CN110532528A - Books similarity calculating method and electronic equipment based on random walk - Google Patents
Books similarity calculating method and electronic equipment based on random walk Download PDFInfo
- Publication number
- CN110532528A CN110532528A CN201910798207.0A CN201910798207A CN110532528A CN 110532528 A CN110532528 A CN 110532528A CN 201910798207 A CN201910798207 A CN 201910798207A CN 110532528 A CN110532528 A CN 110532528A
- Authority
- CN
- China
- Prior art keywords
- books
- user
- similarity
- obtains
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of books similarity calculating method, electronic equipment and storage medium based on random walk, wherein the books similarity calculating method based on random walk includes: to obtain the user mutual behavior data for being directed to books;According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;Random walk calculating is carried out according to books associated diagram, obtains books similarity matrix of the every books relative to other books.The technical solution can be based on the user mutual behavior data for being directed to books, conveniently, books associated diagram is easily constructed, random walk calculating is carried out according to books associated diagram, quickly obtain books similarity matrix of the every books relative to other books, precisely, effectively from the similarity between user perspective reflection books, the accuracy in computation for effectively improving books similarity optimizes books similarity calculation mode.
Description
Technical field
The present invention relates to technical field of data processing, and in particular to a kind of books similarity calculation side based on random walk
Method, electronic equipment and storage medium.
Background technique
The books of e-book form the advantages such as facilitate due to having to obtain, and receive liking for a large number of users.Books are read
Platform is to carry out books recommendation according to the similarity of book contents mostly.Books similarity calculation mode is most in the prior art
To carry out the processing such as text identification, analysis to book contents, the similarity between books is obtained based on analysis result.On however,
Stating books similarity calculation mode is obtained based on content of text, can not be reflected from user perspective similar between books
Degree, accuracy is poor, and then leads to the use in the books for recommended when books recommendation using the similarity between books
Rate is lower, and recommendation effect is bad.
Summary of the invention
In view of the above problems, it proposes on the present invention overcomes the above problem or at least be partially solved in order to provide one kind
State books similarity calculating method, electronic equipment and the storage medium based on random walk of problem.
According to an aspect of the invention, there is provided a kind of books similarity calculating method based on random walk, comprising:
Obtain the user mutual behavior data for being directed to books;
According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
According to another aspect of the present invention, provide a kind of electronic equipment, comprising: processor, memory, communication interface and
Communication bus, processor, memory and communication interface complete mutual communication by communication bus;
Memory makes processor execute following operation for storing an at least executable instruction, executable instruction:
Obtain the user mutual behavior data for being directed to books;
According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
According to another aspect of the invention, a kind of storage medium is provided, it is executable that at least one is stored in storage medium
Instruction, executable instruction make processor execute following operation:
Obtain the user mutual behavior data for being directed to books;
According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
The technical solution provided according to the present invention, can be convenient, convenient based on the user mutual behavior data for being directed to books
Ground constructs books associated diagram, carries out random walk calculating according to books associated diagram, quickly obtains every books relative to other
The books similarity matrix of books, obtained books similarity matrix precisely, effectively can reflect books from user perspective
Between similarity, effectively improve the accuracy in computation of books similarity, optimize books similarity calculation mode.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And it can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can
It is clearer and more comprehensible, the followings are specific embodiments of the present invention.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows a kind of stream of according to embodiments of the present invention one books similarity calculating method based on random walk
Journey schematic diagram;
Fig. 2 shows a kind of streams of according to embodiments of the present invention two books similarity calculating method based on random walk
Journey schematic diagram;
Fig. 3 a shows a kind of books associated diagram schematic diagram;
Fig. 3 b shows another books associated diagram schematic diagram;
Fig. 4 shows the schematic diagram of the corresponding deep tree of books 1;
Fig. 5 shows the structural schematic diagram of according to embodiments of the present invention four a kind of electronic equipment.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
Embodiment one
Fig. 1 shows a kind of stream of according to embodiments of the present invention one books similarity calculating method based on random walk
Journey schematic diagram, as shown in Figure 1, this method comprises the following steps:
Step S101 obtains the user mutual behavior data for being directed to books.
User mutual behavior data for books are for describing the data interacted between user and books, specifically
Can include: user reads data, book review data, books downloading data etc. for the books of books.User mutual behavior number
According to the rule that imply data variation, it can be used for analyzing the incidence relation between books.In step s101, it can be read from books
The books that its storage is obtained in platform read the user mutual behaviors data such as data, book review data, books downloading data.
Step S102 determines the corresponding interactive books sequence of each user according to user mutual behavior data.
