A content distribution management system based on multi-platform experience space

Abstract

The present invention relates to the technical field of content distribution, and in particular to a content distribution management system based on a multi-platform experience space, comprising: a request identification module, used to determine user request information, and perform preference detection according to the user request information to construct a preference adaptation coefficient; a request confirmation module, used to determine the relevance of distribution content corresponding to the user request information and the preference of a storage location of the distribution content according to the acquired preference adaptation coefficient; a distribution determination module, used to divide the distribution content into multiple data sets according to the relevance of the distribution content corresponding to the user request information, and send the data sets to target users according to set priorities; the distribution determination module also includes a first determination unit and a second determination unit; the first determination unit is used to determine the interruption probability of data set transmission, and determine the accuracy of data set segmentation under the corresponding interruption probability; the personalization and accuracy of content distribution are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of content distribution, and in particular to a content distribution management system based on a multi-platform experience space. Background Art

[0002] With the rapid development of XR (extended reality, including virtual reality VR, augmented reality AR and mixed reality MR) technology, multi-platform XR content distribution has become an important technical field. In order to meet the XR experience needs of different users on various devices and platforms, it is necessary to control the devices to send content.

[0003] Since the data formats, data requirements, experience space display requirements, and space delivery methods used by each different platform are completely different, basic XR content needs to be packaged into different resource files according to different platform requirements, and then displayed on different platform display terminals using different parsing methods. At this time, if the actual distributed content is not controlled, it will lead to the inability to accurately capture the user's personalized preferences during distribution, resulting in management difficulties. Summary of the invention

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A content distribution management system based on a multi-platform experience space, comprising:

[0006] A request identification module is used to determine the user's request information, perform preference detection based on the user's request information, and construct a preference adaptation coefficient;

[0007] A request confirmation module, used to determine the relevance of the distribution content corresponding to the user's request information and the preference of the storage location of the distribution content according to the obtained preference adaptation coefficient;

[0008] A distribution determination module, used to divide the distribution content into multiple data sets according to the relevance of the distribution content corresponding to the user's request information, and send the data sets to the target user according to the set priority;

[0009] The distribution determination module further includes a first determination unit and a second determination unit;

[0010] A first determining unit, used to determine the interruption probability of data set transmission, and determine the accuracy of data set segmentation under the corresponding interruption probability;

[0011] The second determining unit is used to determine the influence of each data set on the overall data transmission when the data sets are transmitted.

[0012] The beneficial effects of the present invention are as follows: by constructing a preference relationship network and mining behavioral element vectors and adaptive behavior vectors, the present invention achieves a deep understanding and precise matching of user preferences, thereby improving the personalization and satisfaction of distributed content; by optimizing the transmission interruption probability and segmentation accuracy of data sets, the interruptions and errors in the transmission process are reduced, the transmission efficiency and accuracy are improved, and the user experience is enhanced; by real-time monitoring of key indicators in the transmission process and dynamically adjusting user scheduling strategies, the optimization and adaptation of distribution strategies are achieved, and the flexibility and robustness of the system are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0014] Figure 1 It is a system diagram of a content distribution management system based on a multi-platform experience space. DETAILED DESCRIPTION

[0015] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. If no specific techniques or conditions are specified in the embodiments, the techniques or conditions described in the literature in the art or the product specifications are used.

[0016] See also Figure 1 , a content distribution management system based on a multi-platform experience space, comprising:

[0017] The request identification module is used to determine the user's request information, perform preference detection based on the user's request information, and construct a preference adaptation coefficient; the request information includes video data, instructions, related object tags, and context information related to the requested target object, such as platform type, device information, historical behavior records, etc.

[0018] The request confirmation module is used to determine the relevance of the distribution content corresponding to the user's request information and the preference of the storage location of the distribution content according to the obtained preference adaptation coefficient.

[0019] The distribution determination module is used to divide the distribution content into multiple data sets according to the relevance of the distribution content corresponding to the user's request information, and send the data sets to the target users according to the set priorities.

