CN109635153B

CN109635153B - Migration path generation method, device and storage medium

Info

Publication number: CN109635153B
Application number: CN201811400866.6A
Authority: CN
Inventors: 黄安埠
Original assignee: Tencent Music Entertainment Technology Shenzhen Co Ltd
Current assignee: Tencent Music Entertainment Technology Shenzhen Co Ltd
Priority date: 2018-11-22
Filing date: 2018-11-22
Publication date: 2023-01-13
Anticipated expiration: 2038-11-22
Also published as: CN109635153A

Abstract

The embodiment of the invention discloses a migration path generation method, a migration path generation device and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: dividing objects with the same information on the designated dimension into the same category to obtain a plurality of categories, wherein each category comprises at least one object; respectively extracting the features of the detail information of the object of each category to obtain feature vectors of a plurality of categories, and further determining the similar categories of each category in the plurality of categories; creating a correlation network according to the plurality of categories and the similar category of each category; according to the correlation network, determining a transition path from a starting point category to an end point category, wherein the starting point category and the end point category are any two categories of the multiple categories. The method and the device provide an information basis for subsequent function expansion, and labels do not need to be manually set, so that the waste of artificial resources is avoided, the labor cost is reduced, the problem of low accuracy caused by strong subjectivity of the set labels is avoided, and the accuracy is improved.

Description

Migration path generation method, device and storage medium

Technical Field

The embodiment of the invention relates to the technical field of computers, in particular to a migration path generation method, a migration path generation device and a storage medium.

Background

Listening to songs is a common way for people to enrich their spiritual lives, and users often listen to songs sung by their favorite singers, which may change over time.

In the related art, a plurality of tags may be manually set for each singer, and the plurality of tags may be a gender tag, a time tag, a language tag, and the like. And performing category division according to the labels of each singer to obtain a plurality of categories, wherein each category comprises at least one singer. The method comprises the steps of obtaining playing records of a plurality of users, wherein the playing records of each user comprise data of songs played by the users, singers singing the songs, playing conditions of the songs and the like, and determining the category of each user according to the playing records of each user and the category to which the singers in the playing records belong, wherein the category is the category to which the singers liked by the users belong, so that a user group of each category can be obtained.

The mode needs to manually set the label for each singer, so that higher labor cost is consumed, and the set label has stronger subjectivity, so that the accuracy is lower.

Disclosure of Invention

The embodiment of the invention provides a migration path generation method, a migration path generation device and a storage medium, which can solve the problems in the related art. The technical scheme is as follows:

in one aspect, a migration path generation method is provided, where the method includes:

dividing objects with the same information on a designated dimension into the same category to obtain a plurality of categories, wherein each category comprises at least one object, and the identification of each category is represented by the information of the at least one object on the designated dimension;

respectively extracting features of the detail information of the object of each category to obtain feature vectors of the categories, and determining similar categories of each category in the categories according to the feature vectors of the categories, wherein the similarity between any category and the similar categories is greater than the similarity between non-similar categories;

creating a relevance network according to the plurality of categories and the similar categories of each category, wherein the relevance network comprises the plurality of categories, and each category is associated with the corresponding similar category;

and determining a transition path from a starting point category to an end point category according to the correlation network, wherein the starting point category and the end point category are any two categories of the plurality of categories.

In another aspect, a migration path generation apparatus is provided, the apparatus including:

the dividing module is used for dividing the objects with the same information on the designated dimension into the same category to obtain a plurality of categories, each category comprises at least one object, and the identification of each category is represented by the information of the at least one object on the designated dimension;

the feature extraction module is used for respectively extracting features of the detail information of the objects of each category to obtain feature vectors of the categories, wherein each category comprises at least one object;

the similar category determining module is used for determining similar categories of each category in the multiple categories according to the feature vectors of the multiple categories, and the similarity between any category and the similar categories is greater than the similarity between non-similar categories;

a creating module, configured to create a correlation network according to the multiple categories and the similar categories of each category, where the correlation network includes the multiple categories, and each category is associated with a corresponding similar category;

and the migration path determining module is used for determining a migration path from a starting point category to an end point category according to the correlation network, wherein the starting point category and the end point category are any two categories of the plurality of categories.

In another aspect, an apparatus for migration path generation is provided, the apparatus including a processor and a memory, the memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by the processor to implement the operations as in the migration path generation method.

In yet another aspect, a computer-readable storage medium having stored therein at least one instruction, at least one program, set of codes, or set of instructions loaded by a processor and having an operation to implement the migration path generation method as it has is provided.

According to the method, the device and the storage medium provided by the embodiment of the invention, feature extraction is carried out on the detail information of at least one object in each category to obtain feature vectors of multiple categories, the similar category of each category is obtained according to the feature vectors, each category is associated with the similar category, a correlation network is created, the migration path of any two categories in the multiple categories can be obtained according to the correlation network, the migration probability between any two categories is not only calculated, the migration path can represent other categories passed by the migration between the two categories, the process of step-by-step migration between two or more categories is simulated, other categories used for transition are provided for the migration between any two non-similar categories, the information quantity is improved, and an information basis is provided for subsequent function expansion. In addition, labels do not need to be manually set, so that the waste of manual resources is avoided, the labor cost is reduced, the problem of low accuracy caused by strong subjectivity of the set labels is avoided, and the accuracy of the migration path is improved.

