commerce page identifier are grouped into the transfer set of the source e-commerce page identifier to obtain the weighted jump transfer set of the user behavior record. Using the e-commerce page identifiers of the weighted jump transfer set jump path pairs as topology nodes, the weighted jump transfer set jump path pairs as directed edges, and the transfer weight values ​​of the weighted jump transfer set jump path pairs as edge weight values, a topology graph of the user's e-commerce browsing behavior in the mobile business terminal is constructed.

[0006] In a preferred embodiment, when the intent drift module performs directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering region and the density distribution of the transition directed edges in the e-commerce browsing behavior topology graph to obtain the intent drift vector of the user's corresponding interest change trend, it is specifically used for: The structure of the page identifier corresponding to the topology node in the e-commerce browsing behavior topology graph is parsed to obtain the product category field of the page identifier, and the product category code corresponding to the page identifier is extracted from the product category field; Based on the product category code, the topological nodes with the same product category code in the page identifier are merged into homogeneous nodes to obtain the user's product category node cluster; The ratio between the total number of topological nodes in the product category node cluster and the sum of the edge weights of all directed edges in the product category node cluster is used as the browsing cluster density value of the product category node cluster, and the browsing cluster density value is used as the initial value of the interest intensity of the product category node cluster. Based on the edge weights of the directed edges between the product category node clusters, the directional weights of adjacent product category node clusters are accumulated to obtain the transition strength value of the adjacent product category node clusters. The product category node cluster is used as an interest node; The initial value of the interest intensity of the interest node is used as the node quality value of the interest node, the transfer intensity value between adjacent interest nodes is used as the connection intensity value between adjacent interest nodes, and the interest nodes are connected sequentially according to the magnitude of the transfer intensity value to obtain the directional gradient path of the interest node. Along the direction of the directional gradient path, the directional gradient path is oriented and encoded to obtain the intention drift vector of the user's corresponding interest change trend.

[0007] In a preferred embodiment, when the content prefetching module extracts candidate content components corresponding to the direction of the intent drift vector from a preset content resource set based on the intent drift vector, and generates the loading condition expression for each candidate content component, it is specifically used for: The intention drift vector is parsed to extract the target product category code corresponding to the direction pointed to by the intention drift vector in the user's corresponding product category node cluster, and the target category code set of the intention drift vector is obtained. Based on the product category codes in the target category code set, retrieve content components associated with the product category codes from the preset content resource set, and aggregate and merge the content components to obtain an initial candidate component set for the content components; Based on the vector length value of the intention drift vector, threshold filtering is performed on the content components in the initial candidate component set, retaining the content components that match the vector length value and removing the content components that do not match, to obtain the candidate content component set of the initial candidate component set; Generate corresponding loading condition expressions for the candidate content components in the candidate content component set.

[0008] In a preferred embodiment, when the content prefetching module generates corresponding loading condition expressions for candidate content components in the candidate content component set, it is specifically used for: Resource constraint evaluation is performed on the candidate content components in the candidate content component set to obtain the minimum memory requirement and minimum bandwidth requirement of the candidate content components; The memory determination sub-condition is that the available running memory value of the mobile business terminal is greater than or equal to the minimum memory requirement value, and the bandwidth determination sub-condition is that the current downlink bandwidth value of the mobile business terminal is greater than or equal to the minimum bandwidth requirement value. By performing a logical conjunction operation on the memory determination sub-condition and the bandwidth determination sub-condition, the loading condition expression of the candidate content component is obtained.

[0009] In a preferred embodiment, when the dynamic adaptation module reads the available RAM and current downlink bandwidth of the mobile business terminal, substitutes the available RAM and current downlink bandwidth into the loading condition expression corresponding to the candidate content component for logical determination, and marks the candidate content component with a true determination result as a component to be loaded, it is specifically used for: The mobile business terminal is sampled to obtain its available running memory value and current downlink bandwidth value. The candidate content components are traversed, and the conditions of the current candidate content component are parsed to obtain the loading condition expression of the current candidate content component. Based on the memory determination sub-condition and bandwidth determination sub-condition in the loading condition expression, the available running memory value is substituted into the memory determination sub-condition to perform memory determination, thereby obtaining the memory determination result of the current candidate content component. The current downlink bandwidth value is substituted into the bandwidth determination sub-condition to perform bandwidth determination, thereby obtaining the bandwidth determination result of the current candidate content component. When the memory determination result is true and the bandwidth determination result is true, the currently traversed candidate content component is marked as a component to be loaded; when the memory determination result is false or the bandwidth determination result is false, the currently traversed candidate content component is marked as a non-component to be loaded. The process continues until all candidate content components have been traversed, at which point the component to be loaded is output.

[0010] In a preferred embodiment, when the dependency orchestration module performs dependency hierarchy orchestration on the component dependency identifiers of the components to be loaded to obtain the loading order and loading priority identifiers of the components to be loaded, it is specifically used for: The component to be loaded is associated with a search to obtain the component dependency identifier of the component to be loaded. The component dependency identifier contains a list of component identifiers of other components that the component to be loaded depends on. Using the components to be loaded as graph nodes and the dependency directions in the component dependency identifiers as directed dependency edges, a directed dependency graph of the list of components to be loaded is constructed, wherein the direction of the directed dependency edges is from the dependent component to the dependent component. The directed dependency graph is topologically sorted, and the components to be loaded corresponding to the nodes with an in-degree of zero in the topological sort are arranged into the initial loading level. The components to be loaded in the initial loading level are arranged into the next loading level, thus obtaining the loading level number of the components to be loaded. The loading time distribution of the historical loading time records of the components to be loaded is estimated to obtain the estimated loading time value of the components to be loaded. The estimated loading time value and the loading level number are weighted and arranged to obtain the loading order of the components to be loaded. Based on the loading level number, the basic priority value of the component to be loaded is determined, and the basic priority value is used as the loading priority identifier of the component to be loaded.