[0020] The distribution determination module further includes a first determination unit and a second determination unit.

[0021] The first determining unit is used to determine the interruption probability of data set transmission and determine the accuracy of data set segmentation under the corresponding interruption probability.

[0022] The second determining unit is used to determine the influence of each data set on the overall data transmission when the data sets are transmitted.

[0023] At this time, the data content corresponding to the user's request information is obtained from the cache space. If there is no cache space, the missing file content is determined based on the correlation between the distributed content and the user's request information, and the order of data transmission is adjusted according to priority. At the same time, the corresponding transmission quality of the transmitted data is determined, such as determining the interruption probability of the data set transmission, to determine whether the current transmitted data will have corresponding data loss during data transmission, and whether the current data block size for data segmentation is appropriate when setting the data cache.

[0024] For example, a user wants to obtain cooking-related data using an MR device. At this time, the preference adaptation coefficient indicates the user's preference, and selects video content suitable for the user based on the current main preferred content. Then, the user's request is confirmed based on the correlation between the user's preference adaptation coefficient and the user's request information. At the same time, it is necessary to determine the storage location of the corresponding content to determine the priority of the currently acquired data so that the data required by the user can be sent in a timely manner. At this time, the data is divided into multiple data sets according to the size of the data to be sent, and the interruption probability in the data packet allocated in each data set is determined to ensure that the data set can be accurately retransmitted when it is interrupted. At the same time, it is also necessary to determine whether the division of the data blocks set in the cache position is correct, so as to improve the transmission quality and enable key content to be sent first.

[0025] The request recognition module's preference detection is handled in the following ways:

[0026] Analyze the user's request information to determine the target object corresponding to the request information and the user behavior corresponding to the target object; at this time, the user-related context information is obtained from the request information to determine the platform type, device information, and historical behavior records corresponding to the user, and verify the corresponding target object when using it at this time, such as the corresponding character when currently using the MR device to play, or the corresponding content in the XR space when using other platforms, and determine the user's main target object based on the current user's operation, such as identifying the position and angle of the corresponding camera to determine the user's current behavior bias, and combine historical data to predict the data needed by the user in advance. The data prediction process is to obtain the preference adaptation coefficient and match the data needed by the user according to this coefficient.

[0027] The target object and the user behavior are processed to generate a plurality of preference relationship networks corresponding to the target object and the user behavior; each preference relationship network represents the preference of the target object and the user behavior in one aspect.

[0028] User behavior mining is performed on the preference relationship network to extract behavior element vectors, which include behavior subject, location, time, content, purpose, and emotion.

[0029] The preference relationship network is subjected to adaptation behavior mining to extract an adaptation behavior vector, wherein the adaptation behavior vector includes a first degree of adaptation between a target object and user behavior and a second degree of adaptation that affects each other between the target object and the first degree of adaptation.

[0030] The first fitness degree is used to indicate whether there is a matching relationship between the target object and the user behavior. It focuses on whether the user behavior points to or conforms to the characteristics, needs or expectations of the target object. For example, if the user expects to walk or switch translation, the corresponding data at this time needs to include the data of this operation to meet the user's needs and expectations. The second fitness degree is more inclined to determine whether the first fitness degree will be affected by other target objects after obtaining the first fitness degree, resulting in corresponding abnormalities in user behavior, and to determine the influence of the target object on the user behavior in the context of corresponding order and logical relationship.

[0031] The behavior element vector and the adaptation behavior vector are input into the data training model, the importance of the behavior element vector and the adaptation behavior vector is determined, and the preference relationship network is integrated according to the importance of the behavior element vector and the adaptation behavior vector to generate an integrated preference relationship network.

[0032] According to the integrated preference relationship network, the preference adaptation coefficient between the user request and the user preference is calculated.

[0033] The first degree of fitness indicates the degree of match between the user behavior and the target object. By determining the weight and frequency of the user behavior implemented by the target object, it is determined whether the target object matches the user behavior to determine the first degree of fitness:

[0034] ;

[0035] in, Indicates the first degree of fit between the target object and the user behavior. is the weight of the k-th user behavior, is the frequency of the kth user behavior between the target object and the user behavior, , is the number of user actions.

