A method and system for dynamic rearrangement of items within a main view area of a file manager

Abstract

This invention relates to a method and system for dynamically rearranging items within the main view area of ​​a file manager, belonging to the field of file explorer technology. It solves the problem that existing technologies are unsuitable for scenarios involving the mixing of files and folders in the main view. The method includes: obtaining basic attribute information, preset tags, and stable indexes of all direct sub-items in the current directory; the item can be a folder or a file; for each item: determining a tag level value based on preset tags; determining a type level value based on preset tags and type information; determining a basic score for traditional sorting based on basic attribute information; determining a comprehensive behavior score and a predicted behavior trend score based on user operation data for the item; and determining the comprehensive sorting order of all items based on the tag level value, type level value, comprehensive behavior score, predicted behavior trend score, basic score for traditional sorting, and stable index, and then sorting all items. This invention achieves dynamic mixing of files and folders.

Description

Technical Field

[0001] This invention relates to the field of file explorer technology, and in particular to a method and system for dynamically rearranging items in the main view area of ​​a file explorer. Background Technology

[0002] Most mainstream operating systems and file managers typically provide access points such as "Quick Access," "Frequently Used Locations," and "Favorites" to reduce the search cost for users within directory hierarchies. These features are generally implemented through methods like pinning to the top, manual saving, or simple frequency tracking.

[0003] For example, some file managers pin frequently used folders to easily accessible locations; some system interfaces display recently accessed items in the sidebar or quick access area. However, the above solutions generally have the following shortcomings: 1) The sorting logic is relatively simple, mostly manifested as static pinning or fixed collection, which is difficult to adapt to the actual mixed layout of files and folders in the main view.

[0004] 2) Traditional rules usually prioritize folders over files by default, which cannot meet the need for high-priority files to be displayed across ordinary folders.

[0005] 3) Limited responsiveness to real-time user behavior; interface position adjustments are not timely enough when frequently used items change.

[0006] 4) It lacks a tail constraint mechanism at the pure sorting level, making it impossible to reliably control certain items to always appear at the end of the list. Summary of the Invention

[0007] Based on the above analysis, the embodiments of the present invention aim to provide a method and system for dynamically rearranging items in the main view area of ​​a file manager, in order to solve the problems in the prior art.

[0008] On one hand, embodiments of the present invention provide a method for dynamically rearranging items within the main view area of ​​a file manager, including: Retrieve basic attribute information, preset tags, and stable indexes of all direct sub-projects in the current directory; wherein, the project is a folder or a file; For each item: determine a tag level value based on the preset tag; determine a type level value based on the preset tag and the type information included in the basic attribute information; wherein, the type information is folder type or file type; determine a traditional sorting base score based on the basic attribute information; obtain user operation behavior data for the item, and determine a comprehensive behavior score and a behavior trend prediction score based on the operation behavior data; The overall ranking order of all items is determined based on the tag level value, type level value, behavior comprehensive score, behavior trend prediction score, traditional ranking base score, and stable index of all items. All items are comprehensively ranked based on their overall ranking order.

[0009] Furthermore, the preset markers include top markers, ordinary markers, and last anchor markers.

[0010] Further, determining the tag level value based on the preset tag includes: When the project's preset marker is the top marker, the marker level value is the first marker level value; When the preset marker of the project is a normal marker, the marker level value is the second marker level value; When the preset marker of the project is the last anchor marker, the marker level value is the third marker level value; The first marker level value is greater than the second marker level value, which is greater than the third marker level value.

[0011] Further, determining the type level value based on the preset marker and the type information included in the basic attribute information includes: When the preset marker is the top marker, the type level value of the items whose type information is folder type or file type is the first type level value; When the preset marker is a normal marker, and the type information is a folder type, the project's type level value is the first type level value; when the type information is a file type, the project's type level value is the second type level value. When the preset marker is the last anchor marker, the type level value of the item whose type information is folder type or file type is the third type level value; The first type level value is greater than the second type level value, which is greater than the third type level value.

[0012] Furthermore, The user's operational behavior data for the project includes access frequency, recent access time, dwell time, operation type, and number of times the app has been opened and closed within a preset time. The comprehensive score for behavior is determined based on the aforementioned operational behavior data, including: The access frequency is normalized to obtain the normalized access frequency; Determine the exponential decay function value of the most recent access time and the current time interval, and normalize the exponential decay function value to obtain a normalized most recent access score. The cumulative stay duration and effective stay duration are determined based on the stay duration. The cumulative stay duration is then normalized to obtain the normalized cumulative stay duration. The effective stay score is determined by the percentage of effective stay time based on the effective stay duration and the cumulative stay duration. The operation depth score is determined based on the operation type. The demotion score for fast shutdown is determined based on the number of times the device is opened and closed within a preset time period. The comprehensive behavioral score is determined based on the normalized access frequency, the normalized recent access score, the normalized cumulative dwell time, the effective dwell time score, the operation depth score, and the quick close demotion score.

[0013] Further, determining the behavioral trend prediction score based on the operational behavior data includes: Construct a behavior phase space based on the aforementioned operational behavior data; Calculate the phase velocity and phase acceleration of the behavior phase space; The behavioral trend prediction score is determined based on the phase velocity and the phase acceleration.

[0014] Furthermore, a behavior phase space is constructed based on the aforementioned operational behavior data, including: Construct a three-dimensional behavior phase space, which includes the three-dimensional behavior state vector X_i(t) = [Frequency_i(t), Recency_i(t), Dwell_i(t)] of item i at time t, where Frequency_i(t) is the normalized access frequency of item i at time t, Recency_i(t) is the normalized recent access score of item i at time t, and Dwell_i(t) is the normalized cumulative dwell time of item i at time t; The phase velocity of the behavioral phase space is calculated using the following formula: V_i(t) = X_i(t) X_i(t Δt); Where V_i(t) is the phase velocity of item i at time t, and Δt is the preset time step; X_i(t) Δt) represents the value of item i at time t. The three-dimensional behavioral state vector of Δt; The phase acceleration of the behavior phase space is calculated using the following formula: A_i(t) = V_i(t) V_i(t Δt); Where A_i(t) is the summation velocity of item i at time t; V_i(t) Δt) represents the value of item i at time t. Phase velocity Δt; The behavioral trend prediction score is determined based on the phase velocity and the phase acceleration according to the following formula: TrendScore_i = α × |V_i(t)| × cos(θ_i) + β × |A_i(t)| × cos(φ_i); Where TrendScore_i is the predicted score of the behavioral trend of item i, θ_i is the angle between the phase velocity vector of item i and the direction of the increase in activity, φ_i is the angle between the phase acceleration vector of item i and the direction of the increase in activity, and α and β are weighting coefficients.

