Method for dynamically delivering questionnaire in third-party application through SDK (Software Development Kit)
By using a multi-dimensional factor-weighted decision-making model and SDK integration solution, the problems of multi-terminal compatibility, custom display and data security in existing technologies are solved, and the flexibility and efficiency of dynamically delivering questionnaires in third-party applications are realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU ZHONGYAN NETWORK TECH CO LTD
- Filing Date
- 2026-01-26
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies cannot achieve multi-terminal compatibility, have fixed questionnaire display formats, lack customization capabilities, lack do-not-disturb mechanisms, suffer from data leakage and low statistical efficiency, and have high access costs and long cycles.
Through a multi-dimensional factor-weighted decision-making model, we provide a multi-platform integrated solution that dynamically optimizes the questionnaire display format and appearance. Combined with an intelligent triggering mechanism and a real-time data submission strategy, we use an SDK to dynamically deliver questionnaires in third-party applications.
It achieves multi-terminal compatibility, customizable display, and non-disturbing operation, ensuring data security and real-time performance, reducing access costs, and improving statistical efficiency.
Smart Images
Figure CN122027615A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of computer data processing technology, specifically relating to a method for dynamically delivering questionnaires in third-party applications via an SDK. Background Technology
[0002] With the increasing demand for digital operations, enterprises and organizations need to collect user feedback and conduct research on needs in various third-party applications (Web / H5, WeChat Mini Programs, Android / iOS Apps).
[0003] However, existing technologies have the following shortcomings: most solutions only support a single platform (such as Web only or Android only), failing to achieve "one solution covering multiple terminals" and requiring repeated development; the display format (pop-up only) and appearance of the questionnaire are fixed, unable to be customized according to business needs, and there is no do-not-disturb mechanism; only "pop-up immediately upon entering the page" is supported, without supporting delayed triggering or externally customized triggering; there is no isolation mechanism between the questionnaire and the host application, leading to data leakage and affecting application operation; answer data needs to be uploaded periodically and cannot be submitted to the management backend in real time, resulting in low statistical efficiency; the questionnaire function needs to be embedded through hard coding, which is costly and time-consuming for third-party applications to integrate.
[0004] Therefore, a new method is urgently needed. Summary of the Invention
[0005] The purpose of this invention is to provide a method for dynamically distributing questionnaires in third-party applications via an SDK. This method improves the compatibility and flexibility of questionnaire distribution through a multi-dimensional factor weighted decision model, and is applicable to various digital operation scenarios such as user feedback collection and demand research in third-party applications.
[0006] To achieve the above objectives, this invention provides a method for dynamically delivering questionnaires in a third-party application via an SDK, comprising: The system collects six types of input data: platform environment, host application, user behavior, business configuration, network status, and data characteristics. It then calculates the optimal execution plan using a multi-factor weighted decision model. We provide customized integration solutions for different platform types of third-party applications, configure the display format and appearance of questionnaires according to dynamic optimization rules, and adjust the do-not-disturb strategy based on user behavior data. The intelligent triggering mechanism precisely controls the timing of questionnaire display, and the display effect is optimized by combining hot zone obstacle avoidance and size adaptation technologies. The system performs real-time verification of answer data, adopts a differentiated submission strategy, and achieves efficient data flow through multi-mode synchronization.
[0007] Preferably, the integration scheme includes: Web / H5 applications use JavaScriptSDK and iframe for isolated integration, Android applications use SDK.aar for integration with WebView, iOS applications use SDK.framework for integration, and WeChat mini programs use plugins and state management for integration; and Maven / CocoaPods dependency management methods are also provided.
[0008] Preferably, the dynamic optimization rules include: The formula for calculating the theme color of the questionnaire is: ; Readability is then verified using a contrast formula, expressed as follows: ; In the formula, This represents the brightness value of the foreground area. This represents the brightness value of the background area. The do-not-disturb strategy is adjusted through a dynamic do-not-disturb period, and the formula for calculating the dynamic do-not-disturb period is as follows: ; In the formula, the base period is the preset baseline interval for re-submitting the questionnaire, which is 7 days by default; the dynamic period is the actual interval for re-submitting the questionnaire; and the completion rate is the percentage of users who have participated in the questionnaire in the past. The display formats include card pop-ups, embedded displays, and floating buttons.