[0036] The second fitness degree indicates whether the current target object will be affected by other target objects after obtaining the first fitness degree, resulting in a mismatch between the identified user behavior and the target object, that is, determining the degree of mutual influence between the target object and the first fitness degree:

[0037] ;

[0038] in, The second fitness degree represents the mutual influence between the target object and the first fitness degree, is the degree to which the target object is affected by the mth target object among other target objects, Indicates the first degree of fit between the target object and the user behavior. , is the number of other target objects.

[0039] The degree to which the target object is influenced by the mth target object among other target objects is obtained by taking a weighted average of the similarity values ​​of the current target object and the other target objects in user behaviors, and then obtaining the degree of influence of the current target object and the mth target object among other target objects.

[0040] Methods for determining the importance of the behavior element vector and the adaptation behavior vector include:

[0041] A decision tree is constructed using the behavior element vector and the adaptation behavior vector. The root node of the decision tree is the target object, and the child nodes of each root node are the corresponding user behaviors. The user behaviors are sorted in sequence according to the size of the first fitness degree and the second fitness degree. The importance of the corresponding features on the corresponding nodes of the allocated decision tree is used as the importance of the behavior element vector and the adaptation behavior vector.

[0042] By calculating the reduction in impurity corresponding to the decision tree, the importance of the corresponding features on the corresponding nodes of the decision tree is determined.

[0043] First, determine the Gini impurity of node D on the decision tree, obtain the probability of each user behavior on node D, and determine the total number of user behaviors on node D to determine the Gini impurity of node D:

[0044] ;

[0045] in, is the Gini impurity of node D, is the probability of the user behavior, , is the number of user actions;

[0046] The probability of user behavior is obtained by calculating the probability of each selected user behavior appearing at the node when it is assigned to the node on the decision tree.

[0047] Compared with node D, feature a has less impurity which can be expressed as:

[0048] ;

[0049] in, The impurity of feature a is less than that of node D. , They are the child nodes after node D is split, Is a node The Gini impurity of Is a node The Gini impurity of .

[0050] Based on the overall decision tree, the importance of feature a is expressed as:

[0051] ;

[0052] in, is the importance of feature a, The impurity of feature a is less than that of node t. is the number of decision tree nodes, .

[0053] For the integration method of preferred relationship network:

[0054] According to the corresponding importance, the behavior element vectors and adaptation behavior vectors with the highest similarity are selected for integration, and the adjacent behavior element vectors and adaptation behavior vectors are superimposed on a group of behavior element vectors and adaptation behavior vectors with the greatest similarity to obtain the integrated preference relationship network.

[0055] For the preference adaptation coefficient, the user behavior corresponding to the current request information is extracted, and the average of the importance of the behavior element vector corresponding to the user behavior and the adaptation behavior vector is taken as the preference adaptation coefficient.

[0056] The preference adaptation coefficient obtained at this time is expressed as the average value of the relative importance of the behavior element vector and the adaptation behavior vector for a single behavior or multiple behaviors. It is used to evaluate the importance of the behavior existing in the corresponding request information when the user request information appears, and filter out the content required by the user according to the relatively important parts.

[0057] The request confirmation module is used to determine the relevance of the distribution content corresponding to the user's request information and the preference of the storage location of the distribution content according to the obtained preference adaptation coefficient; its implementation method includes:

[0058] The relevance A of the distributed content corresponding to the user's request information can be expressed as the product of the similarity measurement result sim(c,r) between the distributed content c and the request information r and the preference adaptation coefficient C: ; and use the highest ranked content as the distribution content.

[0059] The preference of the storage location of the distributed content: The preference O can be the geographical distance d between the node and the user and the node load For example: ;in is the maximum load of the node.

[0060] The distribution content and storage situation are determined based on the preference and relevance of the current storage location.

[0061] Optimizing the storage location of distributed content can significantly reduce data transmission delays and increase content loading speed. For example, storing content on a server near the user can greatly shorten user waiting time and improve user experience.