[0015] Furthermore, the stable index is the natural sequence number of all projects in the current directory when each project first enters the current directory.

[0016] Furthermore, the project's basic attribute information, preset tags, traditional sorting base score, comprehensive behavior score, behavior trend prediction score, and stable index are stored in any of the following forms: Extended metadata storage methods; Local configuration file storage method; Local database storage method.

[0017] On the other hand, embodiments of the present invention provide a dynamic reordering system for items within the main view area of ​​a file manager, including: The information acquisition module is used to acquire basic attribute information, preset tags and stable indexes of all direct sub-projects in the current directory, as well as user operation behavior data of the projects; wherein, the projects are folders or files; The numerical calculation module is used for each item to: determine a tag level value based on the preset tag; determine a type level value based on the preset tag and the type information included in the basic attribute information; wherein the type information is folder type or file type; determine a traditional sorting base score based on the basic attribute information; acquire user operation behavior data for the item, and determine a comprehensive behavior score and a behavior trend prediction score based on the operation behavior data; The comprehensive sorting order determination module is used to determine the comprehensive sorting order of all items based on the tag level value, the type level value, the behavior comprehensive score, the behavior trend prediction score, the traditional sorting base score, and the stable index of all items; The sorting module is used to perform a comprehensive sorting of all items based on the overall sorting order of all items.

[0018] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects: 1. Improve the accessibility of frequently used items and reduce the time users spend searching in the directory.

[0019] 2. Set a top-priority tag for files to break the traditional sorting method of "folders always come before files". Through the hierarchical display mechanism, high-priority items are placed in a position that is more in line with the user's actual usage habits.

[0020] 3. Setting a last-position anchor mark introduces a last-position constraint from a sorting perspective, ensuring that a specific item remains stable at the end of the list even if it is frequently accessed.

[0021] 4. By using dynamic hybrid sorting, it takes into account both traditional browsing habits and dynamic usage habits. Users can observe that frequently used files are sorted to the top of the list during use, making it easier for users to understand and use.

[0022] 5. By using time decay and stable indexes, the stability of sorting results and interface consistency are improved, avoiding interface jitter caused by frequent refreshes.

[0023] 6. By introducing an operation depth score, the system can distinguish between shallow browsing and deep usage by users, giving files that have been deeply edited a higher sorting priority than files that have only been viewed, and more accurately reflecting the user's true usage intent.

[0024] 7. By correcting the covariate of the effective dwell time score, noise pollution of the dwell time during idle periods is eliminated, the purity of the user's real usage signal in the behavioral data is improved, and the ranking results are not affected by scenarios such as opening and leaving the computer.

[0025] 8. By quickly disabling the weighted score, statistical bias caused by accidental opening is suppressed, avoiding the incorrect placement of files with similar names due to frequent accidental opening, so that the sorting results more accurately reflect the user's true access intention rather than misoperation.

[0026] 9. Through behavioral trend prediction algorithms, the system can identify projects with rapidly increasing activity and proactively prioritize them, achieving a leap from passive sorting to proactive prediction, thereby improving the efficiency and experience of users in finding target projects.

[0027] 10. By using a stable index debouncing mechanism, the sorting results are prevented from repeatedly jumping between adjacent positions due to slight fluctuations in behavior scores, thus improving interface consistency.

[0028] 11. By extending metadata storage, the sorting state is persistently bound to the file object, and can be accurately restored after renaming, moving or refreshing, thus improving the persistence and recoverability of the sorting state.

[0029] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained through the details specifically pointed out in the description and drawings. Attached Figure Description

[0030] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0031] Figure 1 A flowchart illustrating a method for dynamically rearranging items within the main view area of ​​a file manager, provided in an embodiment of the present invention; Figure 2 This diagram illustrates a comparison of the implementation effects of sorting items within the main view area of ​​a file manager using a traditional sorting method, a partially implemented dynamic sorting method, and a fully implemented method of the present invention, as provided in this embodiment of the invention. Detailed Implementation

[0032] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0033] A specific embodiment of the present invention discloses a method for dynamically rearranging items within the main view area of ​​a file manager. The present invention does not constrain sidebars, navigation bars, or other auxiliary areas, and the protection focus is limited to the arrangement of items within the main view area of ​​the file manager.

[0034] like Figure 1 As shown, the method includes: S110: Obtain the basic attribute information, preset tags, and stable index of all direct sub-projects in the current directory; wherein, the project is a folder or a file.

[0035] In practice, all direct sub-items under the current directory refer to the first-level files and folders that the system directly obtains by reading the directory entries of the folder currently opened by the user, excluding any deep sub-items within subfolders.

[0036] Basic attribute information can include project type, name, modification time, size, and other fundamental attributes. Project type refers to whether the project is a file or a folder, which is used to determine the TypeRank value later. Attributes such as name, modification time, and size are used to calculate the traditional baseRuleScore.

[0037] Alternatively, you can first obtain the parent-child hierarchy information of all items in the current directory, and then determine all direct child items in the current directory based on this information. Parent-child hierarchy refers to whether an item is a direct child of the current directory, used to determine the item's hierarchical position in the directory structure. The specific application of parent-child hierarchy in sorting is: (1) The system only performs the hybrid sorting of the present invention on the first-level sub-items of the current directory.

[0038] (2) If items that are not first-level sub-items are displayed in the main view through search results, virtual folders or aggregate views, these items will not participate in the mixed sorting of this invention and will still be arranged according to the original default rules of the file manager.

[0039] (3) Through this constraint, the sorting mechanism of the present invention is clearly limited to subfolders and files within the current directory level, avoiding interference from cross-level items with the sorting results.

[0040] Among them, the preset tags are the tags for each item (file or folder), including the top tag, the regular tag, and the last anchor tag.