[0011] In a preferred embodiment, when the dependency orchestration module performs time distribution estimation on the historical loading time records of the components to be loaded to obtain the estimated loading time value of the components to be loaded, and then assigns weights to the estimated loading time value and the loading level number to obtain the loading order of the components to be loaded, it is specifically used for: Read the historical loading time record of the component to be loaded from the component attribute record of the component to be loaded. The historical loading time record contains the actual loading time value recorded by the component to be loaded in the historical loading operation. Arrange the actual loading time values ​​of the components to be loaded in chronological order to obtain the loading time sequence of the components to be loaded; A central tendency analysis is performed on the time series, and the average of the mode estimated time and the median estimated time in the time series is used as the initial estimated loading time of the component to be loaded. Using the loading level number of the component to be loaded as the level weight coefficient, the product of the initial estimated loading time value and the level weight coefficient is used as the weighted estimated loading time value of the component to be loaded. Based on the weighted estimated loading time value, the components to be loaded are sorted to obtain the loading order of the components to be loaded. The loading order of the components to be loaded is determined based on the loading order position and the loading level number.

[0012] In a preferred embodiment, the formula for calculating the weighted estimated loading time is as follows: ; In the formula, The number of the component to be loaded is its sequence number in the list of components to be loaded. The first component to be loaded The loading sequence number of each component to be loaded in the history of loading operations. The first component to be loaded The weighted estimated loading time of each component to be loaded. The first component to be loaded The total number of historical loading operations experienced by each component to be loaded. The first component to be loaded The first component to be loaded in history The actual loading time value recorded in this loading operation The first component to be loaded The layer weight coefficient corresponding to the loading layer number of each component to be loaded.

[0013] In a preferred embodiment, when the mounting rendering module mounts the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier to obtain the final content displayed by the mobile business terminal, it is specifically used for: Extract the set of placeholder nodes from the display frame in the mobile business terminal; The component identifier of the component to be loaded is matched with the node identifier in the placeholder node set to obtain the target placeholder node of the component to be loaded. Based on the loading priority identifier, the components to be loaded in the target placeholder node are sorted by priority to obtain the component loading sequence of the target placeholder node; Based on the component loading sequence, the cross-node loading order of the components to be loaded in the node is arranged to obtain the global loading sequence of the components to be loaded in the node; Based on the global loading sequence, the node components to be loaded are sequentially mounted onto the corresponding target placeholder nodes, and after each node component to be loaded is mounted, a node status update signal is sent to the display frame. Once all node status update signals of the target placeholder node have been received, the rendering instruction of the display frame is triggered to merge and render the component content on the placeholder node after it has been mounted, thus obtaining the final presentation content of the mobile business terminal.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention relies on e-commerce browsing behavior topology maps and intent drift vectors to accurately pinpoint user interest directions, completing targeted filtering of candidate content components. This reduces the transmission and loading process of irrelevant content components, decreases terminal network resource consumption, and simultaneously reduces the ineffective use of mobile commerce terminal memory resources, stabilizing terminal operation. Resource adaptation is determined based on loading condition expressions, retaining only components that meet the terminal's real-time memory and bandwidth requirements, ensuring that resource scheduling in the content pre-fetching stage aligns with the terminal's real-time hardware capabilities.

[0015] 2. This invention completes the dependency hierarchy orchestration of components to be loaded, relying on a weighted estimated loading time to establish a uniform loading order and priority identifier, thus avoiding the problem of disordered component loading sequence and shortening the overall loading time. Components are mounted and rendered in an orderly manner according to the global loading sequence, and node status update signals are pushed step by step to complete the merged rendering, ensuring the smoothness of the final rendered content, stabilizing the efficiency of the entire process of remote content updates, and improving the completeness of the terminal page display. Attached Figure Description

[0016] Figure 1 This is a system architecture diagram of a mobile terminal-based remote content update system provided in an embodiment of the present invention.

[0017] 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

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments belong to some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0020] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0021] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0022] In practice, the server-side equipment deployed in a mobile terminal-based remote content update system may consist of one or more devices. This mobile terminal-based remote content update system can be implemented as: a business instance, a virtual machine, or hardware devices. For example, this mobile terminal-based remote content update system can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, this mobile terminal-based remote content update system can be understood as software deployed on a cloud node, used to provide a mobile terminal-based remote content update system for various user terminals. Alternatively, this mobile terminal-based remote content update system can also be implemented as a virtual machine deployed on one or more devices in a cloud node. This virtual machine contains application software for managing various user terminals. Alternatively, this mobile terminal-based remote content update system can also be implemented as a server composed of numerous identical or different types of hardware devices, with one or more hardware devices configured to provide a mobile terminal-based remote content update system for various user terminals.

[0023] In terms of implementation, a mobile terminal-based remote content update system and a user client are mutually compatible. That is, if a mobile terminal-based remote content update system is implemented as an application installed on a cloud service platform, then the user client is implemented as a client that establishes a communication connection with the application; or if a mobile terminal-based remote content update system is implemented as a website, then the user client is implemented as a webpage; or if a mobile terminal-based remote content update system is implemented as a cloud service platform, then the user client is implemented as a mini-program in an instant messaging application.

[0024] like Figure 1The diagram shown is a system architecture diagram of a mobile terminal-based remote content update system provided in an embodiment of the present invention.