And in the process of obtaining the similar categories, a multi-dimensional space is created according to the multi-dimensional feature vector of each category, the multi-dimensional space is divided into a plurality of areas according to the position of each category in the multi-dimensional space to obtain the similar categories of each category, the feature values of each category in a plurality of dimensions are adopted in the process of determining the similar categories, the adopted information is more comprehensive, and the accuracy of the similar categories is improved. And moreover, a multi-dimensional space is created, similar categories are searched in the region to which the multi-dimensional space belongs and the adjacent region after region division is carried out in the multi-dimensional space, the calculation amount is reduced, and the calculation speed is accelerated.

And selecting a migration path with the least number of categories from the multiple migration paths, or selecting a migration path with the smallest migration distance from the multiple migration paths, finding the shortest migration path between the starting point category and the end point category as much as possible, namely finding the least transition categories passing through when the categories migrate from the starting point category to the end point category, and analyzing the fastest migration mode between the starting point category and the end point category.

Moreover, on the basis of the migration path, the migration path can be applied to a recommended object or a scene for performing reverse analysis on the recommended object, so that a reasonable object can be recommended for a user, unreasonable recommendation can be timely searched out, the recommendation accuracy is improved, the application range is expanded, and more functions are expanded.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a migration path generation method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a multi-dimensional space partition according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a binary tree according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a correlation network provided by an embodiment of the present invention;

FIG. 5 is a diagram illustrating an adjacency list structure according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of a display interface of a migration path according to an embodiment of the present invention;

fig. 7 is a schematic structural diagram of a migration path generating apparatus according to an embodiment of the present invention;

fig. 8 is a schematic structural diagram of a terminal according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a server according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings.

Before describing embodiments of the present invention in detail, the concepts involved are explained as follows:

1. object: the data can exist as independent individuals, and from the aspect of data format, the object can comprise data in various formats such as audio data, video data, webpage data, picture data and the like, such as songs, movies and the like.

Each object has detail information including information in multiple dimensions, such as songs including singers, duration, singing language, release time, and movies including director, duration, language, release time, and other details.

2. The category: the method is a category obtained by dividing a plurality of objects according to the condition that whether information on a specified dimension is the same or not, wherein each category comprises at least one object with the same information on the specified dimension, and each category is represented by the information on the specified dimension.

For example, the songs may be divided into a plurality of categories according to the difference of the singers, the singers of the songs in each category are the same, and the category names are the names of the songs. Alternatively, the movies may be divided into categories according to the difference of directors, the directors of the movies in each category are the same, and the category name is the name of the director.

In the related art, a plurality of labels may be manually set for each singer, the plurality of labels being used to describe each singer, and classification may be performed according to the plurality of labels of each singer to obtain a plurality of categories, wherein each category includes at least one singer. And acquiring the play records of a plurality of users, and determining the category of each user according to the play record of each user. And counting according to the playing records of the plurality of users in the plurality of time periods to obtain the user groups of each category in the plurality of time periods, and comparing the user groups of each category in different time periods to obtain the transition probability between any two categories. The mode needs to manually set a label for each singer, so that higher labor cost is consumed, and the set label has stronger subjectivity, so that the accuracy of the determined migration probability is lower.

The embodiment of the invention provides a migration path generation method, which comprises the steps of extracting features of detailed information of at least one object in the same category to obtain feature vectors of multiple categories, obtaining a similar category of each category according to the feature vectors, associating each category with the similar categories, creating a correlation network, and further obtaining migration paths of any two categories in the multiple categories, wherein the migration paths can represent other categories passing through during migration between the two categories, so that a gradual migration process between two or more categories is simulated, and the information quantity is improved. In addition, labels do not need to be manually set, so that the waste of manual resources is avoided, the labor cost is reduced, the problem of low accuracy caused by strong subjectivity of the set labels is avoided, and the accuracy of the migration path is improved.

The embodiment of the invention can be applied to the scene of acquiring the migration path between any two categories. For example, a large number of songs in the internet are used as objects, categories to which the objects belong are obtained by dividing according to singers singing the songs, and each category is represented by the singers, then by adopting the method provided by the embodiment of the present invention, a correlation network including a plurality of singers is created, and a migration path between any two singers can be determined according to the correlation network, wherein the migration path represents a style transition process between the plurality of singers.

Further, other singers may be recommended for the user based on the migration path between any two singers, and the singer who the user currently likes the song. For example, a plurality of singers to which the user listens frequently are acquired, and songs of other singers included in a migration path may be recommended for the user based on the migration path between any two singers in the plurality of singers.

Or, a large number of movies in the internet are used as objects, categories to which the objects belong are obtained by dividing according to directors who make the movies, and each category is represented by a director.

Further, other directors may be recommended for the user based on the migration path between any two directors, as well as the director of the user currently enjoying the movie. For example, a director who acquires a plurality of movies frequently watched by the user may recommend movies of other directors included in a migration path for the user according to the migration path between any two directors in the plurality of directors.

Or, a large number of goods in the internet are used as objects, categories to which the objects belong are obtained by dividing according to the types of the goods to which the goods belong, and each category is represented by a type of the goods, then a correlation network including a plurality of types of the goods is created by using the method provided by the embodiment of the present invention, and a migration path between any two types of the goods can be determined according to the correlation network, wherein the migration path represents a transition path between the plurality of types of the goods.

Further, other types of goods may be recommended to the user based on the migration path between any two goods types, and the goods types that the user currently frequently purchases. For example, a plurality of commodity types frequently purchased by the user are acquired, and according to a migration path between any two commodity types in the plurality of commodity types, commodities of other commodity types included in the migration path may be recommended for the user.