It would know that each user occurred to interact with which books by carrying out data analysis to user mutual behavior data
And books interaction sequences, the books interacted according to each user and books interaction sequences, determine that each user is corresponding
Interaction books sequence.Wherein, user corresponding interactive books sequence refers to interacted the user in books reading platform
Books obtained sequence after being arranged according to books interaction sequences, the corresponding interactive books sequence of user can be from user angle
Reflect the incidence relation between books on degree.
With the continuous growth of time, generated user mutual behavior data are more and more, in order to accurate and effective
Ground is reflected in the incidence relation between the books in the certain time stage, in step s 102 can be according in preset time window
User mutual behavior data determine the corresponding interactive books sequence of each user, and those skilled in the art can be according to actual needs
The window ranges of preset time window are configured, such as window ranges can be set in 7 days.
Step S103, according to the corresponding interactive books sequence of each user, construction obtains books associated diagram.
Specifically, can interactive books sequence corresponding to each user the processing such as count, analyze, two books must be taken office
Incidence relation and corresponding associated weights value between nationality, incidence relation may include that direct correlation relationship and indirect association are closed
System, associated weights value number of users corresponding with incidence relation are related;If having direct correlation relationship between wantonly two books,
Two books are attached, the side between this two books is formed, then according to associated weights value, determine it is each while while
Weighted value, so that construction obtains books associated diagram.It can intuitively, easily reflect the pass between each books by books associated diagram
Connection relationship.Wherein, books associated diagram can be digraph, or non-directed graph.If books associated diagram is digraph, then root
It also would know that user is directed to the interaction sequences of books according to books associated diagram.
Step S104 carries out random walk calculating according to books associated diagram, obtains every books relative to other books
Books similarity matrix.
In the present invention, for ease of description, for a certain books, the books in addition to the books are known as other books
Nationality.After having obtained books associated diagram, it is based on Random Walk Algorithm, carries out random walk calculating, meter according to books associated diagram
Calculation obtains similarity of the every books relative to other books, so that the books for obtaining every books relative to other books are similar
Spend matrix.Wherein, Random Walk Algorithm is the algorithm formed based on diffusive transport law, and the key concept of random walk refers to
The conserved quantity of any random walker institute band is all the ideal mathematics of Brownian movement each corresponding to a diffusive transport law
State.
Specifically, for every books in books associated diagram, from the books, by the side that is connected with the books to
There is one of incidence relation other books migration with the books, by the random walk of iteration, until migration is complete all and is somebody's turn to do
Books have other books of incidence relation, and random walk terminates;Then according to the books migration to each other books
The migration probability on the corresponding side in migration path calculates the similarity between the books and each other books, to be somebody's turn to do
Books similarity matrix of the books relative to other books.
It, can be based on for books using the books similarity calculating method provided in this embodiment based on random walk
User mutual behavior data, conveniently, easily construct books associated diagram, carry out random walk calculating according to books associated diagram, fastly
Books similarity matrix of the every books relative to other books is obtained promptly, and obtained books similarity matrix being capable of essence
It is quasi-, effectively from the similarity between user perspective reflection books, effectively improve the accuracy of books similarity, optimize
Books similarity calculation mode.
Embodiment two
Fig. 2 shows a kind of streams of according to embodiments of the present invention two books similarity calculating method based on random walk
Journey schematic diagram, as shown in Fig. 2, this method comprises the following steps:
Step S201 obtains the user mutual behavior data for being directed to books.
Wherein, the books that each user can be obtained from books reading platform for books read data, book review number
According to user mutual behaviors data such as, books downloading datas.User mutual behavior data include at least: User ID, user are interacted
Books, interaction initial time, interaction terminate the data such as time.Specifically, User ID can be user in books reading platform
Account, such as cell-phone number, user name, mailbox, WeChat ID, QQ number etc..
Step S202 carries out data analysis to the corresponding user mutual behavior data of the user, determines for each user
The books and books interaction sequences that the user interacted.
Specifically, data analysis can be carried out to the corresponding user mutual behavior data of the user in preset time window,
According to the time sequencing that user mutual behavior data generate, successively extracts from user mutual behavior data and occur with the user
Interactive books ID is crossed, so that it is determined that obtaining the books and books interaction sequences that the user interacted.
Step S203 is arranged according to the books that books interaction sequences interacted the user, and it is corresponding to obtain the user
Interaction books sequence.
It, can be according to books interaction sequences after determination obtains the books that the user interacted and books interaction sequences
The books interacted to the user arrange, and obtain the corresponding interactive books sequence of the user.It can according to interaction books sequence
Intuitively know successively interaction has occurred with which books in the user.