[0025] The mobile terminal-based remote content update system 100 described in this invention can be located on a cloud server. In terms of implementation, it can function as one or more service devices, or as an application installed on the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed as a website. Depending on the implemented functions, the mobile terminal-based remote content update system 100 may include a behavior graph module 101, an intent drift module 102, a content prefetching module 103, a dynamic adaptation module 104, a dependency orchestration module 105, and a mounting rendering module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0026] In this embodiment of the invention, in a mobile terminal-based remote content update system, each of the above modules can be implemented independently and can call other modules. Here, "calling" can be understood as a module connecting to multiple modules of another type and providing corresponding services to those connected modules. The mobile terminal-based remote content update system provided by this embodiment of the invention allows for adjustments to the applicability of the system architecture without modifying the program code. This is achieved by adding modules and directly calling them, enabling cluster-based horizontal expansion and flexibly expanding the mobile terminal-based remote content update system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in a virtual device, such as a service instance in a cloud server.

[0027] The following describes, with reference to specific embodiments, the various components and specific workflows of a mobile terminal-based remote content update system: The behavior graph module 101 is used to extract the temporal transition relationship of page jump path in the user behavior record in the mobile business terminal, and generate the e-commerce browsing behavior topology map of the user in the mobile business terminal based on the temporal transition relationship. In this embodiment of the invention, when the behavior graph module extracts the temporal transition relationship of page jump paths in the user behavior records within the mobile commerce terminal, and generates an e-commerce browsing behavior topology map of the user in the mobile commerce terminal based on the temporal transition relationship, it is specifically used for: The system reads user behavior records from the local storage area built into the mobile business terminal. These records contain e-commerce page access events, and each event includes a page identifier and an access timestamp.

[0028] All e-commerce page access events are arranged in chronological order according to the system time stamps of the access timestamps. The page identifiers of two adjacent e-commerce page access events are paired, with the page identifier of the previous access event serving as the source page identifier and the page identifier of the next access event serving as the destination page identifier, thus forming a jump path pair. After processing all adjacent pairings in this way, the initial jump path sequence is obtained.

[0029] Iterate through all jump path pairs in the initial jump path sequence, and accumulate the occurrence count of jump path pairs where the source page identifier and the destination page identifier are exactly the same. Use the accumulated occurrence count as the transfer weight value of the jump path pair.

[0030] All redirection paths are grouped according to the source page identifier. All redirection paths corresponding to the same source page are included in the transfer set corresponding to that source page identifier, resulting in a weighted redirection transfer set of user behavior records.

[0031] Using the e-commerce page identifiers contained in all jump path pairs in the weighted jump transfer set as topology nodes, the page jump direction of the jump path pair as the directed edge between topology nodes, and the transfer weight value of the jump path pair as the edge weight value of the corresponding directed edge, all topology nodes and directed edges are connected and combined according to the corresponding jump relationship to construct the topology graph of the user's e-commerce browsing behavior in the mobile business terminal.

[0032] The beneficial effect is that by structurally deconstructing and weighting the temporal jump relationships of user browsing behavior, it is possible to completely reconstruct the user's e-commerce page browsing trajectory, forming a reusable browsing behavior topology structure, and providing stable data support for subsequent determination of user interest direction.

[0033] As another optional implementation of this embodiment, in addition to using the adjacent pairing jump path generation method, a sliding window covering 10 consecutive e-commerce page access events can be set up. The first event in the window is paired with the page identifiers of the second to tenth events in the window, respectively. Then, the window is moved forward 5 event positions, and the above pairing operation is repeated until all events are covered. The extended jump path sequence is obtained by summing all the pairing results generated in all windows. Compared with adjacent pairing, this method can capture indirect jump patterns with multiple access steps in between, and is suitable for scenarios where users frequently backtrack or jump across levels in their browsing behavior.

[0034] As another optional implementation of this embodiment, in addition to counting the occurrences of the same redirect path pair, a decay coefficient based on time difference can be assigned to each occurrence. This decay coefficient is 1.0 when the time difference is 0 minutes, decays to 0.5 when the time difference reaches 30 minutes, decays to 0.1 when the time difference reaches 120 minutes, and decreases by 0.15 to 0.20 for every 30-minute increase in time difference. The decay coefficients of each occurrence are summed to obtain a decay weighted value, which is used as the transfer weight value of the redirect path pair. This statistical method enables recently visited redirect path pairs to receive higher transfer weight values, and is suitable for scenarios that need to highlight users' recent browsing preferences, enabling faster response to the dynamic changes in e-commerce users' interests over time.

[0035] The intent drift module 102 is used to perform directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering area and the density distribution of the transfer directed edge in the e-commerce browsing behavior topology graph, so as to obtain the intent drift vector of the user's corresponding interest change trend. In this embodiment of the invention, when the intent drift module performs directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering region and the density distribution of the transition directed edges in the e-commerce browsing behavior topology graph to obtain the intent drift vector of the user's corresponding interest change trend, it is specifically used for: The structure of the page identifier corresponding to each topology node in the e-commerce browsing behavior topology diagram is parsed. The page identifier is a string generated according to a fixed hierarchy rule, with a total length of 32 bytes. The 5th to 12th character segments constitute the product category field. After parsing, the product category code corresponding to the product category is extracted from this field.

[0036] Topological nodes with identical product category codes are grouped into the same set. The set retains all directed edges and edge weights of the corresponding topological nodes, thus obtaining the user's product category node cluster.

[0037] The total number of topological nodes within a single product category node cluster is counted, and the sum of the edge weights of all directed edges within the cluster is then calculated. The ratio of the total number of topological nodes to the sum of edge weights is used as the browsing cluster density value of the cluster, and this density value is directly used as the initial value of the interest intensity of the cluster.

[0038] Two product category node clusters that are directly connected by a directed edge are identified as adjacent clusters. Along the direction of the directed edge, the weight values ​​of all directed edges from the starting cluster to the ending cluster are accumulated, and the accumulated value is used as the transition strength value between adjacent clusters.

[0039] Each product category node cluster is mapped to an independent interest node. The node inherits the initial interest intensity value of the corresponding cluster and the transition strength value with neighboring nodes. The initial interest intensity value is used as the node quality value, and the transition strength value between neighboring nodes is used as the connection strength value.