Fig. 1 is a flowchart of a migration path generation method according to an embodiment of the present invention. The execution subject of the embodiment of the present invention is a generation apparatus, and referring to fig. 1, the method includes:

101. the generation device divides objects with the same information in a specified dimension into the same category to obtain a plurality of categories, and performs feature extraction on detailed information of the objects in each category to obtain feature vectors of the plurality of categories.

The generating means may be a terminal or a server. When the generating device is a server, the server may collect a plurality of objects accessed by a plurality of terminals and detailed information of the plurality of objects, or collect a plurality of objects uploaded by a plurality of publishers and detailed information of the plurality of objects, store the collected objects and the detailed information thereof by the server, and divide and count the stored detailed information to obtain feature vectors of a plurality of categories. Or, when the generating device is a terminal, the terminal receives the plurality of objects and the detail information of the plurality of objects transmitted by the server, stores the plurality of objects and the detail information of the plurality of objects, and divides and counts the stored detail information to obtain the feature vectors of the plurality of categories.

In the embodiment of the invention, each object in the plurality of objects has detail information, and the detail information of each object comprises information of a plurality of dimensions, and is used for describing the corresponding object from the plurality of dimensions. According to the information of the detail information of each object on the designated dimension, the objects can be divided into different categories to obtain a plurality of categories, wherein each category comprises at least one object with the same information on the designated dimension, and the identification of each category is represented by the information of the at least one object on the designated dimension.

The specified dimension is any one of the multiple dimensions, and can be determined according to the requirements of the user, or the dimension most concerned by the user can be determined as the specified dimension according to the popularity degree of the information on each dimension in the internet.

For example, when the object is a song, the detail information of each song may include details of a singer, a time length, a singing language, a release time, and the like. For the plurality of songs, if the plurality of songs are divided according to the singer, a plurality of categories can be obtained, each category is represented by one singer, and at least one object of each category represents at least one song sung by the singer.

Alternatively, when the plurality of objects are a plurality of commodities, the detail information of each commodity may include information such as a color, price, model, manufacturer, and commodity type of the commodity. For the plurality of commodities, if the commodities are classified according to commodity types, a plurality of categories can be obtained, each category represents a commodity type, and at least one object of each category represents at least one commodity belonging to each commodity type.

The generating device acquires the detail information of each object, and performs feature extraction on the detail information of each object to obtain a feature vector of each object, wherein the feature vector of each object is used for describing the features of each object. Since the same category includes at least one object, the generation device performs statistics on the feature vectors of the at least one object belonging to the same category, respectively, to obtain a feature vector of each of the plurality of categories.

In a possible implementation manner, the generating device may also obtain the detail information of at least one object of each category, and obtain at least one word group included in the detail information of each object, and for each word group, may use a preset conversion algorithm to perform conversion from a word to a word vector, so as to obtain a word vector corresponding to the word group, thereby obtaining at least one word vector corresponding to the at least one word group, and combine the at least one word vector to form the feature vector of the object. The preset conversion algorithm may be word2vec (word conversion vector) algorithm or other algorithms.

In another possible implementation manner, after the generating device obtains the feature vector of at least one object of any category, the generating device performs mean calculation on the feature vector of the at least one object, and uses a vector obtained after the mean calculation as the feature vector of the category. And respectively calculating each category according to the same calculation mode to obtain the feature vector of each category.

102. The generating means calculates the degree of similarity between any two categories based on the feature vectors of the plurality of categories.

In the embodiment of the present invention, after the generating device acquires the feature vectors of multiple categories, the generating device compares the feature vector of each category with the feature vectors of other categories except the category according to the feature vectors of multiple categories, so as to obtain a similarity between any two categories, and then determines the similar category of each category according to the obtained similarity.

The similarity between any two categories represents the similarity degree between any two categories, and the higher the similarity is, the more similar the two categories are, and the lower the similarity is, the more dissimilar the two categories are.

In one possible implementation manner, after the generating device obtains the feature vectors of the plurality of categories, the generating device performs point multiplication calculation on the feature vector of each category and the feature vectors of other categories to obtain the similarity between each category and the other categories, and further obtains the similarity between any two categories of the plurality of categories.

Besides the point-and-multiply calculation, the similarity between any two categories can be calculated by calculating Euclidean distance, calculating cosine similarity and the like.

103. For each category, the generating means determines a similar category of the category among the plurality of categories according to a similarity between the category and another category of the plurality of categories other than the category.

Wherein the similarity between any one category and the similar category is greater than the similarity between the non-similar categories.

To determine the degree of similarity between the various categories, for each category, the other categories may be divided into similar and non-similar categories for that category according to the degree of similarity between that category and the other categories. Wherein the similarity between the category and the similar category is greater than the similarity between the category and the dissimilar category.

In a possible implementation manner, the detail information of each object includes information of the object in multiple dimensions, and the feature vector of each category also includes feature values of the category in multiple dimensions, so that the feature values of each category in multiple dimensions can be comprehensively considered to determine a similar category of each category.

For this purpose, as shown in fig. 2, the generating device creates a multidimensional space according to the plurality of dimensions, determines the position of each category in the multidimensional space according to the feature value of the feature vector of each category in each dimension of the plurality of dimensions, and divides the multidimensional space into a plurality of regions according to the positions of the plurality of categories in the multidimensional space, wherein each region includes at least one category. For each category, taking the region to which the category belongs and the adjacent region of the region to which the category belongs as candidate regions of the category, and respectively obtaining the similarity between the category and other categories except the category in the candidate regions, thereby determining the similar category of the category in the candidate regions.