[0009] Preferably, the intelligent triggering mechanism includes automatic delayed triggering and externally customized triggering, and the formula for calculating the actual delay time of automatic delayed triggering is: ; In the formula, The delay time for triggering the questionnaire; the base delay is the preset baseline waiting time for triggering the questionnaire. The percentage of page DOM elements that have been loaded; the resource load rate is the percentage of page resources that have been loaded. Whether or not the questionnaire is displayed is determined by the probability of allowing its display, and the formula for calculating the probability of allowing its display is as follows: ; In the formula, the probability of being allowed to be displayed is the probability value of the questionnaire triggering the display; The time interval since the user last viewed the questionnaire; the negative behavior weight is the weight coefficient corresponding to the user's negative actions on the questionnaire.
[0010] Preferably, the optimization of the display effect specifically includes: The coordinates of the floating button are determined using the following formula: ; ; In the formula, the optimal coordinates are the final display positions of the questionnaire floating trigger button on the page; the button is the floating control that invokes the questionnaire; the hot zone is the area in the host application where users frequently operate; the button size is the actual display size of the questionnaire floating button; and the shortest side of the screen is the shorter side length of the device screen. The formula for determining the questionnaire pop-up offset is as follows: ; In the formula, the offset is the vertical displacement distance that the questionnaire pop-up needs to be adjusted; the overlap height is the height of the vertical overlap between the initial position of the questionnaire and the high-frequency operation area of the host application. Additional safety clearance is reserved; The duration of the questionnaire's interactive animation is determined using the following formula: ; In the formula, the animation duration is the actual execution duration of the questionnaire interactive animation; the platform commonly used duration is the default duration of pop-up interactive animation in the operating system / platform corresponding to the host application where the questionnaire is located. The formula for determining the card height is as follows: ; In the formula, card height is the actual height of the questionnaire when displayed in card form; base height is the fixed baseline height of the card; number of questions is the number of questions included in the questionnaire; and complexity coefficient is a coefficient set according to the question type, with the complexity coefficient set to [value missing]. , , .
[0011] Preferably, the differentiated submission strategy is as follows: When network bandwidth is less than 0.5 Mbps, fragmented submission is used. The fragment size is calculated using the following formula: ; ; In the formula, the fragment size is the maximum data size of a single fragment when the answer data is transmitted in fragments; The maximum size of the fragment is the preset limit; the bandwidth is the actual network bandwidth of the current device; the fragment data is the answer data content corresponding to a single fragment; It is a cyclic redundancy check algorithm; the check code is an encoding used to verify the integrity of fragmented data and prevent tampering. The multi-mode synchronization method includes real-time synchronization and polling backup synchronization; the polling interval calculation formula is: ; In the formula, the polling interval is the actual waiting time between two adjacent requests when the questionnaire data is synchronized through polling; the initial interval is the preset baseline interval for polling requests, which is 5 seconds by default; and the delay is the network latency of the current device.
[0012] Preferably, the real-time data verification is achieved through a parameter validity scoring formula, which is: ; In the formula, the parameter validity score is the score for the validity of the relevant parameters in the questionnaire; This represents the sum of the calculation results of all problematic parameter items; the error field is the parameter field with validity issues; the error weight is the importance coefficient of the corresponding error field.
[0013] Preferably, an exponential backoff reconnection strategy is adopted after a network outage, and the reconnection interval is calculated according to the following formula: ; In the formula, the reconnection interval is the waiting time for the next attempt to reconnect after the network connection is lost; This represents the number of reconnection attempts. It is the basic unit of time.