[0040] The node of interest with the highest node quality value is selected as the starting point of the path. Starting from the starting point, the adjacent unconnected nodes with the highest connection strength value are selected in sequence to connect to the path. If the connection strength values ​​of multiple adjacent nodes are all lower than the preset threshold of 0.3, the connection is stopped. After the connection is completed, the directional gradient path is obtained.

[0041] Along the direction from the start node to the end node, the path direction is used as the vector direction, and the sum of all connection strength values ​​within the path is used as the vector length attribute. After encoding, the intention drift vector corresponding to the user's interest change trend is obtained.

[0042] The beneficial effect is that by aggregating user browsing behavior topology at the category level and constructing gradient paths, it is possible to quantitatively represent the user's interest migration trend in the product category space, forming an intent drift vector that can be directly used for content matching, providing accurate directional basis for the subsequent content pre-fetching stage, and ensuring the degree of matching between content selection and user interest trends.

[0043] The content prefetching module 103 is used to extract candidate content components corresponding to the direction of the intent drift vector from a preset content resource set according to the intent drift vector, and generate the loading condition expression for the candidate content components respectively; In this embodiment of the invention, when the content prefetching module extracts candidate content components corresponding to the direction of the intent drift vector from a preset content resource set according to the intent drift vector, and generates the loading condition expression for each candidate content component, it is specifically used for: The intention drift vector is analyzed for direction. The analysis is based on a pre-established spatial orientation mapping table of product category node clusters, which covers 28 product category node clusters, each cluster corresponding to fixed orientation coordinates in the category space. The target product category codes corresponding to the product category node clusters covered by the vector's direction are extracted, and all extracted target codes constitute the target category code set.

[0044] Based on the codes in the target category code set, content components that are bound to the codes are retrieved from the preset content resource set. The preset content resource set is a remotely updatable collection of components categorized and archived according to product category dimensions. Each content component in the set is bound to a corresponding product category code and its own resource attribute information. All retrieved and matched components are aggregated, merged, and duplicates are removed to obtain an initial candidate component set.

[0045] Based on the vector length value of the intended drift vector, the initial candidate component set is threshold-filtered according to a preset threshold. The preset threshold is a set of rules corresponding to the vector length range and the component resource size admission standard: the admission limit is 5MB when the vector length value is in the range of 1 to 3, the admission limit is 12MB when the vector length value is in the range of 4 to 6, and the admission limit is 20MB when the vector length value is in the range of 7 to 10. Components whose own resource size is within the corresponding admission limit range are retained, and those that exceed the range are removed, resulting in the candidate content component set.

[0046] Resource constraint assessment is performed on each candidate content component in the candidate content component set. The pre-defined file size and runtime memory usage of the component are read, and the minimum runtime memory capacity required for normal loading of the component is derived according to the terminal runtime memory usage conversion rules. This capacity is the minimum memory requirement value.

[0047] Evaluate the transmission resource constraints of candidate content components. Read the total file data size and transmission fragmentation attribute data of the component, and derive the minimum downlink transmission rate required for the complete download of the component according to the network downlink transmission rate conversion rules. This rate is the minimum bandwidth requirement.

[0048] A memory determination sub-condition is used, where the available RAM of the mobile business terminal is greater than or equal to the minimum memory requirement, to determine whether the terminal's remaining memory meets the basic requirements for component loading and operation. A bandwidth determination sub-condition is used, where the terminal's current downlink bandwidth is greater than or equal to the minimum bandwidth requirement, to determine whether the terminal's network downlink transmission capacity meets the requirements for complete component download.

[0049] The memory determination sub-condition and the bandwidth determination sub-condition are logically combined. If both conditions are true, the overall determination result is true. After the operation is completed, the loading condition expression of the candidate content component is obtained.

[0050] The beneficial effects are that the target category range is locked by the direction dimension of the intent drift vector, and the scale filtering is completed by combining the threshold corresponding to the vector length. The content range corresponding to the direction and intensity of user interest migration is accurately matched. At the same time, the minimum memory requirement value and minimum bandwidth requirement value are derived based on the actual attributes of the component. The generated loading condition expression has a clear judgment benchmark, which provides a clear execution basis for subsequent terminal resource adaptation and screening, ensuring the accuracy of content prefetching and resource adaptability.

[0051] The dynamic adaptation module 104 is used to read the available running memory value and the current downlink bandwidth value of the mobile business terminal, substitute the available running memory value and the current downlink bandwidth value into the loading condition expression corresponding to the candidate content component for logical judgment, and mark the candidate content component with the judgment result as the component to be loaded. In this embodiment of the invention, when the dynamic adaptation module reads the available running memory value and the current downlink bandwidth value of the mobile business terminal, substitutes the available running memory value and the current downlink bandwidth value into the loading condition expression corresponding to the candidate content component for logical determination, and marks the candidate content component with the determination result as a component to be loaded, it is specifically used for: Status sampling is performed by calling the standard hardware status reading interface of the mobile business terminal operating system. Sampling is executed at the start of the decision-making process after the candidate content component set is generated, and the response time of a single interface call does not exceed 200 milliseconds. The interface directly reads the total amount of unused running memory in the system as the available running memory value, and simultaneously reads the real-time downlink data transmission rate of the current network link as the current downlink bandwidth value.

[0052] The components are traversed one by one according to the pre-defined storage arrangement order of the candidate content component set. The loading condition expression structure of the currently traversed component is parsed, and the memory determination sub-condition and bandwidth determination sub-condition inside the expression are extracted.

[0053] The available running memory value obtained from the sampling is compared with the minimum memory requirement value marked in the sub-condition to determine whether the relationship of being greater than or equal to is satisfied, thus obtaining the memory determination result. The current downlink bandwidth value is compared with the minimum bandwidth requirement value marked in the sub-condition to determine whether the relationship of being greater than or equal to is satisfied, thus obtaining the bandwidth determination result.