For example, the generating device may perform region division on the multidimensional space by using a preset division algorithm. The preset partitioning algorithm may be a KD tree (a data structure for partitioning a k-dimensional data space) algorithm or other algorithms. As shown in fig. 2 and fig. 3, the KD tree algorithm constructs a binary tree according to the multidimensional space, where a node in the binary tree represents an area in the multidimensional space, and each time any area in the multidimensional space is divided, a child node is generated for a node of the area corresponding to the binary tree, until the area division of the multidimensional space is completed, each leaf node in the binary tree may represent an area of a minimum unit, and a label is added to each leaf node, where the label may represent the number of categories in the corresponding area. By adopting the KD tree algorithm, the rapid division can be realized, the division speed is increased, and the speed of determining similar categories is increased.

In one possible implementation manner, for each category, according to the similarity between the category and other categories, the category with the highest similarity between the categories in the candidate area is taken as the similar category of the category.

In another possible implementation manner, for each category, the other categories are arranged in an order of descending similarity to the category according to the similarity between the category and the other categories, and a preset number of categories are selected from the other categories according to the arrangement order to serve as the similar categories.

In another possible implementation manner, the similarity between the category and other categories is obtained, and after the other categories are ranked from the highest similarity to the category, a similarity list of the category may be created, where the similarity list includes the other categories ranked in the ranking order. Then the first category may be selected as the similar category of the category according to the similarity list, or a preset number of categories may be selected as the similar categories of the category in sequence.

Of course, the generating means may also determine similar categories for each category in other ways.

104. The generating means creates a correlation network from the plurality of categories and the similar category for each category.

The generation device acquires the similar category of each category, associates each category with the similar category of the category, and creates a correlation network, where the correlation network includes the multiple categories, and each category is associated with the similar category of the category, and the correlation network may be as shown in fig. 4.

In a possible implementation manner, after the correlation network is obtained, an adjacency list may be created by using Protobuf (a data exchange format), the correlation network is stored as a graph network by using the structure of the adjacency list, and the correlation degree between each category can be visually seen according to the graph network.

Wherein at least one similar category of each category is stored in the adjacency list, and the at least one similar category is arranged in order. For example, as shown in fig. 5, the similar categories in the order of category a are category B and category D, the similar categories in the order of category B are category a, category C and category D, the similar category of category C is category B, and the similar categories in the order of category D are category a and category B.

105. The generation means determines a migration path that passes from the start point category to the end point category based on the correlation network.

The start point category and the end point category are any two categories of a plurality of categories, and the migration path at least includes the two categories of the start point category and the end point category, and may also include other categories besides the start point category and the end point category.

Optionally, after determining the start point category and the end point category, the generating device may calculate a migration path that the start point category passes through to reach the end point category by using a preset algorithm, where the preset algorithm may be Dijkstra (Dijkstra) algorithm or another algorithm, where Dijkstra is a single-source shortest path algorithm for calculating a shortest path from one node to all other nodes.

In a possible implementation manner, the generating device may search for the start point category by using a Dijkstra algorithm until the end point category is found, at which time the migration path search is finished, and acquire and store the migration path from the start point category to the end point category.

In another possible implementation manner, the generating device may search for the end point category by using a Dijkstra algorithm until the start point category is found, at which time the migration path search is finished, and acquire and store the migration path from the start point category to the end point category.

In another possible implementation manner, the generating device may perform search by using Dijkstra algorithm at the same time for the starting point category and the ending point category until the search performed from the starting point category and the ending point category meet at a certain intermediate category, at which time the search for the migration path is completed, the generating device obtains the migration path from the starting point category to the intermediate category and the migration path from the intermediate category to the ending point category, and saves the migration path formed by the two migration paths as the migration path from the starting point category to the ending point category.

If a plurality of migration paths which are passed from the starting point category to the end point category are determined in the correlation network, any one migration path is selected from the plurality of migration paths.

In a possible implementation manner, a migration path with the least number of categories is selected from the plurality of migration paths. Or selecting the transition path with the minimum transition distance from the plurality of transition paths. The migration distance between any two of the categories and the similarity between any two of the categories are in a negative correlation relationship, that is, the higher the similarity is, the shorter the migration distance is.

For example, when the category is a singer, the migration path that passes from the starting point category to the end point category is the migration path between any two singers in the plurality of singers, as shown in fig. 6, a search field for searching the starting point category and the end point category is displayed on the migration path display interface, and a "search" button is displayed, so that the user can input the starting point category and the end point category in the two search fields respectively, and then click the "search" button to display the migration path from the starting point category to the end point category. If the starting point category is singer A and the end point category is singer F, running Dijkstra algorithm for searching at the same time in the starting point category and the end point category to obtain a migration path of the singer A → the singer B → the singer C → the singer D → the singer E → the singer F, and the migration path is the migration path from the singer A to the singer F.

The embodiment of the invention can be applied to a scene of displaying the migration path for the user, the user can set the starting point category and the end point category on the generating device, the user can be shown the migration path from the starting point category to the end point category after the generating device generates the migration path, and the user can know the migration process between the starting point category and the end point category and other categories related to the migration process, so that a novel display mode is provided, and the interestingness is enhanced.