[0041] The stable index (StableIndex) is assigned a unique StableIndex value to each item when it first enters the current directory. This value is equal to the item's natural sequence number among all items in the current directory (e.g., the first item's StableIndex = 1, the second item's StableIndex = 2, and so on). Once assigned, the StableIndex is persistently stored and will not change due to user behavior unless the item is deleted or the user manually resets it.

[0042] S120: For each item: determine the tag level value based on the preset tag; determine the type level value based on the preset tag and the type information included in the basic attribute information; wherein, the type information is folder type or file type; determine the traditional sorting base score based on the basic attribute information; obtain user operation behavior data for the item, and determine the comprehensive behavior score and behavior trend prediction score based on the operation behavior data.

[0043] In some embodiments, determining a tag level value based on the preset tag includes: When the project's preset marker is the top marker, the marker level value is the first marker level value; When the preset marker of the project is a normal marker, the marker level value is the second marker level value; When the preset marker of the project is the last anchor marker, the marker level value is the third marker level value; The first marker level value is greater than the second marker level value, which is greater than the third marker level value.

[0044] In specific implementation, the weights of different tags are set as follows: the last anchor tag has the lowest weight. The goal of the last anchor tag is not to hide or control access, but to stably place the item at the end of the list through sorting constraints. A fixed last tag level value is set for this type of item so that it is always less than items with ordinary tags when compared; the top tag has the highest weight to ensure its priority display, and items with ordinary tags have a moderate weight. Based on this, the assignment rule for the tag level value TagRank in this invention is as follows: 1) Items marked as pinned to the top have the highest TagRank (i.e., the first tag level is 3); 2) Items with ordinary tags have a TagRank in the middle (i.e., the second tag level value is 2); 3) Items with last anchor tags have the lowest TagRank (i.e., the third tag level value is 1).

[0045] In some embodiments, determining a type level value based on the preset marker and the type information included in the basic attribute information includes: When the preset marker is the top marker, the type level value of the items whose type information is folder type or file type is the first type level value; When the preset marker is a normal marker, and the type information is a folder type, the project's type level value is the first type level value; when the type information is a file type, the project's type level value is the second type level value. When the preset marker is the last anchor marker, the type level value of the item whose type information is folder type or file type is the third type level value; The first type level value is greater than the second type level value, which is greater than the third type level value.

[0046] In practical implementation, for different types: within the same tag layer, either the traditional rule of "folder priority" or "file priority" can still be retained as the default order. For example, for items in a normal layer, folders can be placed before files by default. However, when an item has a pinned tag, a conflict occurs between pinned files and unpinned folders, allowing files with pinned tags to cross over folders without pinned tags. The traditional folder priority rule for this layer is replaced by a skip-level rule, breaking the traditional sorting method of "folders always come before files." In other words, folders no longer naturally have absolute priority; instead, the traditional folder priority rule is only used when the tags are the same and the sorting priorities are similar. Based on this, the assignment rule for the TypeRank value in this invention is as follows: 1) Files and folders marked as pinned to the top, as well as folders marked as ordinary (not pinned), all received a higher TypeRank (i.e., a first type rank value of 2). 2) Files with ordinary tags (not pinned) receive a lower TypeRank (i.e., a second type rank value of 1); 3) Files and folders with last-anchor tags have the lowest TypeRank (i.e., a third-type level value of 0). Because items with last-anchor tags all have the same TagRank and a TypeRank of 0, TypeRank no longer plays a differentiating role within the last-anchor layer. The order of two last-anchored items is determined by the StableIndex (explained later). Combined with the TagRank value, files marked as pinned to the top can be displayed across folders that are not pinned to the top.

[0047] In practice, the traditional sorting base score is determined based on the aforementioned basic attribute information; this base score is derived from traditional sorting fields such as name, modification time, and size. The specific calculation method is as follows: traditional fields are comprehensively scored according to the current main view's sorting rules (e.g., ascending by name, descending by modification time, descending by size, etc.). For example, when the sorting rule is "descending by modification time," items with more recent modification times receive a higher traditional sorting base score (BaseRuleScore); when the sorting rule is "ascending by name," items that appear earlier in alphabetical order receive a higher traditional sorting base score (BaseRuleScore). If multiple traditional fields are considered simultaneously, a weighted sum can be calculated based on the weight of each field.

[0048] In some embodiments, the user's operational behavior data for the project includes access frequency, recent access time, dwell time, operation type, and number of times the device is opened and closed within a preset time. The comprehensive score for behavior is determined based on the aforementioned operational behavior data, including: The access frequency is normalized to obtain the normalized access frequency; Determine the exponential decay function value of the most recent access time and the current time interval, and normalize the exponential decay function value to obtain a normalized most recent access score. The cumulative stay duration and effective stay duration are determined based on the stay duration. The cumulative stay duration is then normalized to obtain the normalized cumulative stay duration. The effective stay score is determined by the percentage of effective stay time based on the effective stay duration and the cumulative stay duration. The operation depth score is determined based on the operation type. The demotion score for fast shutdown is determined based on the number of times the device is opened and closed within a preset time period. The comprehensive behavioral score is determined based on the normalized access frequency, the normalized recent access score, the normalized cumulative dwell time, the effective dwell time score, the operation depth score, and the quick close demotion score.

[0049] In practice, for items with ordinary tags that are not forcibly pinned to the top or bottom, they are dynamically sorted based on user behavior data. The BehaviorScore comprehensively characterizes user behavior across six dimensions: frequency, timeliness, duration, operation depth, effectiveness, and exclusion of erroneous operations. These dimensions collectively constitute a comprehensive measure of the user's true usage intent. Details of the six dimensions are as follows: 1) Access Frequency: The number of times the project is opened within a preset time window. The access frequency needs to be normalized (mapping the original value to the 0~1 range) to obtain a normalized access frequency. For example, max-min linear normalization can be used: Normalized access frequency = (count min) / (max min) .in, count For access frequency, min To minimize the access frequency, max This represents the maximum access frequency. min and max should be set based on the actual situation.

[0050] 2) Recent Visit Score: The exponentially decaying score calculated from the user's last visit to this item. The exponential decay function... e λ（t-tf） t represents the current time, and tf represents the time of the most recent access. λ is the attenuation coefficient (λ>0, control) (Speed ​​of decay) Then, the exponential decay function value is normalized to obtain the normalized most recently visited score.