[0054] When both the memory and bandwidth checks are true, a "pending loading" status flag is added to the component, marking it as a component to be loaded. When either check is false, a "not pending loading" status flag is added, marking it as a component not to be loaded. This process of traversal, parsing, and marking continues until all components have been processed. All components with the "pending loading" flag are then aggregated and output uniformly, resulting in a complete list of components that meet the terminal's operating conditions and are ready to be loaded.

[0055] The beneficial effects are that accurate real-time hardware operating parameters are obtained by sampling through the official terminal interface, and memory and bandwidth are determined item by item in two dimensions by combining loading condition expressions. This accurately selects components that are suitable for the current operating state of the terminal, avoids components that exceed the carrying capacity from entering the loading process, ensures the compatibility of content updates with terminal hardware resources, and maintains the stability of terminal operation.

[0056] As another optional implementation of this embodiment, three consecutive state sampling operations are performed before each determination is initiated, with each sampling interval being 500 milliseconds. The average of the available running memory values ​​obtained from the three samplings is used as the representative value of available running memory for that determination period, and the average of the current downlink bandwidth values ​​obtained from the three samplings is used as the representative value of current downlink bandwidth. The representative values ​​are used instead of single sample values ​​for determination. This method can filter out the influence of instantaneous fluctuations in the terminal's operating state, obtaining parameter values ​​that better reflect the actual average carrying capacity of the terminal. It is suitable for scenarios where the terminal's operating state fluctuates greatly and a single sample cannot accurately represent the actual state.

[0057] The dependency orchestration module 105 is used to perform dependency hierarchy orchestration on the component dependency relationship identifiers of the components to be loaded, so as to obtain the loading order and loading priority identifiers of the components to be loaded. In this embodiment of the invention, when the dependency orchestration module performs dependency hierarchy orchestration on the component dependency relationship identifiers of the components to be loaded to obtain the loading order and loading priority identifiers of the components to be loaded, it is specifically used for: Iterate through all components to be loaded, retrieve the prerequisite dependency information of each component from the pre-established component attribute information library, and obtain the component dependency relationship identifier of the component to be loaded. The component dependency relationship identifier contains a list of component identifiers of other components that must be loaded in advance for the component to be loaded to run normally.

[0058] Each component to be loaded is treated as an independent graph node. The direction of the directed dependency edge is determined by the dependency pointer recorded in the component dependency identifier. The direction of the directed dependency edge is from the dependent component to the component to be loaded that depends on it. All nodes are associated and combined with their corresponding directed edges to construct a directed dependency graph of the list of components to be loaded.

[0059] The directed dependency graph is hierarchically divided. First, all nodes without incoming edges are selected as the components to be loaded, and they are arranged into the initial loading level. Then, all nodes and their corresponding outgoing edges in this level are removed. The process of selecting nodes without incoming edges and arranging them into the next loading level is repeated until all nodes are hierarchically divided, and the loading level number of the components to be loaded is obtained.

[0060] The historical loading time records of each component to be loaded are read from the component property information library. The historical loading time records are the actual running data automatically stored after each component loading operation is completed, including the actual loading time value recorded by the component to be loaded in each historical loading operation.

[0061] Arrange all the actual loading time values ​​corresponding to a single component to be loaded in chronological order of the loading operations to obtain the loading time sequence of the component to be loaded.

[0062] The central tendency of the time series is statistically analyzed. The actual loading time value that appears most frequently in the statistical series is taken as the mode estimated time value, and the actual loading time value in the middle position of the series is taken as the median estimated time value. The two values ​​are averaged to obtain the initial estimated loading time value of the component to be loaded.

[0063] The loading level number corresponding to the component to be loaded is used as the level weight coefficient. The value of the level weight coefficient corresponds one-to-one with the value of the loading level number. The initial estimated loading time value is multiplied by the corresponding level weight coefficient to obtain the weighted estimated loading time value of the component to be loaded.

[0064] The layer weight coefficient is a pre-set weight value that corresponds one-to-one with the loading layer number. This value is based on the engineering experience of prioritizing the loading of underlying dependent components. The smaller the loading layer number, the closer the corresponding component is to the underlying dependent component, and the larger the corresponding layer weight coefficient value. Each loading layer number corresponds to a unique and fixed layer weight coefficient.

[0065] The formula for calculating the weighted estimated loading time is as follows: ; The sequence number of the component to be loaded in the list of components to be loaded The unique sequential number is assigned to each component after the list of components to be loaded is generated, according to a predetermined storage arrangement, and is used to identify different individual components to be loaded in the list.

[0066] The first component to be loaded The loading sequence number of each component to be loaded in historical loading operations. , from the The historical loading time records of each component to be loaded are sorted according to the order in which the loading operations occurred, and a unique sequential number is assigned to each loading operation to distinguish different historical loading records of the same component.

[0067] The first component to be loaded The total number of historical loading operations experienced by each component to be loaded. This total number of entries is derived from the total number of valid historical loading time records read from the component property records of the component to be loaded. Each record corresponds to one complete historical loading execution process. The total number of valid records is counted to obtain this total number of times.

[0068] The first component to be loaded The first component to be loaded in history The actual loading time recorded in the next loading operation , from the After each historical loading operation is completed, the official timing interface built into the mobile business terminal system automatically collects and stores the actual loading time data in the component attribute record. Each loading operation corresponds to a unique actual loading time value, and the collection standard is uniform and fixed.

[0069] The first component to be loaded The hierarchy weight coefficient corresponding to the loading hierarchy number of each component to be loaded. It originates from the pre-defined mapping rules between loading levels and weights. These rules are based on the engineering experience of prioritizing the loading of underlying dependent components. The smaller the loading level number, the larger the corresponding level weight coefficient. Each loading level number corresponds to a unique and fixed level weight coefficient value.