For example, when the object is a song and the category is a singer, the user sets two singers a and B with greatly different styles at will, so that a migration path between the singer a and the singer B can be generated, and the user can know other singers existing between the singer a and the singer B and know the style evolution process from the singer a to the singer B.

In addition, the embodiment of the invention can also be applied to a scene of recommending the object for the user, and after the migration path from the starting point category to the end point category is determined, a list of the object to be recommended can be established according to the migration path and recommended to the user. The recommendation list includes at least one object in other categories between the start point category and the end point category on the migration path in addition to the objects in the start point category and the end point category.

For example, when the object is a song and the category is a singer, according to a plurality of singers currently liked by the user, songs sung by the other singers on the migration paths of any two of the plurality of singers may be recommended for the user. Or, when the object is a movie and the category is director, according to a plurality of directors currently liked by the user, a movie produced by another director on a migration path of any two directors among the plurality of directors can be recommended to the user. Alternatively, when the object is a product and the category is a product type, a product included in another product type on the migration path of any two product types among the plurality of product types may be recommended to the user according to a product type frequently purchased by the user at present.

In addition, the embodiment of the invention can also be applied to a scene of carrying out reverse analysis on the recommended object, after the object is recommended to the user by adopting some recommendation algorithms, the method provided by the embodiment of the invention can be adopted to obtain the migration path from the category to which the recommended object belongs to a certain category liked by the user, and according to the migration distance of the migration path, whether the recommended object is reasonable or not can be analyzed, whether the preference of the user is met or not can be analyzed, the situation of wrong recommendation can be timely searched out, and therefore, the reverse analysis is realized.

For example, when the object is a song and the category is a singer, and when a certain song is recommended for the user, whether the recommended singer is reasonable or not can be judged according to a migration path between the singer currently liked by the user and the singer singing the recommended song.

According to the method provided by the embodiment of the invention, feature extraction is carried out on the detail information of at least one object of each category to obtain feature vectors of multiple categories, similar categories of each category are obtained according to the feature vectors, each category is associated with the similar categories to create a correlation network, the migration path of any two categories in the multiple categories can be obtained according to the correlation network, the migration probability between any two categories is not only calculated, the migration path can represent other categories passed by the migration between the two categories, the process of step-by-step migration between two or more categories is simulated, other categories used for transition are provided for the migration between any two non-similar categories, the information quantity is improved, and an information basis is provided for subsequent function expansion. In addition, labels do not need to be manually set, so that the waste of manual resources is avoided, the labor cost is reduced, the problem of low accuracy caused by strong subjectivity of the set labels is avoided, and the accuracy of the migration path is improved.

And in the process of obtaining the similar categories, a multi-dimensional space is created according to the multi-dimensional feature vector of each category, the multi-dimensional space is divided into a plurality of areas according to the position of each category in the multi-dimensional space to obtain the similar categories of each category, the feature values of each category in the plurality of dimensions are adopted in the process of determining the similar categories, the adopted information is more comprehensive, and the accuracy of the similar categories is improved. And moreover, a multi-dimensional space is created, similar categories are searched in the region to which the multi-dimensional space belongs and the adjacent region after region division is carried out in the multi-dimensional space, the calculation amount is reduced, and the calculation speed is accelerated.

Fig. 7 is a schematic structural diagram of a migration path generating apparatus according to an embodiment of the present invention, and referring to fig. 7, the apparatus includes:

a dividing module 701, configured to perform the step of dividing the objects with the same information in the specified dimension into the same category to obtain multiple categories;

a feature extraction module 702, configured to perform feature extraction on the detail information of the object in each category respectively in the foregoing embodiment, to obtain feature vectors of multiple categories;

a similar category determining module 703, configured to perform the step of determining a similar category of each category in the multiple categories according to the feature vectors of the multiple categories in the foregoing embodiment;

a creating module 704, configured to perform the step of creating the correlation network according to the multiple categories and the similar category of each category in the foregoing embodiment;

the migration path determining module 705 is configured to perform the step of determining a migration path that passes from the starting point category to the ending point category according to the correlation network in the foregoing embodiment.

Optionally, the feature extraction module 702 includes:

a feature extraction unit, configured to perform feature extraction on the detail information of each object in the foregoing embodiment, to obtain a feature vector of each object;

and the counting unit is used for performing the step of counting the characteristic vectors of at least one object of each category respectively to obtain the characteristic vectors of a plurality of categories.

Optionally, the similar category determining module 703 includes:

a calculating unit, configured to perform a step of calculating a similarity between any two categories according to feature vectors of a plurality of categories in the above embodiment;

and a similar category determining unit, configured to perform, for each category in the above-described embodiment, a step of determining, according to a similarity between the category and another category, except the category, of the plurality of categories, a similar category of the category among the plurality of categories.

Optionally, the detail information of the object includes information of the object in multiple dimensions, and the feature vector of the category includes feature values of the category in multiple dimensions;

the similar category determining unit is configured to perform the steps of creating a multidimensional space according to a plurality of dimensions and determining similar categories of the categories in the candidate area based on the multidimensional space in the above embodiment.