[0014] Therefore, the present invention employs the above-mentioned method of dynamically distributing questionnaires in third-party applications via SDK. Compared with the prior art, the technical solution of the present invention has the following beneficial effects: (1) Adopting multi-platform dedicated integration solutions (such as JavaScriptSDK for Web / H5 and SDK.framework for iOS) and dependency management methods such as Maven / CocoaPods, we overcome the problems of traditional solutions that only support a single platform, require repeated development, and have high costs and long cycles due to hard coding. In this way, we can achieve "one SDK covering multiple terminals", without hard coding, reducing access costs and shortening the access cycle. (2) Adopting multiple display forms such as card pop-up / embedded / floating button, dynamic adaptation of appearance according to host color, dynamic do-not-disturb period (associated with historical completion rate) and automatic delay (combined with page load rate) / external custom triggering technology, it overcomes the problems of fixed display and appearance, no customization capability, no do-not-disturb mechanism and single triggering method of traditional solutions, thereby meeting the business customization needs, reducing user harassment, and adapting to different business scenarios; (3) By adopting the partitioned warehouse style isolation technology and real-time verification and submission of answer data, selecting overall / fragmented submission according to network status, and WebSocket + polling backup and synchronization technology, the traditional solution has overcome the problems of no isolation mechanism between questionnaire and host application, easy data leakage and style / code pollution, and the inability to synchronize answer data in real time and low statistical efficiency. In this way, secure isolation is achieved, application operation security is guaranteed, and real-time data flow is achieved, improving statistical efficiency and data security.
[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0016] Figure 1 This is a flowchart illustrating an embodiment of a method for dynamically delivering questionnaires in a third-party application using an SDK according to the present invention; Figure 2 This is an architecture diagram of an embodiment of a method for dynamically delivering questionnaires in a third-party application using an SDK, according to the present invention. Detailed Implementation
[0017] 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 are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used in the present invention should have the ordinary meaning understood by those skilled in the art.
[0018] Example 1 like Figures 1-2 As shown, this embodiment provides a method for dynamically distributing questionnaires in a third-party application via an SDK. It should be understood that the specific parameters, models, and protocols mentioned in this embodiment are merely examples to help those skilled in the art understand the present invention, and are not intended to limit the present invention.
[0019] This invention discloses a method for dynamically delivering questionnaires in third-party applications via an SDK. Aiming for multi-platform compatibility, business scenario adaptability, user-friendly experience, and efficient data flow, it inputs five types of basic data: platform environment data, host application data, user behavior data, business configuration data, and network status data. Through multi-dimensional factor weighted calculation and a decision-making model, it outputs the optimal execution parameters for each stage. The overall formula is as follows: ; In the formula, The optimal execution plan; It is a multi-dimensional decision function; For platform environmental factors; For host application factors; User behavior factors; Configure factors for business; Network state factors; For data feature factors; , , , , , The weights are dynamic; specifically: By analyzing platform environment factors and host application factors, an integrated solution for the target platform is output, specifically: pass Calculate adaptation priority, and match the decision function with a preset integrated template library, including: If the platform type is Web / H5 ( ): Outputs a "JavaScriptSDK + iframe isolation + postMessage communication" solution; This solution employs JavaScriptSDK packaging technology, using webpack to package the core modules into a UMD format file named embed_deliver.js (with a built-in generateData variable storing delivery configuration). Third-party websites include the transition.js bootstrap file via a script tag. The bootstrap file dynamically requests the latest version of embed_deliver.js based on the cache validity period calculated by the algorithm (cache validity period, where N is the network stability factor). It also uses iframe technology to dynamically create secure isolation containers and uses the postMessage method to achieve cross-domain communication, ensuring secure isolation between the questionnaire and the host page. If the platform type is WeChat Mini Program ( 3): Output a solution of "uni-app plugin + vuex state management + style isolation in separate repositories"; This solution packages the questionnaire function into an independent plugin based on the uni-app plugin mechanism. It ensures style independence through separate repository management, uses Vuex state management to replace