[0070] The first component to be loaded Weighted estimated loading time of each component to be loaded The final result of the calculation is used as the core criterion for determining the loading order of components to be loaded.

[0071] The components to be loaded within the same loading level are sorted in ascending order according to the weighted estimated loading time value to obtain the loading order position of the components within their respective levels.

[0072] The components to be loaded are arranged in ascending order of their loading level numbers. Within the same level, they are arranged according to their loading order position, forming a complete loading sequence to obtain the loading order of the components to be loaded.

[0073] The basic priority value of the component to be loaded is determined based on the loading level number. The smaller the loading level number, the higher the basic priority value. The basic priority value is used as the loading priority identifier of the component to be loaded.

[0074] The beneficial effects are that by constructing a directed dependency graph, the component dependency hierarchy is divided in an orderly manner, the pre-loading conditions of each component to be loaded are clarified, component operation abnormalities caused by disordered loading order are avoided, and the weighted sorting is completed by combining historical loading time data to optimize the loading order of components at the same level, thereby improving the overall loading process efficiency and providing a clear and orderly execution basis for subsequent mounting and rendering.

[0075] As another optional implementation of this embodiment, in addition to topologically sorting the directed dependency graph and arranging nodes with an in-degree of zero as the initial loading level, after constructing the directed dependency graph, the total path length from each graph node to all its dependent descendant nodes can be calculated. Graph nodes whose total path length reaches a preset threshold are marked as key nodes. All key nodes are arranged in descending order of total path length to obtain a key node sequence. Key nodes in the key node sequence are preferentially arranged in the initial loading level and subsequent levels close to the initial loading level. Non-key nodes are arranged in the remaining levels according to the topological sorting result. Compared with simple topological sorting, this arrangement method ensures that key components with long dependency chains and large impact ranges receive higher priority loading positions, making it suitable for application scenarios with complex dependencies between components and large differences in dependency depth.

[0076] The mounting rendering module 106 is used to mount the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier, so as to obtain the final presentation content of the mobile business terminal.

[0077] In this embodiment of the invention, when the mounting rendering module mounts the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier to obtain the final content displayed by the mobile business terminal, it is specifically used for: The layout configuration file of the mobile commerce terminal's built-in display frame is read. The display frame is the standard layout container for rendering e-commerce pages on the terminal. The configuration file is a predefined page structure description file, which marks all reserved container nodes used to host external content components. All reserved container nodes are extracted and aggregated to obtain a placeholder node set. A typical e-commerce page has 3 to 5 reserved container nodes, corresponding to the top product recommendation slot, the middle promotional content slot, and the bottom function extension slot, respectively.

[0078] The unique component identifier of each component to be loaded is read. The component identifier is then compared one by one with the node identifiers of each node in the placeholder node set according to the key-value mapping rules. The placeholder node that matches successfully is the target placeholder node for the corresponding component to be loaded. The key-value mapping rules are preset to establish a fixed matching relationship between 12 types of functional components and their corresponding reserved container nodes.

[0079] Collect all the components to be loaded for a single target placeholder node, read the loading priority identifier of each component, and sort them in descending order of priority value. The higher the priority value, the earlier the corresponding position in the sorting. After sorting, the component loading sequence of the target placeholder node is obtained.

[0080] Based on the hierarchical priority rules corresponding to the loading order, the component loading sequences of all target placeholder nodes are integrated across nodes. The integration rule is: components at higher loading levels are loaded first, and within the same level, they are executed according to the order of their loading sequences. Assuming a total of 8 components to be loaded are distributed across 3 loading levels, with 3 components in level 1, 3 components in level 2, and 2 components in level 3, then after integration, the 3 components in level 1 are executed first, followed by the 3 components in level 2, and finally the 2 components in level 3, resulting in the global loading sequence.

[0081] Following the global loading sequence, the content data of each component is written into the content container of the corresponding target placeholder node to complete the mounting. After the mounting operation of each component is completed, the status update signal of the corresponding node is sent to the status listening channel of the display framework. The status listening channel is the node status receiving port pre-set by the display framework, and the signal reception confirmation response time is within 50 milliseconds.

[0082] The display frame's status monitoring channel is continuously monitored. Once the status update signals corresponding to all target placeholder nodes are successfully received, a rendering instruction is sent to the rendering execution unit of the display frame. The rendering execution unit merges and renders the content of all components mounted in the placeholder nodes according to the page layout rules. The execution time of a single rendering instruction is approximately 100 milliseconds. After rendering is completed, the final content presented on the mobile business terminal is obtained.

[0083] The beneficial effects are that key-value matching can accurately locate the component mounting position, combined with the global loading sequence to achieve orderly component mounting, and the batch feedback mechanism of state update signals can trigger unified merging rendering, avoiding page flickering caused by the sequential rendering of single components, ensuring the integrity of the final content and the smoothness of rendering, and improving the page display effect after content update.

[0084] As another optional implementation of this embodiment, a dynamic priority value is set at each target placeholder node, with an initial value of 0. This value is incremented by 1 each time a component is attached to the node. The dynamic priority value and the component's loading priority flag are weighted and summed in a 7:3 ratio to obtain the overall loading priority of the target placeholder node. Components mapped to each target placeholder node are then interleaved and attached according to their overall loading priority from high to low, with components on nodes with higher priority being attached first. This method allows for a more balanced loading progress among multiple target placeholder nodes and is suitable for scenarios requiring multiple page areas to be displayed quickly and simultaneously.

[0085] As another optional implementation of this embodiment, after receiving a node state update signal, it checks whether there is a dependency relationship between the component corresponding to the target placeholder node and the component corresponding to a node that has not yet received a signal. If no dependency relationship exists, local rendering of that node is triggered independently. If a dependency relationship exists, it is not triggered temporarily, and the merging rendering of the relevant nodes is triggered only after the state update signal of the node containing the dependent component is received. Compared with global merging rendering, this method allows non-dependent components to be presented to the user earlier, which is suitable for scenarios with high requirements for the speed of the first screen display.