Optionally, the similar category determining unit is further configured to perform, according to the similarity between the category and each category in the candidate area, taking the category with the greatest similarity between the category and the candidate area as the similar category; or selecting a preset number of categories from the alternative area as similar categories according to the sequence of the similarity between the categories from large to small;

optionally, the migration path determining module 705 includes:

a migration path determining unit, configured to perform a step of determining, according to the correlation network in the foregoing embodiment, a plurality of migration paths that pass through from the start point category to the end point category;

the selecting unit is used for selecting the migration path with the least number of types from the plurality of migration paths in the embodiment; or selecting a migration path with the minimum migration distance from the multiple migration paths, wherein the migration distance between any two categories and the similarity between any two categories are in a negative correlation relationship.

It should be noted that: in the migration path generation apparatus provided in the foregoing embodiment, when generating a migration path, only the division of each functional module is illustrated, and in practical applications, the above function distribution may be completed by different functional modules as needed, that is, the internal structure of the generation apparatus is divided into different functional modules to complete all or part of the above described functions. In addition, the migration path generation apparatus and the migration path generation method provided in the foregoing embodiments belong to the same concept, and specific implementation processes thereof are described in detail in the method embodiments, and are not described herein again.

Fig. 8 is a schematic structural diagram of a terminal 800 according to an exemplary embodiment of the present invention. The terminal 800 may be a portable mobile terminal such as: smart phones, tablet computers, MP3 players (Moving Picture Experts Group Audio Layer III, moving Picture Experts compression standard Audio Layer 3), MP4 players (Moving Picture Experts Group Audio Layer IV, moving Picture Experts compression standard Audio Layer 4), laptops, desktop computers, head-mounted devices, or any other intelligent terminals. The terminal 800 may also be referred to as a user equipment, portable terminal, laptop terminal, desktop terminal, or the like, among other names.

In general, the terminal 800 includes: a processor 801 and a memory 802.

The processor 801 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and so forth. The processor 801 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 801 may also include a main processor and a coprocessor, where the main processor is a processor for Processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 801 may be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing content that the display screen needs to display. In some embodiments, the processor 801 may further include an AI (Artificial Intelligence) processor for processing computing operations related to machine learning.

Memory 802 may include one or more computer-readable storage media, which may be non-transitory. Memory 802 can also include high-speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In some embodiments, a non-transitory computer readable storage medium in the memory 802 is used to store at least one instruction for the processor 801 to have to implement the migration path generation method provided by the method embodiments of the present application.

In some embodiments, the terminal 800 may further include: a peripheral interface 803 and at least one peripheral. The processor 801, memory 802, and peripheral interface 803 may be connected by buses or signal lines. Various peripheral devices may be connected to the peripheral interface 803 by a bus, signal line, or circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 804, a touch screen display 805, a camera 806, an audio circuit 807, a positioning component 808, and a power supply 809.

The peripheral interface 803 may be used to connect at least one peripheral related to I/O (Input/Output) to the processor 801 and the memory 802. In some embodiments, the processor 801, memory 802, and peripheral interface 803 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 801, the memory 802, and the peripheral interface 803 may be implemented on separate chips or circuit boards, which is not limited by the present embodiment.

The Radio Frequency circuit 804 is used for receiving and transmitting RF (Radio Frequency) signals, also called electromagnetic signals. The radio frequency circuitry 804 communicates with a communication network and other communication devices via electromagnetic signals. The rf circuit 804 converts an electrical signal into an electromagnetic signal to be transmitted, or converts a received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 804 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so forth. The radio frequency circuit 804 may communicate with other terminals via at least one wireless communication protocol. The wireless communication protocols include, but are not limited to: metropolitan area networks, various generation mobile communication networks (2G, 3G, 4G, and 8G), wireless local area networks, and/or WiFi (Wireless Fidelity) networks. In some embodiments, the radio frequency circuit 804 may further include NFC (Near Field Communication) related circuits, which are not limited in this application.

The display screen 805 is used to display a UI (User Interface). The UI may include graphics, text, icons, video, and any combination thereof. When the display 805 is a touch display, the display 805 also has the ability to capture touch signals on or above the surface of the display 805. The touch signal may be input to the processor 801 as a control signal for processing. At this point, the display 805 may also be used to provide virtual buttons and/or a virtual keyboard, also referred to as soft buttons and/or a soft keyboard. In some embodiments, the display 805 may be one, providing a front panel of the terminal 800; in other embodiments, the display 805 may be at least two, respectively disposed on different surfaces of the terminal 800 or in a folded design; in still other embodiments, the display 805 may be a flexible display disposed on a curved surface or a folded surface of the terminal 800. Even further, the display 805 may be configured as a non-rectangular irregular figure, i.e., a shaped screen. The Display 805 can be made of LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), and other materials.

The camera assembly 806 is used to capture images or video. Optionally, camera assembly 806 includes a front camera and a rear camera. Generally, a front camera is disposed at a front panel of a terminal, and a rear camera is disposed at a rear surface of the terminal. In some embodiments, the number of the rear cameras is at least two, and each rear camera is any one of a main camera, a depth-of-field camera, a wide-angle camera and a telephoto camera, so that the main camera and the depth-of-field camera are fused to realize a background blurring function, the main camera and the wide-angle camera are fused to realize panoramic shooting and a VR (Virtual Reality) shooting function or other fusion shooting functions. In some embodiments, camera head assembly 806 may also include a flash. The flash lamp can be a single-color temperature flash lamp or a double-color temperature flash lamp. The double-color-temperature flash lamp is a combination of a warm-light flash lamp and a cold-light flash lamp and can be used for light compensation under different color temperatures.