window global variables to achieve code sharing, and the plugin completes two-way data communication with the host mini-program through component instances and event listener mechanism. The algorithm automatically generates plugin namespace prefix (based on the hash of the host application package name) to avoid style pollution. If the platform type is Android ( ): Outputs the "SDK.aar + WebView + JSBridge communication" solution; This solution packages the deployment configuration and questionnaire functionality into an SDK.aar file, which is then imported by third-party Android projects via Maven dependency management. The algorithm matches the host dependency version and outputs conflict resolution (prioritizing higher version compatibility). WebView technology is used to load the questionnaire page, and JSBridge technology is used to achieve bidirectional communication between native code and JavaScript. If the platform type is iOS ( Outputs the "SDK.framework+WebView+JSBridge communication" solution; The deployment configuration and questionnaire functionality are packaged into an SDK.framework file and integrated into third-party iOS projects via CocoaPods dependency management. WebView technology is used to load the questionnaire page, and JSBridge technology is used to achieve bidirectional communication between native code and JavaScript, ensuring cross-language interaction stability. By analyzing business configuration factors, host application factors, and user behavior factors, a personalized optimal configuration set is output, specifically: Through calculation The decision function optimizes configuration parameters based on business objectives (feedback collection rate / user experience), including: Optimize basic configuration: Output the optimal embedding form based on the trigger scenario weight (output "card pop-up" in the e-commerce order placement scenario, and "card embedding" in the education course page scenario); Supports three configuration forms: card pop-up (pop-up at the bottom of the current page), card embedding (page layout placeholder), and floating button (pop-up after clicking). The configuration information is stored on the server and obtained through the API interface when the plugin is loaded. Optimize appearance configuration: Press The host's primary color is used to calculate the theme color adaptation value. The calculation formula is as follows: ; Readability is then verified using a contrast formula, expressed as follows: ; In the formula, This represents the brightness value of the foreground area. This represents the brightness value of the background area. It supports customization of parameters such as theme color, text color, and background color. The plugin dynamically applies the configuration after obtaining it, ensuring consistency with the host's visual style. Optimize external parameter configuration: Verify the validity of input parameters using parameter validity scores. The calculation formula is as follows: ; In the formula, the parameter validity score is the score for the validity of the relevant parameters in the questionnaire; This represents the sum of the calculation results of all problematic parameter items; the error field is the parameter field with validity issues; the error weight is the importance coefficient of the corresponding error field; Parameters with a valid score of ≥80 are directly bound; otherwise, a default parameter set is output. The host application can pass in external data parameters in JSON format. After the plugin is verified, it is associated with the questionnaire function to realize the dynamic binding of questionnaire content with external data (such as automatically filling in user ID and order information). Optimize display logic configuration: based on The dynamic do-not-disturb period for calculating historical completion rate is calculated using the following formula: ; In the formula, the base period is the preset baseline interval for re-submitting the questionnaire, which is 7 days by default; the dynamic period is the actual interval for re-submitting the questionnaire; and the completion rate is the percentage of users who have participated in the questionnaire in the past. The cycle is extended when the completion rate is less than 30% and shortened when the completion rate is greater than 70%. It supports display timing configuration (display immediately upon entering the page, delayed display, custom trigger) and do-not-disturb policy configuration (display once every 1 month / 15 days / 3 days or custom cycle). The plugin records the device identifier and the last display time through a local caching mechanism to realize local verification of the do-not-disturb policy. By analyzing user behavior factors, page state factors, and trigger rule factors, a trigger decision is output, specifically: Through calculation The decision function determines the triggering conditions, including: Trigger timing determination: The algorithm parses the trigger_type field to determine the trigger type (0=floating button, 1=pop-up mode, 2=card embedding), and determines the trigger mode based on the btn_show_immediate field (a value of 2 indicates a custom trigger); during plugin initialization, the deployment configuration is obtained through the API interface, the field parsing and trigger mode determination are completed, and subsequent trigger instructions are awaited; Automatic triggering: The actual delay is calculated based on the page loading