[0051] 3) Dwell Time Score: The cumulative dwell time of users on this item within a preset time window. The cumulative dwell time needs to be normalized to obtain a normalized cumulative dwell time. Normalization can be done using a reference upper limit: Normalized Cumulative Dwell Time = Cumulative Dwell Time / maxt. maxt is a reference upper limit for the cumulative dwell time, set based on actual conditions.

[0052] 4) Effective Stay Score: The ratio of effective stay time to cumulative stay time within a preset time window. This score distinguishes between effective stays and ineffective idle time, correcting measurement noise in the cumulative stay time for ineffective periods such as "opening and leaving". Effective stay time is calculated by removing periods of continuous ineffective idle time from the cumulative stay time. Periods of continuous ineffective idle time refer to cumulative stay times that continuously exceed a threshold (e.g., 60 seconds, which marks the boundary between pauses and actual departures, eliminating most minute-level departures while retaining pauses of several seconds to tens of seconds; this threshold is adjustable).

[0053] 5) Operation Depth Score: Operational behaviors are categorized into shallow browsing (e.g., opened 10 times but only spent 3 seconds each time) and deep usage (e.g., opened 3 times but edited for 20 minutes each time). Relying solely on access frequency and dwell time is insufficient for accurate differentiation. Therefore, this invention upgrades the evaluation from "statistically what users did" to "distinguishing how well users did it" by using operation type in the operation depth score, providing the strongest signal reflecting true user intent. Operation type directly reflects the user's level of cognitive engagement. Specifically, different weights are assigned based on operation type (open / view only: 0.2; preview / quick browse: 0.3; rename / move / copy: 0.5; edit / modify content: 0.7; save / export / share: 0.9). The scores are ranked in ascending order according to the level of cognitive input: read-only operations (view / preview) have the lowest scores, structured organization operations (rename / move / copy) are in the middle, and output operations (save / export / share) have the highest scores. The highest score retains a margin of 0.1 to prevent normalization saturation and is an adjustable example value. The operation depth score is calculated as the weighted average of the scores of each operation type within a preset time window.

[0054] 6) Quick Close Demotion Score: By statistically analyzing the ratio of the number of times an item is closed within a threshold time (e.g., 5 seconds, shorter than the minimum reading time required for a typical user to complete one file recognition; closing within 5 seconds indicates that the user has confirmed that the file is "not the target file" and has terminated its use; the quick close event is used to identify accidental openings and is adjustable) to the total number of times an item is opened, a negative correction is made for false high-frequency access caused by accidental openings.

[0055] First, the raw values ​​for each dimension are statistically analyzed and normalized. Then, the overall behavioral score is calculated using a weighted formula. BehaviorScore = w1 × Normalized access frequency + w2 × Normalized recent access score + w3 × Normalized cumulative dwell time + w4 × Operation depth score + w5 × Effective dwell time score + w6 × Quick closure demotion score; Among them, w1 to w6 are preset weight coefficients, w1 to w5 are positive numbers, and the sum of w1 to w5 and the absolute value of w6 satisfy the weighted balance relationship, and w6 is a negative value.

[0056] Items with higher BehaviorScores rank higher within the same tag and type layers.

[0057] Dynamic refresh mechanism for the main view: When user behavior data changes, the system automatically recalculates the BehaviorScore of the affected items and re-sorts the items in the main view accordingly. The updated sorting order is then refreshed in real time to the file manager's main view interface, allowing frequently used items to automatically move to the front of the page as user habits change, without requiring manual intervention from the user.

[0058] In some embodiments, determining a behavioral trend prediction score based on the operational behavior data includes: Construct a behavior phase space based on the aforementioned operational behavior data; Calculate the phase velocity and phase acceleration of the behavior phase space; The behavioral trend prediction score is determined based on the phase velocity and the phase acceleration.

[0059] In practical implementation, to further enhance the intelligence level of the ranking, a behavior trend prediction algorithm is introduced on top of the comprehensive behavior score. This algorithm predicts the usage trend of items by tracking the changes in user behavior data over time, and prioritizes items with "increasing activity". The core idea of ​​the algorithm is to model the behavior changes of each item over time as a trajectory in phase space, and predict the future trend of item behavior by calculating the phase velocity and phase acceleration of the trajectory.

[0060] Among them, behavioral data is sampled at a time step of Δt=1 day to form a time-series snapshot sequence, and trend assessment is based on a fixed-length sliding time window: the time window length W is preset to 7 (i.e., the 7 most recent time points, corresponding to 7 days), and the window slides forward by one step each time a new snapshot arrives; phase velocity and phase acceleration are calculated on the state snapshots of adjacent time points within the time window; when the actual number of time points n within the time window is less than 7 (cold start scenario), the trend prediction score is weighted by a weighting factor ω=min(1,n / 7).

[0061] The reason for choosing a time window length of 7 is twofold: First, it aligns with the full confidence window for cold starts (ω=1 when n≥7); second, 7 days precisely cover a complete "week" usage rhythm. Periodically fluctuating items will simultaneously include periods of rising and falling activity within a time window, with alternating positive and negative phase velocity directions and discontinuous direction, resulting in a trend prediction score that tends towards zero, thus avoiding misjudging periodic usage as an upward or downward trend.

[0062] The specific implementation method is as follows: 1. Construct a behavior phase space based on operational behavior data, including: constructing a three-dimensional behavior phase space. The three-dimensional behavior phase space includes the three-dimensional behavior state vector X_i(t) = [Frequency_i(t), Recency_i(t), Dwell_i(t)] of item i at time t, where Frequency_i(t) is the normalized access frequency of item i at time t, Recency_i(t) is the normalized recent access score of item i at time t, and Dwell_i(t) is the normalized cumulative dwell time of item i at time t; the values ​​of all are in the range of [0,1].

[0063] 2. Calculate the phase velocity. The phase velocity describes the rate of change of the project's behavioral state. It can be calculated using the following formula: V_i(t) = X_i(t) X_i(t Δt); where V_i(t) is the phase velocity of item i at time t, and Δt is the preset time step; X_i(t) Δt) represents the value of item i at time t. The three-dimensional behavior state vector of Δt; the larger the magnitude of the phase velocity |V_i(t)|, the more drastic the behavior change of the project.