The audio circuitry 807 may include a microphone and a speaker. The microphone is used for collecting sound waves of a user and the environment, converting the sound waves into electric signals, and inputting the electric signals to the processor 801 for processing or inputting the electric signals to the radio frequency circuit 804 to realize voice communication. The microphones may be provided in a plurality, respectively, at different portions of the terminal 800 for the purpose of stereo sound collection or noise reduction. The microphone may also be an array microphone or an omni-directional acquisition microphone. The speaker is used to convert electrical signals from the processor 801 or the radio frequency circuit 804 into sound waves. The loudspeaker can be a traditional film loudspeaker and can also be a piezoelectric ceramic loudspeaker. When the speaker is a piezoelectric ceramic speaker, the speaker can be used for purposes such as converting an electric signal into a sound wave audible to a human being, or converting an electric signal into a sound wave inaudible to a human being to measure a distance. In some embodiments, the audio circuitry 807 may also include a headphone jack.

The positioning component 808 is used to locate the current geographic Location of the terminal 800 for navigation or LBS (Location Based Service). The Positioning component 808 may be a Positioning component based on the GPS (Global Positioning System) in the united states, the beidou System in china, the graves System in russia, or the galileo System in the european union.

Power supply 809 is used to provide power to various components in terminal 800. The power supply 809 can be ac, dc, disposable or rechargeable. When the power source 809 comprises a rechargeable battery, the rechargeable battery may support wired or wireless charging. The rechargeable battery may also be used to support fast charge technology.

In some embodiments, terminal 800 also includes one or more sensors 810. The one or more sensors 810 include, but are not limited to: acceleration sensor 811, gyro sensor 812, pressure sensor 813, fingerprint sensor 814, optical sensor 815 and proximity sensor 816.

The acceleration sensor 811 may detect the magnitude of acceleration in three coordinate axes of the coordinate system established with the terminal 800. For example, the acceleration sensor 811 may be used to detect the components of the gravitational acceleration in three coordinate axes. The processor 801 may control the touch screen 805 to display the user interface in a landscape view or a portrait view according to the gravitational acceleration signal collected by the acceleration sensor 811. The acceleration sensor 811 may also be used for acquisition of motion data of a game or a user.

The gyro sensor 812 may detect a body direction and a rotation angle of the terminal 800, and the gyro sensor 812 may acquire a 3D motion of the user on the terminal 800 in cooperation with the acceleration sensor 811. From the data collected by the gyro sensor 812, the processor 801 may implement the following functions: motion sensing (such as changing the UI according to a user's tilting operation), image stabilization while shooting, game control, and inertial navigation.

Pressure sensors 813 may be disposed on the side bezel of terminal 800 and/or underneath touch display 805. When the pressure sensor 813 is disposed on the side frame of the terminal 800, the holding signal of the user to the terminal 800 can be detected, and the processor 801 performs left-right hand recognition or shortcut operation according to the holding signal collected by the pressure sensor 813. When the pressure sensor 813 is disposed at the lower layer of the touch display screen 805, the processor 801 controls the operability control on the UI interface according to the pressure operation of the user on the touch display screen 805. The operability control comprises at least one of a button control, a scroll bar control, an icon control, and a menu control.

The fingerprint sensor 814 is used for collecting a fingerprint of the user, and the processor 801 identifies the identity of the user according to the fingerprint collected by the fingerprint sensor 814, or the fingerprint sensor 814 identifies the identity of the user according to the collected fingerprint. Upon identifying the user as a trusted identity, the processor 801 authorizes the user to have relevant sensitive operations including unlocking the screen, viewing encrypted information, downloading software, paying for and changing settings, etc. Fingerprint sensor 814 may be disposed on the front, back, or side of terminal 800. When a physical button or a vendor Logo is provided on the terminal 800, the fingerprint sensor 814 may be integrated with the physical button or the vendor Logo.

The optical sensor 815 is used to collect the ambient light intensity. In one embodiment, the processor 801 may control the display brightness of the touch screen 805 based on the ambient light intensity collected by the optical sensor 815. Specifically, when the ambient light intensity is high, the display brightness of the touch display screen 805 is increased; when the ambient light intensity is low, the display brightness of the touch display 805 is turned down. In another embodiment, the processor 801 may also dynamically adjust the shooting parameters of the camera assembly 806 based on the ambient light intensity collected by the optical sensor 815.

A proximity sensor 816, also known as a distance sensor, is typically disposed on a front panel of the terminal 800. The proximity sensor 816 is used to collect a distance between the user and the front surface of the terminal 800. In one embodiment, when the proximity sensor 816 detects that the distance between the user and the front surface of the terminal 800 is gradually reduced, the processor 801 controls the touch display 805 to switch from a bright screen state to a dark screen state; when the proximity sensor 816 detects that the distance between the user and the front surface of the terminal 800 becomes gradually larger, the processor 801 controls the touch display 805 to switch from the screen-on state to the screen-on state.

Those skilled in the art will appreciate that the configuration shown in fig. 8 is not intended to be limiting of terminal 800 and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components may be used.

Fig. 9 is a schematic structural diagram of a server according to an embodiment of the present invention, where the server 900 may generate a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 901 and one or more memories 902, where the memory 902 stores at least one instruction, and the at least one instruction is loaded and executed by the processors 901 to implement the methods provided by the foregoing method embodiments. Certainly, the server may further have a wired or wireless network interface, a keyboard, an input/output interface, and other components to facilitate input and output, and the server may further include other components for implementing functions of the device, which are not described herein again.

The server 900 may be configured to execute the steps executed by the generation apparatus in the migration path generation method.