progress. The formula for calculating the actual delay is: ; In the formula, The delay time for triggering the questionnaire; the base delay is the preset baseline waiting time for triggering the questionnaire. The percentage of page DOM elements that have been loaded; the resource load rate is the percentage of page resources that have been loaded. When the loading completion rate is ≥90%, the delay is shortened; the plugin sets a timer based on the configured btn_show_delay field and the actual delay time output by the algorithm, and displays the corresponding questionnaire component after the specified time is reached; Custom trigger: The `trigger_rule` in the middle is parsed into decision conditions. After receiving the parameters passed to the host by calling the `bestCemTrigger` method, it matches the rule threshold and outputs "satisfied / not satisfied with the trigger condition". The plugin waits for the host application to call this method, validates the passed parameters, and displays the questionnaire component if the trigger condition is met, thus avoiding the risk of code injection. Do Not Disturb Verification: Calculate the probability of allowing display, using the following formula: ; In the formula, the probability of being allowed to be displayed is the probability value of the questionnaire triggering the display; The time interval since the user last viewed the questionnaire; the negative behavior weight is the weight coefficient corresponding to the user's negative actions on the questionnaire; If the probability of displaying the output is ≥0.8, triggering is allowed; otherwise, triggering is disabled. The plugin records the device identifier and the last display time through a local storage mechanism. Combined with the allowed display results output by the algorithm, the component is hidden if the display cycle has not been reached to avoid user annoyance. By analyzing host page factors, display configuration factors, and platform interaction factors, a seamless display solution is output, specifically: Through calculation The decision function optimizes the display coordinates, dimensions, and animation parameters, including: Optimize the floating button: Determine the coordinates of the floating button using the following formula: ; ; In the formula, the optimal coordinates are the final display positions of the questionnaire floating trigger button on the page; the button is the floating control that invokes the questionnaire; the hot zone is the area in the host application where users frequently operate; the button size is the actual display size of the questionnaire floating button; and the shortest side of the screen is the shorter side length of the device screen. Create a floating button component at a specified location on the page. The button style is rendered according to the appearance configuration. When the user clicks it, the questionnaire content is displayed through the uni-popup popup component. Optimize pop-up display: Output center coordinates; if it overlaps with the bottom navigation, adjust upwards by the offset. Simultaneously, output animation parameters to match platform interaction characteristics. The calculation formula is: ; ; In the formula, the offset is the vertical displacement distance that the questionnaire pop-up needs to be adjusted; the overlap height is the height of the vertical overlap between the initial position of the questionnaire and the high-frequency operation area of the host application. The additional safety margin is reserved; the animation duration is the actual execution duration of the questionnaire interactive animation; the platform's commonly used duration is the default duration of pop-up interactive animations in the operating system / platform corresponding to the host application where the questionnaire is located. Create a pop-up component based on the configured display timing (immediate / delayed). The pop-up's position and style are set according to the appearance configuration and algorithm output parameters. The questionnaire page is embedded in the pop-up. Optimize card embedding: Calculate the adaptive size using the following formula: ; In the formula, card height is the actual height of the questionnaire when displayed in card form; base height is the fixed baseline height of the card; number of questions is the number of questions included in the questionnaire; and complexity coefficient is a coefficient set according to the question type, with the complexity coefficient set to [value missing]. , , ; Create placeholder components in the page layout. The component size and position are set according to the configuration and algorithm output parameters. The questionnaire page is directly embedded in the component to achieve seamless integration of visuals and layout. By analyzing network state factors, data feature factors, and platform interaction factors, a seamless display solution is output, specifically: Through calculation The decision letter selects the synchronization method and parameters, including: Data collection: Output answer data verification rules. The plugin captures answer changes through an event listening mechanism, stores the data in the local state management module, and performs real-time verification (such as required fields and logical association of options) and logical rule processing according to algorithm rules. Data Submission: Based on the network bandwidth output submission strategy (bandwidth ≥ 1Mbps outputs "full