[0062] The implementation of the distribution determination module includes:

[0063] The distribution determination module first divides the obtained preference adaptation coefficient and the correlation of the distribution content into multiple data sets of the same size, and sends them according to the priority of each data set.

[0064] The priority set by the distribution determination module may be: the product of the relevance of the distribution content and the preference of the storage location of the distribution content, and the data sets are sorted from large to small according to the product, and the sorted data sets are output in sequence.

[0065] The distribution determination module further includes a first determination unit and a second determination unit;

[0066] The processing method of the first determination unit includes:

[0067] The total number of data packets and the number of lost data packets when the data set is sent are obtained, and the interruption probability of the data set is determined according to the total number of data packets and the number of lost data packets.

[0068] The number of data blocks into which the data set is segmented during transmission and the number of data blocks recovered by the receiving end are obtained to determine the accuracy of the data set segmentation.

[0069] The interruption probability of the segmented data set is determined by using network monitoring tools or APIs to collect the real-time status of the network, including but not limited to bandwidth, latency, packet loss rate, etc.; record previous data transmission conditions, including statistics on successful and failed transmissions, and determine the interruption probability under specific network conditions in historical transmission data, as well as whether interruptions will occur when accessing specific data and storage content, so as to predict possible packet loss rates and interruption possibilities in the future.

[0070] If there is a chance of interruption, redundant coding is added to the identified data set to improve the network's ability to recover when problems occur, and the size of the segmented data set and the number of data blocks in the data set are adjusted to adjust the error resolution capability in the face of transmission problems.

[0071] The interruption probability is expressed as:

[0072] ;

[0073] in, is the interruption probability, is the number of packets lost in the data set, is the total number of packets in the data set.

[0074] The segmentation accuracy is expressed as follows: based on the number of data blocks into which the data set is segmented during transmission, the number of redundant backups corresponding to each data block and the number of data blocks recovered by the receiving end are obtained to determine the segmentation accuracy:

[0075] ;

[0076] in, is the segmentation accuracy, is the number of blocks into which the data set is divided, Indicates the number of redundant copies of each data block. is the number of data blocks recovered by the receiving end.

[0077] The data packet described at this time is in the form of a combination of the data in the original data set and the data header, message, etc. of the target during transmission. The data block represents the original data set that has been segmented. The accuracy of the data set segmentation is verified at this time to determine that in the event of an interruption, the number of data blocks successfully received and recovered by the receiving end should be as close as possible to or equal to the number of data blocks in the original data set to maintain the integrity and consistency of the current data transmission.

[0078] When the probability of transmission interruption is low, the risk of data blocks being lost or damaged during transmission is small, so the receiving end can more accurately restore the original data set; in this case, the segmentation accuracy is high and can meet the user's needs for data integrity and consistency; on the contrary, when the probability of transmission interruption is high, data blocks are easily lost or damaged during transmission.

[0079] The second determining unit determines the influence of the data set on the overall data transmission, including setting a user scheduling strategy according to the interruption probability and the segmentation accuracy.

[0080] Setting the user scheduling strategy includes determining the data transmission time corresponding to the interruption probability, and dividing the data transmission time into the communication time with the base station and the user transmission time. The user transmission time is divided into the adaptive transmission time and the set time gap.

[0081] The adaptive transmission time is the transmission time set according to the size of the data packet. The time gap is used to determine the transmission success rate of the corresponding transmitted data packet within the base station communication time. When the data transmission is successful within the base station communication time, the interruption probability of the currently transmitted data packet is determined according to the time gap. If the interruption probability is less than the preset interruption probability value, the corresponding data packet is transmitted according to the adaptive transmission time; otherwise, the data packet with the smallest interruption probability is transmitted. When the data packet is successfully transmitted, the duration of the time gap is determined, and the cached data blocks stored in the data set within the duration of the time gap are determined, and the segmentation accuracy of each data block is determined, and each data block is divided into multiple data packets for transmission in turn according to the timing feedback instruction.