[0086] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0087] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A content remote update system based on a mobile terminal, characterized in that, The system includes a behavior graph module, an intent drift module, a content prefetching module, a dynamic adaptation module, a dependency orchestration module, and a mounting and rendering module, wherein: The behavior graph module is used to extract the temporal transition relationship of page jump path in the user behavior record in the mobile business terminal, and generate the e-commerce browsing behavior topology map of the user in the mobile business terminal based on the temporal transition relationship. The intent drift module is used to perform directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering area and the density distribution of the transfer directed edge in the e-commerce browsing behavior topology graph, so as to obtain the intent drift vector of the user's corresponding interest change trend. The content prefetching module is used to extract candidate content components corresponding to the direction of the intent drift vector from a preset content resource set, and generate loading condition expressions for the candidate content components respectively. The dynamic adaptation module is used to read the available running memory value and the current downlink bandwidth value of the mobile business terminal, substitute the available running memory value and the current downlink bandwidth value into the loading condition expression corresponding to the candidate content component for logical judgment, and mark the candidate content component with the judgment result as the component to be loaded. The dependency orchestration module is used to perform dependency hierarchy orchestration on the component dependency relationship identifiers of the components to be loaded, so as to obtain the loading order and loading priority identifiers of the components to be loaded. The mounting rendering module is used to mount the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier, so as to obtain the final display content of the mobile business terminal.

2. The content remote update system based on a mobile terminal as described in claim 1, characterized in that, When the behavior mapping module extracts the temporal transition relationships of page jump paths from user behavior records within the mobile commerce terminal and generates an e-commerce browsing behavior topology map of the user in the mobile commerce terminal based on the temporal transition relationships, it is specifically used for: Read user behavior records from the local storage of the mobile business terminal. The user behavior records include e-commerce page access events, page identifiers and access timestamps of the e-commerce page access events. Based on the time sequence of the access timestamps, the page identifiers of two adjacent e-commerce page access events are paired to obtain the initial jump path sequence of the e-commerce page access events. The initial jump path sequence contains jump path pairs arranged in time sequence, and the jump path pairs are composed of source e-commerce page identifiers and destination e-commerce page identifiers. The number of times the same jump path pair appears in the initial jump path sequence is counted, and the number of occurrences is used as the transition weight value of the jump path pair; The jump path pairs with the same source e-commerce page identifier are grouped into the transfer set of the source e-commerce page identifier to obtain the weighted jump transfer set of the user behavior record. Using the e-commerce page identifiers of the weighted jump transfer set jump path pairs as topology nodes, the weighted jump transfer set jump path pairs as directed edges, and the transfer weight values ​​of the weighted jump transfer set jump path pairs as edge weight values, a topology graph of the user's e-commerce browsing behavior in the mobile business terminal is constructed.

3. The content remote update system based on a mobile terminal as described in claim 1, characterized in that, When the intent drift module performs directional gradient encoding on the interest migration direction in the user's corresponding product category space based on the browsing node clustering region and the density distribution of the transitioning directed edges in the e-commerce browsing behavior topology graph to obtain the intent drift vector of the user's corresponding interest change trend, it is specifically used for: The structure of the page identifier corresponding to the topology node in the e-commerce browsing behavior topology graph is parsed to obtain the product category field of the page identifier, and the product category code corresponding to the page identifier is extracted from the product category field; Based on the product category code, the topological nodes with the same product category code in the page identifier are merged into homogeneous nodes to obtain the user's product category node cluster; The ratio between the total number of topological nodes in the product category node cluster and the sum of the edge weights of all directed edges in the product category node cluster is used as the browsing cluster density value of the product category node cluster, and the browsing cluster density value is used as the initial value of the interest intensity of the product category node cluster. Based on the edge weights of the directed edges between the product category node clusters, the directional weights of adjacent product category node clusters are accumulated to obtain the transition strength value of the adjacent product category node clusters. The product category node cluster is used as an interest node; The initial value of the interest intensity of the interest node is used as the node quality value of the interest node, the transfer intensity value between adjacent interest nodes is used as the connection intensity value between adjacent interest nodes, and the interest nodes are connected sequentially according to the magnitude of the transfer intensity value to obtain the directional gradient path of the interest node. Along the direction of the directional gradient path, the directional gradient path is oriented and encoded to obtain the intention drift vector of the user's corresponding interest change trend.

4. The content remote update system based on a mobile terminal as described in claim 1, characterized in that, When the content prefetching module extracts candidate content components corresponding to the direction of the intent drift vector from a preset content resource set based on the intent drift vector, and generates the loading condition expression for each candidate content component, it is specifically used for: The intention drift vector is parsed to extract the target product category code corresponding to the direction pointed to by the intention drift vector in the user's corresponding product category node cluster, and the target category code set of the intention drift vector is obtained. Based on the product category codes in the target category code set, retrieve content components associated with the product category codes from the preset content resource set, and aggregate and merge the content components to obtain an initial candidate component set for the content components; Based on the vector length value of the intention drift vector, threshold filtering is performed on the content components in the initial candidate component set, retaining the content components that match the vector length value and removing the content components that do not match, to obtain the candidate content component set of the initial candidate component set; Generate corresponding loading condition expressions for the candidate content components in the candidate content component set.