[0064] 3. Calculate the phase acceleration. The phase acceleration describes the accelerating trend of changes in the project's behavior. It can be calculated in the phase space of the behavior using the following formula: A_i(t) = V_i(t) V_i(t Δt); where A_i(t) is the summation velocity of item i at time t; V_i(t) Δt) represents the value of item i at time t. Phase velocity Δt.

[0065] 4. Calculate the trend prediction score. The behavioral trend prediction score can be determined based on the phase velocity and phase acceleration using the following formula: TrendScore_i = α × |V_i(t)| × cos(θ_i) + β × |A_i(t)| × cos(φ_i); where TrendScore_i is the behavioral trend prediction score for item i, θ_i is the angle between the phase velocity vector of item i and the direction of increasing activity, φ_i is the angle between the phase acceleration vector of item i and the direction of increasing activity, and α and β are weighting coefficients. Example: α=1, β=0.5; since all components of the state vector are normalized to [0,1], the maximum single-step magnitude of the phase velocity |V| ≤ The maximum modulus of the additive velocity is |A|≤2 The value is approximately twice the upper limit of |V|. Taking β = α / 2 = 0.5 ensures that the contribution dimensions of the velocity and acceleration terms to the trend score are consistent and comparable in scale, preventing acceleration from dominating the trend score due to its larger numerical range. α and β are both adjustable parameters. A positive score is given when the direction of behavioral change is towards an increase in activity, and a negative score is given when it is towards a decrease in activity. The direction of increasing activity refers to the direction in which each component in the three-dimensional behavioral state space increases (i.e., in the same direction as the state vector [1,1,1]). A positive score is obtained when the phase velocity or phase acceleration has a positive projection in this direction.

[0066] This algorithm tracks the trajectory of behavioral data over time, which is a dynamic trajectory analysis. Compared with static cross-sectional analysis that calculates scores based on the current snapshot of behavioral data, it is better able to capture the trend differences between increasing and decreasing activity.

[0067] S130: Determine the overall ranking order of all items based on the tag level value, the type level value, the behavior comprehensive score, the behavior trend prediction score, the traditional ranking base score, and the stable index.

[0068] In practice, the final sorting of each file or folder item is a multi-field tuple, for example: FinalKey = (TagRank, TypeRank, BehaviorScore, BaseRuleScore, StableIndex) When sorting, compare the sort keys field by field: First, compare the TagRank of different items; items with higher TagRank are ranked higher. If the TagRank is the same, then the TypeRank is compared, and the item with the higher TypeRank is ranked first. If the TypeRank is also the same, then the BehaviorScore is compared. If the difference between the BehaviorScores of the two items exceeds the threshold δ (a preset adjustable parameter, such as 0.15, which can be marked according to the low percentile of the distribution of the absolute value of the difference between the scores of each pair of items, for example, taking the 10th to 15th percentile, so that about 10% to 15% of the items with the closest scores enter the trend-based ranking), the order is determined by the BehaviorScore. If the absolute value of the difference between the BehaviorScores of the two items does not exceed the preset threshold δ, then the TrendScore is used as the behavior-based ranking factor for comparison, with the item with the higher TrendScore ranking first. Items with higher TrendScores (items with rapidly increasing activity) receive an additional boost within the threshold range. Notably, TrendScore is not directly used as a sorting field, but is used as the behavior-based ranking factor for comparison when the difference between the BehaviorScores of the two items does not exceed the preset threshold δ. The example value δ=0.15 has the following advantages: the comprehensive behavioral score is a weighted sum of the six-dimensional normalized scores, and its value range is approximately [ [0.14, 0.86], with a range of approximately 1.0, 0.15 accounts for approximately 15% of the full range, providing moderate coverage; 0.15 is slightly higher than the minimum positive dimension contribution, ensuring that the trend prediction score only intervenes in the final decision when the behavioral score difference has not yet widened by one dimension and behavioral evidence is insufficient, thus avoiding frequent rewriting of the behavioral layer ranking.

[0069] If the BehaviorScore or TrendScore are the same, then the BaseRuleScore is compared, and the item with the higher BaseRuleScore is ranked first. If all the first few items are the same, then the StableIndex is compared. Items with smaller StableIndex values ​​are ranked first. StableIndex serves as the final criterion for breaking ties, ensuring the stability of the sorting results.

[0070] The following example illustrates the calculation process of the sort key: Example 1: Suppose there are three items in the main view: Project A (folder, ordinary tag, accessed 10 times in the last 30 days, last accessed 1 day ago, duration of stay 60 minutes). Project B (file, pinned, accessed 5 times in the last 30 days, last accessed 3 days ago, with a stay of 20 minutes). Project C (folder, ordinary tag, accessed 3 times in the last 30 days, last accessed 10 days ago, with a stay of 10 minutes).

[0071] First, calculate the sort key for each item: Project A has a TagRank of 2 (normal), a TypeRank of 2 (normal folder), and a BehaviorScore of 0.85 (hypothetical value). Project B has TagRank=3 (top), TypeRank=2 (top file), and BehaviorScore=0.55; Project C has TagRank=2 (normal), TypeRank=2 (normal folder), and BehaviorScore=0.30.

[0072] During sorting, item B has the highest TagRank (3), so it ranks first. Items A and C have the same TagRank (both are 2). Further comparing the BehaviorScore, item A (0.85) is higher than item C (0.30), 0.85-0.30=0.55>δ (0.15). The order is determined by the BehaviorScore, so item A ranks second and item C ranks third. The final sort is: item B (pinned file) > item A (normal folder) > item C (normal folder). Although item B is a file, because it has a pinned tag, its TagRank is higher than that of a normal folder, so it can be displayed at the top, bypassing normal folders.

[0073] Example 2: Suppose there are two items in the main view: Item D (file, pinned, accessed 20 times in the last 30 days, last accessed 1 day ago, with a stay of 90 minutes); BehaviorScore calculated to be 0.92; Project E (folder, pinned, accessed 5 times in the last 30 days, last accessed 5 days ago, with a stay of 30 minutes); BehaviorScore calculated to be 0.45.