[0082] The timing feedback instruction is expressed as: according to the segmentation accuracy of each data block, the transmission time required for each data block is determined, and the transmission time of the data block is less than the adaptive transmission time. If the time gap is exceeded, the current data block is re-segmented until the transmission time of each data block is less than the adaptive transmission time.

[0083] Assume that there is a data set that needs to be transmitted from a user equipment to a base station; the data set is divided into multiple data blocks, and each data block will be further divided into data packets for transmission.

[0084] Based on historical data or prediction models, the interruption probability during data transmission is determined; the overall data transmission time is divided into communication time with the base station and user transmission time; the user transmission time is further subdivided into adaptive transmission time and set time gap.

[0085] The adaptive transmission time is dynamically set according to the size of the data packet; the time gap is used to evaluate the transmission success rate of the data packet within the base station communication time.

[0086] When a data packet is successfully transmitted within the base station communication time, the time gap is used to evaluate the interruption probability of the data packet; if the interruption probability is less than the preset threshold, the next data packet is transmitted; if the interruption probability is greater than or equal to the preset threshold, the data packet with the smallest interruption probability is selected for transmission.

[0087] When a data packet is successfully transmitted, the data blocks cached in the data set for the duration of the time gap are determined; the segmentation accuracy of each data block is evaluated and transmitted according to the timing feedback instructions.

[0088] The timing feedback instruction determines the transmission time of each data block according to the segmentation accuracy; if the transmission time of a data block exceeds the adaptive transmission time, the data block is re-segmented; this process is repeated until the transmission time of all data blocks is less than the adaptive transmission time and they are transmitted in sequence.

[0089] In this way, the current scheme is able to dynamically adjust the user scheduling strategy to optimize the data transmission process based on the interruption probability and segmentation accuracy. This helps to improve the efficiency and reliability of data transmission, especially in unstable communication environments or when the packet sizes vary greatly.

[0090] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention and they are still covered by the protection scope of the present invention.

Claims

1. A content distribution management system based on a multi-platform experience space, characterized in that: include: The request identification module is used to determine the user's request information, perform preference detection based on the user's request information, and construct a preference adaptation coefficient; A request confirmation module, used to determine the relevance of the distribution content corresponding to the user's request information and the preference of the storage location of the distribution content according to the obtained preference adaptation coefficient; A distribution determination module, used to divide the distribution content into multiple data sets according to the relevance of the distribution content corresponding to the user's request information, and send the data sets to the target user according to the set priority; The distribution determination module further includes a first determination unit and a second determination unit; A first determining unit, used to determine the interruption probability of data set transmission, and determine the accuracy of data set segmentation under the corresponding interruption probability; A second determining unit, used to determine the impact of each data set on the overall data transmission when the data set is transmitted; The request recognition module's preference detection is handled in the following ways: Analyze the user's request information to determine the target object corresponding to the request information and the user behavior corresponding to the target object; Process the target object and the user behavior to generate a plurality of preference relationship networks corresponding to the target object and the user behavior; each preference relationship network represents the preference of the target object and the user behavior in one aspect; Conduct user behavior mining on the preference relationship network and extract behavior element vectors; Performing adaptive behavior mining on the preference relationship network to extract an adaptive behavior vector, wherein the adaptive behavior vector includes a first degree of adaptation between the target object and the user behavior and a second degree of adaptation that affects each other between the target object and the first degree of adaptation; Taking the behavior element vector and the adaptation behavior vector as input, inputting them into the data training model, determining the importance of the behavior element vector and the adaptation behavior vector, integrating the preference relationship network according to the importance of the behavior element vector and the adaptation behavior vector, and generating an integrated preference relationship network; According to the integrated preference relationship network, the preference adaptation coefficient between the user request and the user preference is calculated.

2. A content distribution management system based on a multi-platform experience space according to claim 1, characterized in that: First fitness: Among them, A 1 represents the first degree of fit between the target object and the user behavior, ω k is the weight of the k-th user behavior, f k is the frequency of the kth user behavior between the target object and the user behavior, k = 1, 2,,,K, where K is the number of user behaviors.