5. The content remote update system based on a mobile terminal as described in claim 4, characterized in that, When the content prefetching module generates corresponding loading condition expressions for candidate content components in the candidate content component set, it is specifically used for: Resource constraint evaluation is performed on the candidate content components in the candidate content component set to obtain the minimum memory requirement and minimum bandwidth requirement of the candidate content components; The memory determination sub-condition is that the available running memory value of the mobile business terminal is greater than or equal to the minimum memory requirement value, and the bandwidth determination sub-condition is that the current downlink bandwidth value of the mobile business terminal is greater than or equal to the minimum bandwidth requirement value. By performing a logical conjunction operation on the memory determination sub-condition and the bandwidth determination sub-condition, the loading condition expression of the candidate content component is obtained.

6. The content remote update system based on a mobile terminal as described in claim 1, characterized in that, The dynamic adaptation module, when performing operations such as reading the available RAM and current downlink bandwidth of the mobile business terminal, substituting these values ​​into the loading condition expression corresponding to the candidate content component for logical judgment, and marking the candidate content component with a true judgment result as a component to be loaded, is specifically used for: The mobile business terminal is sampled to obtain its available running memory value and current downlink bandwidth value. The candidate content components are traversed, and the conditions of the current candidate content component are parsed to obtain the loading condition expression of the current candidate content component. Based on the memory determination sub-condition and bandwidth determination sub-condition in the loading condition expression, the available running memory value is substituted into the memory determination sub-condition to perform memory determination, thereby obtaining the memory determination result of the current candidate content component. The current downlink bandwidth value is substituted into the bandwidth determination sub-condition to perform bandwidth determination, thereby obtaining the bandwidth determination result of the current candidate content component. When the memory determination result is true and the bandwidth determination result is true, the currently traversed candidate content component is marked as a component to be loaded; When the memory determination result is false or the bandwidth determination result is false, the currently traversed candidate content component is marked as a non-loadable component; The process continues until all candidate content components have been traversed, at which point the component to be loaded is output.

7. The content remote update system based on a mobile terminal as described in claim 1, characterized in that, When the dependency orchestration module performs dependency hierarchy orchestration on the component dependency identifiers of the components to be loaded to obtain the loading order and loading priority identifiers of the components to be loaded, it is specifically used for: The component to be loaded is associated with a search to obtain the component dependency identifier of the component to be loaded. The component dependency identifier contains a list of component identifiers of other components that the component to be loaded depends on. Using the components to be loaded as graph nodes and the dependency directions in the component dependency identifiers as directed dependency edges, a directed dependency graph of the list of components to be loaded is constructed, wherein the direction of the directed dependency edges is from the dependent component to the dependent component. The directed dependency graph is topologically sorted, and the components to be loaded corresponding to the nodes with an in-degree of zero in the topological sort are arranged into the initial loading level. The components to be loaded in the initial loading level are arranged into the next loading level, thus obtaining the loading level number of the components to be loaded. The loading time distribution of the historical loading time records of the components to be loaded is estimated to obtain the estimated loading time value of the components to be loaded. The estimated loading time value and the loading level number are weighted and arranged to obtain the loading order of the components to be loaded. Based on the loading level number, the basic priority value of the component to be loaded is determined, and the basic priority value is used as the loading priority identifier of the component to be loaded.

8. The content remote update system based on a mobile terminal as described in claim 7, characterized in that, The dependency orchestration module, when performing time distribution estimation on the historical loading time records of the components to be loaded to obtain the estimated loading time value of the components to be loaded, and then weighting the estimated loading time value with the loading level number to obtain the loading order of the components to be loaded, is specifically used for: Read the historical loading time record of the component to be loaded from the component attribute record of the component to be loaded. The historical loading time record contains the actual loading time value recorded by the component to be loaded in the historical loading operation. Arrange the actual loading time values ​​of the components to be loaded in chronological order to obtain the loading time sequence of the components to be loaded; A central tendency analysis is performed on the time series, and the average of the mode estimated time and the median estimated time in the time series is used as the initial estimated loading time of the component to be loaded. Using the loading level number of the component to be loaded as the level weight coefficient, the product of the initial estimated loading time value and the level weight coefficient is used as the weighted estimated loading time value of the component to be loaded. Based on the weighted estimated loading time value, the components to be loaded are sorted to obtain the loading order of the components to be loaded. The loading order of the components to be loaded is determined based on the loading order position and the loading level number.

9. A content remote update system based on a mobile terminal as described in claim 8, characterized in that, The formula for calculating the weighted estimated loading time is as follows: ； In the formula, The number of the component to be loaded is its sequence number in the list of components to be loaded. The first component to be loaded The loading sequence number of each component to be loaded in the history of loading operations. The first component to be loaded The weighted estimated loading time of each component to be loaded. The first component to be loaded The total number of historical loading operations experienced by each component to be loaded. The first component to be loaded The first component to be loaded in history The actual loading time value recorded in this loading operation The first component to be loaded The layer weight coefficient corresponding to the loading layer number of each component to be loaded.

10. A content remote update system based on a mobile terminal as described in claim 1, characterized in that, When the mounting and rendering module mounts the component to be loaded onto the display frame placeholder node of the mobile business terminal based on the loading order and the loading priority identifier, and obtains the final content to be displayed on the mobile business terminal, it is specifically used for: Extract the set of placeholder nodes from the display frame in the mobile business terminal; The component identifier of the component to be loaded is matched with the node identifier in the placeholder node set to obtain the target placeholder node of the component to be loaded. Based on the loading priority identifier, the components to be loaded in the target placeholder node are sorted by priority to obtain the component loading sequence of the target placeholder node; Based on the component loading sequence, the cross-node loading order of the components to be loaded in the node is arranged to obtain the global loading sequence of the components to be loaded in the node; Based on the global loading sequence, the node components to be loaded are sequentially mounted onto the corresponding target placeholder nodes, and after each node component to be loaded is mounted, a node status update signal is sent to the display frame. Once all node status update signals of the target placeholder node have been received, the rendering instruction of the display frame is triggered to merge and render the component content on the placeholder node after it has been mounted, thus obtaining the final presentation content of the mobile business terminal.

Images