[0074] During sorting, both items were marked as top-ranked and had the same TagRank (3). Further comparing TypeRank, both items were also top-ranked and had the same TypeRank (2). Next, comparing BehaviorScore, item D had a calculated BehaviorScore of 0.92, while item E had 0.45. 0.92 - 0.45 = 0.47 > δ (0.15). Therefore, the order is determined by the BehaviorScore, with item D higher than item E. Thus, the final ranking is: Item D > Item E. In other words, the order of the two top-ranked items is determined by their overall behavioral score.

[0075] Example 3: Suppose there are three folders with ordinary tags in the main view: Project A: Current BehaviorScore = 0.70, 7 days ago it was 0.40, and the phase velocity direction is increasing activity; Project B: Current BehaviorScore = 0.85, 7 days ago it was 0.88, and the phase velocity direction is decreasing activity; Project C: BehaviorScore = 0.50, which remains unchanged before and after 7 days.

[0076] Based on the behavioral trend prediction algorithm, project A has a TrendScore of 0.85, and project B has a TrendScore of... Project C's TrendScore is 0.10, and its BehaviorScore is 0.00. Projects A and B have BehaviorScores of 0.70 and 0.85 respectively, with a difference of 0.15 that does not exceed the preset threshold δ. Therefore, they proceed to the TrendScore final round. Project A wins with a TrendScore of 0.85, significantly higher than Project B's. 0.10) Offside early, even though its current BehaviorScore is lower than item B.

[0077] The final ranking is: Item A > Item B > Item C. The behavioral trend prediction algorithm identified that Item A was being used frequently and proactively moved it to the top, transforming the ranking from a passive response to user behavior to a proactive prediction of user needs.

[0078] Example 4: Suppose there are three folders with ordinary tags in the main view: Project F (BehaviorScore=0.60, BaseRuleScore=0.70, StableIndex=1); Project G (BehaviorScore=0.60, BaseRuleScore=0.70, StableIndex=2); Item H (BehaviorScore=0.60, BaseRuleScore=0.70, StableIndex=3).

[0079] During sorting, the TagRank, TypeRank, BehaviorScore, TrendScore, and BaseRuleScore of the three items are all identical. At this point, the system compares the StableIndex, and the item with the smaller StableIndex value is ranked first. The final ranking is: Item F (StableIndex=1) > Item G (StableIndex=2) > Item H (StableIndex=3).

[0080] Now suppose a user opens item G, causing an increase in the frequency of access to item G, and the BehaviorScore rises from 0.60 to 0.85. At this point, the BehaviorScore of item G (0.85) is higher than that of items F (0.60) and H (0.60), and 0.85 - 0.60 = 0.25 > δ (0.15). Therefore, item G is moved to the first position in the list after reordering. The sorting result becomes: item G (0.85) > item F (0.60) > item H (0.60).

[0081] Assuming time progresses, the BehaviorScore of item G drops from 0.85 to 0.60 due to time decay, becoming the same as item F again. At this point, the system compares the StableIndex. Item F's StableIndex (1) is less than item G's StableIndex (2), so item F returns to the top of item G. The sorting result becomes: Item F (0.60) > Item G (0.60) > Item H (0.60), consistent with the initial state. Here, time decay refers to the fact that the most recent access time index decreases as time increases from the current point in time; that is, the longer an item has not been accessed, the lower its most recent access score, thus gradually reducing its priority in the sorting.

[0082] Through the above mechanism, StableIndex guarantees that: (1) when all sorting fields are the same, the relative position of the item is uniquely determined by StableIndex and will not change due to the randomness of the sorting algorithm or small floating-point errors; (2) when the BehaviorScore recovers after a short fluctuation, the sorting position of the item can accurately return to its original state, avoiding the "jittering" phenomenon of the sorting result repeatedly jumping between two positions.

[0083] In practice, the last-anchor tag and the cross-type skip-level display rule work together during the sorting process. Items with a last-anchor tag have the lowest priority, and their TagRank is always the minimum. Therefore, no matter how high the BehaviorScore or BaseRuleScore of an item with a last-anchor tag is, it will always be at the end of the list during sorting. The cross-type skip-level display rule only affects the comparison of TypeRank within the same tag level (i.e., when TagRank is the same), and does not affect the hierarchical division of TagRank. Therefore, even if a file is set with a last-anchor tag and has a high BehaviorScore, its TagRank is still lower than all ordinary items and the top item, and it will eventually remain stable at the end of the list.

[0084] S140: Perform a comprehensive sorting of all items based on the comprehensive sorting order of all items.

[0085] In practice, the final sorting result is written back to the main view interface and dynamically refreshed when subsequent behaviors change.

[0086] Ultimately, this invention enables a hybrid view in the main view area that prioritizes dynamic priority objects while sorting ordinary objects according to traditional rules: a small number of high-value items are prominently prioritized, while the remaining items maintain a familiar, conventional arrangement.

[0087] In some embodiments, the project's basic attribute information, preset tags, traditional sorting base score, comprehensive behavior score, behavior trend prediction score, and stable index are stored in any of the following forms: Extended metadata storage methods; Local configuration file storage method; Local database storage method.

[0088] In practice, to ensure the continuous and stable operation of dynamic reordering, top-position marking, bottom-position anchoring, and behavior-driven sorting within the main view area, the system adopts an extended metadata storage method to persistently manage project-related information.

[0089] Extended metadata storage refers to using the extended attributes mechanism provided by the operating system's file system to directly write sorting-related additional information as key-value pairs into the metadata of file or folder objects. For example, in file systems that support extended attributes, several custom attribute keys can be set for each file or folder (such as "com.app.pinned" for pinned, "com.app.tailanchor" for tailanchor, "com.app.frequency" for access frequency, etc.), and the corresponding values ​​can be persistently stored. The characteristics of this storage method are: data is directly bound to the file or folder object and moves or copies with the object; it does not rely on external databases or configuration files; and it can be read and written at the file system level without requiring additional storage services.

[0090] Specifically, the system writes the project's basic attributes, preset tags, behavioral statistics, and sorting auxiliary fields into the extended metadata supported by the file system, and establishes a binding relationship between this information and the corresponding file or folder objects. The sorting auxiliary fields include a stable index (StableIndex) and behavioral scores (BehaviorScore), TrendScore, and traditional sorting base score (BaseRuleScore) used for dynamic reordering. When a project is renamed, because the extended metadata is bound to the file object rather than depending on the name, the preset tags will not be lost due to the name change (the system assigns a unique underlying identifier to each file / folder. The extended metadata is directly associated with this identifier, not the file path or name). When the directory is refreshed or the sorting is updated, the system rereads the preset tags and sorting auxiliary information of each project from the extended metadata, recalculates the sorting priority, and refreshes the view, rather than reinitializing the preset tags of each project. When re-entering the main view, the system also restores the complete sorting information of each project from the extended metadata, thus ensuring the persistence of preset tags and sorting association information. It should be noted that each project can only be under one of the preset tags (top tag, normal tag, or bottom anchor tag) at any given time, and there is no situation where it can have two tags at the same time.