3. A content distribution management system based on a multi-platform experience space according to claim 1, characterized in that: Second suitability: Among them, A 2 The second fitness degree, α, represents the interaction between the target object and the first fitness degree. m is the degree to which the target object is affected by the mth target object among other target objects, A 1 represents the first degree of fitness between the target object and the user behavior, m=1,2,,,M, where M is the number of other target objects.

4. The content distribution management system based on multi-platform experience space according to claim 1, characterized in that: The importance of the behavior element vector and the adaptation behavior vector is determined by: A decision tree is constructed using the behavior element vector and the adaptation behavior vector. The root node of the decision tree is the target object, and the child nodes of each root node are the corresponding user behaviors. The user behaviors are sorted in sequence according to the size of the first fitness degree and the second fitness degree. The importance of the corresponding features on the corresponding nodes of the allocated decision tree is used as the importance of the behavior element vector and the adaptation behavior vector.

5. The content distribution management system based on multi-platform experience space according to claim 1, characterized in that: For the integration method of preferred relationship network: According to the corresponding importance, the behavior element vectors and the adaptation behavior vectors are selected to integrate the behavior element vectors and the adaptation behavior vectors with the highest similarity, and the adjacent behavior element vectors and the adaptation behavior vectors are superimposed to a group of behavior element vectors and the adaptation behavior vectors with the greatest similarity to obtain the integrated preference relationship network; For the preference adaptation coefficient, the user behavior corresponding to the current request information is extracted, and the average of the importance of the behavior element vector corresponding to the user behavior and the adaptation behavior vector is taken as the preference adaptation coefficient.

6. A content distribution management system based on a multi-platform experience space according to claim 1, characterized in that: The request confirmation module implementation methods include: The relevance A of the distributed content corresponding to the user's request information is expressed as the product of the similarity measurement result sim(c,r) between the content vector c and the user's request vector r and the preference adaptation coefficient C: A = sim(c,r)·C; the content with the highest ranking is used as the distributed content; The preference of the storage location for distributing content: The preference O is a function of the geographical distance d between the node and the user and the node load nl. where nl max is the maximum load of the node.

7. The content distribution management system based on multi-platform experience space according to claim 1, characterized in that: The processing method of the first determination unit includes: Obtaining the total number of data packets and the number of lost data packets when the data set is being sent, and determining the interruption probability of the data set according to the total number of data packets and the number of lost data packets; The number of data blocks into which the data set is segmented during transmission and the number of data blocks recovered by the receiving end are obtained to determine the accuracy of the data set segmentation.

8. The content distribution management system based on multi-platform experience space according to claim 1, characterized in that: The second determining unit determines the impact of the data set on the overall data transmission, including setting a user scheduling strategy according to the interruption probability and the segmentation accuracy; Setting the user scheduling strategy includes determining the data transmission time corresponding to the interruption probability, and dividing the data transmission time into the communication time with the base station and the user transmission time, and the user transmission time is divided into the adaptive transmission time and the set time interval; When data transmission is successful within the base station communication time, the interruption probability of the currently transmitted data packet is determined according to the time gap. If the interruption probability is less than the preset interruption probability value, the corresponding data packet is transmitted at the adaptive transmission time; Otherwise, the data packet with the smallest probability of interruption will be transmitted. When the data packet is successfully transmitted, the duration of the time gap is determined, and the cached data blocks stored in the data set during the duration of the time gap are determined, and the segmentation accuracy of each data block is determined, and each data block is divided into multiple data packets for transmission in turn according to the timing feedback instruction.

9. A content distribution management system based on a multi-platform experience space according to claim 8, characterized in that: The timing feedback instruction is expressed as: according to the segmentation accuracy of each data block, the transmission time required for each data block is determined, and the transmission time of the data block is less than the adaptive transmission time. If the time gap is exceeded, the current data block is re-segmented until the transmission time of each data block is less than the adaptive transmission time.

Images