An embodiment of the present invention further provides a migration path generation apparatus, where the apparatus includes a processor and a memory, where the memory stores at least one instruction, at least one program, a code set, or an instruction set, and the instruction, the program, the code set, or the instruction set is loaded by the processor and has an operation in implementing the migration path generation method according to the foregoing embodiment.

An embodiment of the present invention further provides a computer-readable storage medium, where at least one instruction, at least one program, a code set, or a set of instructions is stored in the computer-readable storage medium, and the instruction, the program, the code set, or the set of instructions is loaded by a processor and has an operation in a migration path generation method for implementing the foregoing embodiment.

It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

The above description is only a preferred embodiment of the present invention, and should not be taken as limiting the invention, and any modifications, equivalents, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A migration path generation method, comprising:

and determining a migration path from a starting point category to an end point category according to the correlation network, wherein the starting point category and the end point category are any two categories of the plurality of categories.

2. The method according to claim 1, wherein said performing feature extraction on the detail information of the object in each category respectively to obtain feature vectors of the plurality of categories comprises:

extracting the features of the detail information of each object to obtain a feature vector of each object;

and respectively counting the characteristic vectors of at least one object of each category to obtain the characteristic vectors of the plurality of categories.

3. The method of claim 1, wherein the determining similar classes of the each class in the plurality of classes according to the feature vectors of the plurality of classes comprises:

calculating the similarity between any two categories according to the feature vectors of the categories;

for each of the categories, determining a similar category of the category in the plurality of categories according to a similarity between the category and other categories of the plurality of categories except the category.

4. The method of claim 3, wherein the detail information of the object comprises information of the object in a plurality of dimensions, and the feature vector of the category comprises feature values of the category in the plurality of dimensions;

the determining, according to the similarity between the category and other categories except the category in the multiple categories, a similar category of the category in the multiple categories includes:

creating a multi-dimensional space according to the plurality of dimensions, and determining the position of each category in the multi-dimensional space according to the characteristic value of the characteristic vector of each category in the plurality of dimensions;

dividing the multi-dimensional space into a plurality of regions according to the positions of the plurality of categories in the multi-dimensional space;

and for each category, taking the region to which the category belongs and the adjacent region of the region as candidate regions, and determining the similar category of the category in the candidate regions according to the similarity between the category and each category in the candidate regions.

5. The method of claim 4, wherein determining similar categories of the category in the candidate area according to the similarity between the category and each category in the candidate area comprises:

according to the similarity between the category and each category in the alternative area, taking the category with the maximum similarity between the category and the alternative area as the similar category; or,

and selecting a preset number of categories from the alternative area as the similar categories according to the sequence of the similarity between the categories from large to small.

6. The method of claim 1, wherein determining a migration path from a starting point category to an ending point category according to the correlation network comprises:

determining a plurality of migration paths from the starting point category to the end point category according to the correlation network;

selecting a migration path with the least number of categories from the plurality of migration paths; or,

and selecting the migration path with the minimum migration distance from the multiple migration paths, wherein the migration distance between any two categories and the similarity between any two categories are in a negative correlation relationship.

7. A migration path generation apparatus, characterized in that the apparatus comprises:

8. The apparatus of claim 7, wherein the feature extraction module comprises:

the characteristic extraction unit is used for extracting the characteristics of the detail information of each object to obtain a characteristic vector of each object;

and the counting unit is used for counting the characteristic vectors of at least one object of each category respectively to obtain the characteristic vectors of the plurality of categories.

9. The apparatus of claim 7, wherein the similar category determining module comprises:

the calculating unit is used for calculating the similarity between any two categories according to the feature vectors of the categories;

a similar category determining unit, configured to determine, for each of the categories, a similar category of the category in the multiple categories according to a similarity between the category and another category of the multiple categories except the category.

10. The apparatus according to claim 9, wherein the detail information of the object includes information of the object in a plurality of dimensions, and the feature vector of the category includes feature values of the category in the plurality of dimensions;

the similar category determining unit is further configured to create a multidimensional space according to the multiple dimensions, and determine a position of each category in the multidimensional space according to the feature value of the feature vector of each category in the multiple dimensions; dividing the multi-dimensional space into a plurality of regions according to the positions of the plurality of categories in the multi-dimensional space; and for each category, taking the region to which the category belongs and the adjacent region of the region as candidate regions, and determining the similar category of the category in the candidate regions according to the similarity between the category and each category in the candidate regions.

11. The apparatus according to claim 10, wherein the similar category determining unit is further configured to, according to a similarity between the category and each category in the candidate area, regard, as the similar category, a category in the candidate area with a highest similarity to the category; or selecting a preset number of categories from the alternative area as the similar categories according to the sequence of the similarity between the categories from large to small.

12. The apparatus of claim 7, wherein the migration path determining module comprises:

a migration path determination unit, configured to determine, according to the correlation network, a plurality of migration paths that pass through from a starting point category to an end point category;

a selecting unit, configured to select a migration path with the smallest number of categories from the multiple migration paths; or,

the selecting unit is configured to select a migration path with the smallest migration distance from the multiple migration paths, where the migration distance between any two categories and the similarity between any two categories are in a negative correlation relationship.

13. An apparatus for migration path generation, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by the processor to implement the operations performed in the migration path generation method according to any one of claims 1 to 6.

14. A computer-readable storage medium having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by a processor to perform the operations as recited in any one of claims 1 to 6.