[0091] Through the above methods, the system can ensure that the sorting status can be accurately restored after operations such as file renaming, directory structure adjustment or interface refresh, while improving the persistence, consistency and recoverability of the sorting status, and ensuring that the hybrid sorting mechanism described in this invention is stable and effective in the main view area of ​​the file manager for a long time.

[0092] Data storage can also be replaced by the following methods: 1. Local configuration file storage method: Information such as pinned tags, last anchor tags, access frequency, etc. can be stored in a local configuration file, for example, in JSON, INI or XML format.

[0093] 2. Local database storage method: Tag information and behavior statistics can be stored in a local database such as SQLite, and query efficiency can be improved through indexed fields.

[0094] The items in the main view area of ​​the file manager are sorted using traditional sorting methods, partially implementing dynamic sorting methods, and fully implementing the method of this invention. The sorting results are as follows: Figure 2 As shown, where: The characteristics of traditional sorting methods are: folders are naturally prioritized, frequently accessed files are still placed after folders, and there is no tail constraint mechanism. Therefore, the sorting is: 1. Folder A (normal), 2. Folder B (normal), 3. File C (normal), 4. File D (frequently accessed), 5. Folder E (normal), 6. File F (frequently accessed).

[0095] The characteristics of the partially implemented dynamic sorting method are: it supports behavior-driven dynamic reordering; it promotes frequently accessed items; it still does not support top-level cross-level display; and it does not implement fixed tail. Therefore, its sorting is: 1 Folder A (normal), 2 File D (frequently accessed), 3 Folder B (normal), 4 File C (normal), 5 File F (frequently accessed), 6 Folder E (normal).

[0096] The key features of fully implementing this invention's method are: dynamic sorting + top placement + last-place anchoring. Top-placed files support cross-folder front placement and dynamic reordering of the main view. Last-place anchored objects are stably placed at the end, forming a complete hybrid sorting mechanism. Therefore, the sorting is: 1. File D (top placement), 2. File C (high-frequency access), 3. Folder A (normal), 4. Folder B (normal), 5. File F (normal), 6. Folder E (last-place anchoring).

[0097] Compared with the prior art, the dynamic rearrangement method for items in the main view area of ​​the file manager provided in this embodiment can achieve the following beneficial effects: 1. Improve the accessibility of frequently used items and reduce the time users spend searching in the directory.

[0098] 2. Set a top-priority tag for files to break the traditional sorting method of "folders always come before files". Through the hierarchical display mechanism, high-priority items are placed in a position that is more in line with the user's actual usage habits.

[0099] 3. Setting a last-position anchor mark introduces a last-position constraint from a sorting perspective, ensuring that a specific item remains stable at the end of the list even if it is frequently accessed.

[0100] 4. By using dynamic hybrid sorting, it takes into account both traditional browsing habits and dynamic usage habits. Users can observe that frequently used files are sorted to the top of the list during use, making it easier for users to understand and use.

[0101] 5. By using time decay and stable indexes, the stability of sorting results and interface consistency are improved, avoiding interface jitter caused by frequent refreshes.

[0102] 6. By introducing an operation depth score, the system can distinguish between shallow browsing and deep usage by users, giving files that have been deeply edited a higher sorting priority than files that have only been viewed, and more accurately reflecting the user's true usage intent.

[0103] 7. By correcting the covariate of the effective dwell time score, noise pollution of the dwell time during idle periods is eliminated, the purity of the user's real usage signal in the behavioral data is improved, and the ranking results are not affected by scenarios such as opening and leaving the computer.

[0104] 8. By quickly disabling the weighting score, statistical bias caused by accidental openings is suppressed, preventing the incorrect prioritization of files with similar names due to frequent accidental openings. This ensures that the sorting results more accurately reflect the user's true access intent rather than erroneous operations. Negative correction is used to exclude non-genuine usage from a negative perspective.

[0105] 9. Through behavioral trend prediction algorithms, the system can identify projects with rapidly increasing activity and proactively prioritize them, achieving a leap from passive sorting to proactive prediction, thereby improving the efficiency and experience of users in finding target projects.

[0106] 10. By using a stable index debouncing mechanism, the sorting results are prevented from repeatedly jumping between adjacent positions due to slight fluctuations in behavior scores, thus improving interface consistency.

[0107] 11. By extending metadata storage, the sorting state is persistently bound to the file object, and can be accurately restored after renaming, moving or refreshing, thus improving the persistence and recoverability of the sorting state.

[0108] A specific embodiment of the present invention discloses a dynamic rearrangement system for items within the main view area of ​​a file manager, comprising: The information acquisition module is used to acquire basic attribute information, preset tags and stable indexes of all direct sub-projects in the current directory, as well as user operation behavior data of the projects; wherein, the projects are folders or files; The numerical calculation module is used for each item to: determine a tag level value based on the preset tag; determine a type level value based on the preset tag and the type information included in the basic attribute information; wherein the type information is folder type or file type; determine a traditional sorting base score based on the basic attribute information; acquire user operation behavior data for the item, and determine a comprehensive behavior score and a behavior trend prediction score based on the operation behavior data; The comprehensive sorting order determination module is used to determine the comprehensive sorting order of all items based on the tag level value, the type level value, the behavior comprehensive score, the behavior trend prediction score, the traditional sorting base score, and the stable index of all items; The sorting module is used to perform a comprehensive sorting of all items based on the overall sorting order of all items.

[0109] It may also include a view rendering module to refresh the sorting results to the main view of the file manager; and a tag management module to maintain the top tag, the bottom anchor tag and the normal state, and support tag switching.

[0110] The above-described method and system embodiments are based on the same principles, and their related aspects can be referenced from each other to achieve the same technical effects. For specific implementation processes, please refer to the foregoing embodiments, which will not be repeated here.