submission", < 0.5Mbps outputs "fragmented submission"), fragmented submission splits data according to the fragment size bandwidth. The fragment size calculation formula is as follows: ; ; In the formula, the fragment size is the maximum data size of a single fragment when the answer data is transmitted in fragments; The maximum size of the fragment is the preset limit; the bandwidth is the actual network bandwidth of the current device; the fragment data is the answer data content corresponding to a single fragment; It is a cyclic redundancy check algorithm; the check code is an encoding used to verify the integrity of fragmented data and prevent tampering. Additional verification code data fragments; after the user submits the questionnaire, the plugin submits the answer data (whole / fragments) to the server through the API interface. After the server completes the data verification and storage, it returns the submission result. If the network is interrupted, only the fragments are submitted and not received. Data synchronization: The algorithm outputs the synchronization method ("WebSocket synchronization" for high real-time requirements, and "WebSocket + polling backup" for unstable networks). The WebSocket reconnection interval is calculated based on the reconnection interval (n≤5), and the reconnection interval calculation formula is as follows: ; In the formula, the reconnection interval is the waiting time for the next attempt to reconnect after the network connection is lost; This represents the number of reconnection attempts. The basic unit of time; The polling interval is dynamically adjusted based on the initial interval delay, and the calculation formula is as follows: ; In the formula, the polling interval is the actual waiting time between two adjacent requests when the questionnaire data is synchronized via polling; the initial interval is the preset baseline interval for polling requests, which is 5 seconds by default; and the delay is the network latency of the current device. After receiving the data, the server updates the database in real time and synchronizes the data to the backend management system through the synchronization method specified by the algorithm, supporting real-time viewing of questionnaire results and data statistical analysis.
[0020] Therefore, the present invention adopts the above-mentioned method of dynamically distributing questionnaires in third-party applications through SDK. This method improves the compatibility and flexibility of questionnaire distribution through a multi-dimensional factor weighted decision model, and is applicable to various digital operation scenarios such as user feedback collection and demand research in third-party applications.
[0021] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0022] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. 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 still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A method for dynamically distributing questionnaires in a third-party application via an SDK, characterized in that, include: The system collects six types of input data: platform environment, host application, user behavior, business configuration, network status, and data characteristics. It then calculates the optimal execution plan using a multi-factor weighted decision model. We provide customized integration solutions for different platform types of third-party applications, configure the display format and appearance of questionnaires according to dynamic optimization rules, and adjust the do-not-disturb strategy based on user behavior data. The intelligent triggering mechanism precisely controls the timing of questionnaire display, and the display effect is optimized by combining hot zone obstacle avoidance and size adaptation technologies. The system performs real-time verification of answer data, adopts a differentiated submission strategy, and achieves efficient data flow through multi-mode synchronization.
2. The method for dynamically distributing questionnaires in a third-party application via SDK according to claim 1, characterized in that, The integration scheme includes: Web / H5 applications use JavaScriptSDK and iframe for isolated integration, Android applications use SDK.aar for integration with WebView, iOS applications use SDK.framework for integration, and WeChat mini programs use plugins and state management for integration; and Maven / CocoaPods dependency management methods are also provided.
3. The method for dynamically delivering questionnaires in a third-party application via SDK according to claim 1, characterized in that, The dynamic optimization rules include: The formula for calculating the theme color of the questionnaire is: ； Readability is then verified using a contrast formula, expressed as follows: ； In the formula, This refers to the brightness value of the foreground area. This represents the brightness value of the background area. The do-not-disturb strategy is adjusted through a dynamic do-not-disturb period, and the formula for calculating the dynamic do-not-disturb period is as follows: ； In the formula, the base period is the preset baseline interval for re-submitting the questionnaire, which is 7 days by default; the dynamic period is the actual interval for re-submitting the questionnaire; and the completion rate is the percentage of users who have participated in the questionnaire in the past. The display formats include card pop-ups, embedded displays, and floating buttons.
4. The method for dynamically distributing questionnaires in a third-party application via SDK according to claim 1, characterized in that, The intelligent triggering mechanism includes automatic delayed triggering and externally customized triggering. The formula for calculating the actual delay time of automatic delayed triggering is as follows: ； In the formula, The delay time for triggering the questionnaire; the base delay is the preset baseline waiting time for triggering the questionnaire. This represents the percentage of the page's DOM elements that have finished loading. Resource load rate is the percentage of page resources that have been loaded. Whether or not the questionnaire is displayed is determined by the probability of allowing its display, and the formula for calculating the probability of allowing its display is as follows: ； In the formula, the probability of being allowed to be displayed is the probability value of the questionnaire triggering the display; The time interval since the user last viewed the questionnaire; the negative behavior weight is the weight coefficient corresponding to the user's negative actions on the questionnaire.
5. The method for dynamically distributing questionnaires in a third-party application via SDK according to claim 1, characterized in that, The optimization of the display effect specifically includes: The coordinates of the floating button are determined using the following formula: ； ； In the formula, the optimal coordinates are the final display positions of the questionnaire floating trigger button on the page; the button is the floating control that invokes the questionnaire; the hot zone is the area in the host application where users frequently operate; the button size is the actual display size of the questionnaire floating button; and the shortest side of the screen is the shorter side length of the device screen. The formula for determining the questionnaire pop-up offset is as follows: ； In the formula, the offset is the vertical displacement distance that the questionnaire pop-up needs to be adjusted; the overlap height is the height of the vertical overlap between the initial position of the questionnaire and the high-frequency operation area of the host application. Additional safety clearance is reserved; The duration of the questionnaire's interactive animation is determined using the following formula: ； In the formula, the animation duration is the actual execution duration of the questionnaire interactive animation; the platform commonly used duration is the default duration of pop-up interactive animation in the operating system / platform corresponding to the host application where the questionnaire is located. The card height is determined using the following formula: ； In the formula, card height is the actual height of the questionnaire when displayed in card form; base height is the fixed baseline height of the card; number of questions is the number of questions included in the questionnaire; and complexity coefficient is a coefficient set according to the question type, with the complexity coefficient set to [value missing]. , , .
6. The method for dynamically delivering questionnaires in a third-party application via SDK according to claim 1, characterized in that, The differentiated submission strategy is as follows: When network bandwidth is less than 0.5 Mbps, fragmented submission is used. The fragment size is calculated using the following formula: ； ； In the formula, the fragment size is the maximum data size of a single fragment when the answer data is transmitted in fragments; This is the preset maximum size for the partition; The bandwidth refers to the actual network bandwidth of the current device. The segmented data consists of the answer data content corresponding to a single segment. It is a cyclic redundancy check algorithm; the check code is an encoding used to verify the integrity of fragmented data and prevent tampering. The multi-mode synchronization method includes real-time synchronization and polling backup synchronization; the polling interval calculation formula is: ； In the formula, the polling interval is the actual waiting time between two adjacent requests when the questionnaire data is synchronized through polling; the initial interval is the preset baseline interval for polling requests, which is 5 seconds by default; and the delay is the network latency of the current device.
7. The method for dynamically distributing questionnaires in a third-party application via SDK according to claim 1, characterized in that, The real-time data verification is achieved through a parameter validity scoring formula, which is as follows: ； In the formula, the parameter validity score is the score for the validity of the relevant parameters in the questionnaire; This indicates that the calculation results of all problematic parameter items are summed up. Error fields are parameter fields with validity issues; error weights are the importance coefficients of the corresponding error fields.
8. The method for dynamically distributing questionnaires in a third-party application via SDK according to claim 1, characterized in that, After a network outage, an exponential backoff reconnection strategy is adopted, with the reconnection interval calculated using the following formula: ； In the formula, the reconnection interval is the waiting time for the next attempt to reconnect after the network connection is lost; This represents the number of reconnection attempts. It is the basic unit of time.
9. A computer device, characterized in that, include: A processor configured to be coupled to memory, read and execute instructions and / or program code in the memory to perform the method as described in any one of claims 1-8.
10. A computer-readable medium, characterized in that, The computer-readable medium stores computer program code that, when executed on a computer, causes the computer to perform the method as described in any one of claims 1-8.