[0111] Electronic device example: One specific embodiment of this application discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the dynamic rearrangement method for items in the main view area of ​​the file manager in the method embodiment.

[0112] Examples of readable storage media: One specific embodiment of this application discloses a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the dynamic rearrangement method for items within the main view area of ​​the file manager in the method embodiment.

[0113] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0114] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for dynamically rearranging items within the main view area of ​​a file manager, characterized in that, include: Retrieve basic attribute information, preset tags, and stable indexes of all direct sub-projects in the current directory; wherein, the project is a folder or a file; For each item: determine a tag level value based on the preset tag; determine a type level value based on the preset tag and the type information included in the basic attribute information; wherein, the type information is folder type or file type; determine a traditional sorting base score based on the basic attribute information; obtain user operation behavior data for the item, and determine a comprehensive behavior score and a behavior trend prediction score based on the operation behavior data; The overall ranking order of all items is determined based on the tag level value, type level value, behavior comprehensive score, behavior trend prediction score, traditional ranking base score, and stable index of all items. All items are comprehensively ranked based on their overall ranking order.

2. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 1, characterized in that, The preset markers include top markers, ordinary markers, and last anchor markers.

3. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 2, characterized in that, Determining the tag level value based on the preset tag includes: When the project's preset marker is the top marker, the marker level value is the first marker level value; When the preset marker of the project is a normal marker, the marker level value is the second marker level value; When the preset marker of the project is the last anchor marker, the marker level value is the third marker level value; The first marker level value is greater than the second marker level value, which is greater than the third marker level value.

4. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 3, characterized in that, Determining the type level value based on the preset marker and the type information included in the basic attribute information includes: When the preset marker is the top marker, the type level value of the items whose type information is folder type or file type is the first type level value; When the preset marker is a normal marker, and the type information is a folder type, the project's type level value is the first type level value; when the type information is a file type, the project's type level value is the second type level value. When the preset marker is the last anchor marker, the type level value of the item whose type information is folder type or file type is the third type level value; The first type level value is greater than the second type level value, which is greater than the third type level value.

5. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 1, characterized in that, The user's operational behavior data for the project includes access frequency, recent access time, dwell time, operation type, and number of times the app has been opened and closed within a preset time. The comprehensive score for behavior is determined based on the aforementioned operational behavior data, including: The access frequency is normalized to obtain the normalized access frequency; Determine the exponential decay function value of the most recent access time and the current time interval, and normalize the exponential decay function value to obtain a normalized most recent access score. The cumulative stay duration and effective stay duration are determined based on the stay duration. The cumulative stay duration is then normalized to obtain the normalized cumulative stay duration. The effective stay score is determined by the percentage of effective stay time based on the effective stay duration and the cumulative stay duration. The operation depth score is determined based on the operation type. The demotion score for fast shutdown is determined based on the number of times the device is opened and closed within a preset time period. The comprehensive behavioral score is determined based on the normalized access frequency, the normalized recent access score, the normalized cumulative dwell time, the effective dwell time score, the operation depth score, and the quick close demotion score.

6. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 5, characterized in that, Determining behavioral trend prediction scores based on the aforementioned operational behavior data includes: Construct a behavior phase space based on the aforementioned operational behavior data; Calculate the phase velocity and phase acceleration of the behavior phase space; The behavioral trend prediction score is determined based on the phase velocity and the phase acceleration.

7. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 6, characterized in that, Constructing a behavior phase space based on the aforementioned operational behavior data includes: Construct a three-dimensional behavior phase space, which includes the three-dimensional behavior state vector X_i(t) = [Frequency_i(t), Recency_i(t), Dwell_i(t)] of item i at time t, where Frequency_i(t) is the normalized access frequency of item i at time t, Recency_i(t) is the normalized recent access score of item i at time t, and Dwell_i(t) is the normalized cumulative dwell time of item i at time t; The phase velocity of the behavioral phase space is calculated using the following formula: V_i(t) = X_i(t) X_i(t Δt); Where V_i(t) is the phase velocity of item i at time t, and Δt is the preset time step; X_i(t) Δt) represents the value of item i at time t. The three-dimensional behavioral state vector of Δt; The phase acceleration of the behavior phase space is calculated using the following formula: A_i(t) = V_i(t) V_i(t Δt); Where A_i(t) is the summation velocity of item i at time t; V_i(t) Δt) represents the value of item i at time t. Phase velocity Δt; The behavioral trend prediction score is determined based on the phase velocity and the phase acceleration according to the following formula: TrendScore_i = α × |V_i(t)| × cos(θ_i) + β × |A_i(t)| × cos(φ_i); Where TrendScore_i is the predicted score of the behavioral trend of item i, θ_i is the angle between the phase velocity vector of item i and the direction of the increase in activity, φ_i is the angle between the phase acceleration vector of item i and the direction of the increase in activity, and α and β are weighting coefficients.

8. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 1, characterized in that, The stable index is the natural sequence number of each project when it first enters the current directory, among all projects in the current directory.

9. The method for dynamically rearranging items within the main view area of ​​a file manager according to claim 1, characterized in that, The project's basic attribute information and preset tags, the traditional sorting base score, the behavior comprehensive score, the behavior trend prediction score, and the stable index are stored in any of the following forms: Extended metadata storage methods; Local configuration file storage method; Local database storage method.

10. A dynamic rearrangement system for items within the main view area of ​​a file manager, characterized in that, include: The information acquisition module is used to acquire basic attribute information, preset tags and stable indexes of all direct sub-projects in the current directory, as well as user operation behavior data of the projects; wherein, the projects are folders or files; The numerical calculation module is used for each item to: determine a tag level value based on the preset tag; determine a type level value based on the preset tag and the type information included in the basic attribute information; wherein the type information is folder type or file type; determine a traditional sorting base score based on the basic attribute information; acquire user operation behavior data for the item, and determine a comprehensive behavior score and a behavior trend prediction score based on the operation behavior data; The comprehensive sorting order determination module is used to determine the comprehensive sorting order of all items based on the tag level value, the type level value, the behavior comprehensive score, the behavior trend prediction score, the traditional sorting base score, and the stable index of all items; The sorting module is used to perform a comprehensive sorting of all items based on the overall sorting order of all items.

Images