For example, the user for being uid1 for User ID, determines that obtaining the books that the user interacted includes books 1, books
3, books 10 and books 20, books interaction sequences are " 10 → books of books 1 → books, 3 → books 20 ", then User ID is
The corresponding interactive books sequence of the user of ID1 is represented by [books 1, books 3, books 10, books 20];For another example, for user
ID is the user of uid2, determines that obtaining the books that the user interacted includes books 1, books 2, books 21 and books 22, book
Nationality interaction sequences are " 21 → books of books 1 → books, 2 → books 22 ", then User ID is the corresponding interactive book of user of ID2
Nationality sequence is represented by [books 1, books 2, books 21, books 22].Wherein, interaction books sequence will be handed over [] according to books
The books that the user that mutually sequence obtains after arranging interacted include together, indicate there is incidence relation, user between these books
Successively interacted with these books.Those skilled in the art can indicate interactive books sequence with other forms, herein
Without limitation.
It is read in scene in practical books, user may occur to interact with many books, but be not every books
The books read all are consumed a longer time for user, for example, user can click during selecting books checks more
The brief introduction of books or several paragraphs, to judge whether book contents are its interested content.If in interaction books sequence
All books that user interacted are remained with, then not only making the data volume of interactive books sequence larger, but also can not be quasi-
Really, effectively reflect true incidence relation between books.
In view of the above problem, it can be directed to each user, data are carried out to the corresponding user mutual behavior data of the user
Analysis, determines the corresponding interactive duration of every books, wherein interaction duration can be according to the corresponding user mutual behavior number of the user
Terminate what the time was calculated for the interaction initial time of every books and interaction in.Every books are obtained in determination
After corresponding interactive duration, interactive duration can be screened out from the corresponding interactive books sequence of the user less than preset duration
Books.By preset duration be 10 minutes for, it is assumed that the corresponding interactive books sequence of the user that User ID is ID2 be [books 1,
Books 2, books 21, books 22], wherein the corresponding interactive duration of books 1, the corresponding interactive duration of books 2 and books 22 are right
The interaction duration answered is all larger than 10 minutes, and the only corresponding interactive duration of books 21 was less than 10 minutes, then can be from the interaction book
Books 21 are screened out in nationality sequence, after screening out processing, obtained interaction books sequence is [books 1, books 2, books 22].
The corresponding interactive books sequence of each user is split according to books interaction sequences, is obtained more by step S204
A books association pair.
It, can be according to books interaction sequences by the corresponding interactive books sequence of each user for the ease of constructing books associated diagram
It is split as the association pair of multiple books, each books association is to including wantonly two books and its incidence relation.For example, being directed to User ID
For the user of ID1, the corresponding interactive books sequence of the user is [books 1, books 3, books 10, books 20], then splitting
To books association to including: (books 1, books 3), (books 3, books 10) and (books 10, books 20), wherein books
Two books are included together available () by association, indicate there is incidence relation, books associated pair institute between this two books
The incidence relation for including is direct correlation relationship, which first interacts with the left side books of books associated pair, afterwards and book
The right side books of nationality associated pair interact.
Step S205 counts the association of each books to corresponding number of users, obtains each books association pair
Associated weights value.
After the corresponding interactive books sequence of each user all is completed to split, each books association is counted to corresponding use
Amount amount determines the associated weights of each books association pair according to the association of each books to corresponding number of users statistical result
Value.Specifically, for the association pair of each books, which can be associated with and the book is determined as to corresponding number of users statistical result
The associated weights value of nationality association pair.
Assuming that obtaining through statistics, there are the corresponding interactive books sequences of 500 users, and books association pair is obtained after fractionation
(books 1, books 3), there are the corresponding interactive books sequences of 300 users, and books association is obtained after fractionation to (books 3, book
Nationality 10), books are obtained after fractionation there are the corresponding interactive books sequence of 1200 users is associated with to (books 10, books 20),
So books association is 500 to (books 1, books 3) corresponding number of users, and books association is corresponding to (books 3, books 10)
Number of users is 300, and books association is 1200 to (books 10, books 20) corresponding number of users, then can be by books association pair
The associated weights value of (books 1, books 3) is determined as 500, and books association is determined as the associated weights value of (books 3, books 10)
300, books association is determined as 1200 to the associated weights value of (books 10, books 20).
Step S206 determines the side between each books according to the association pair of each books, and according to the association pair of each books
Associated weights value, determine the side right weight values on each side, construction obtains books associated diagram.
The books are associated with for the association pair of each books using each books as the node books in books associated diagram
Two books to included in are attached, and form the side between this two books;It is associated with according to all books to complete
After connecting between books two-by-two in pairs, the side between each books is obtained.Then according to the association power of each books association pair
Weight values determine the side right weight values on each side, to complete the construction to books associated diagram.Wherein, books associated diagram can be to have
To figure or non-directed graph,
If desired the books associated diagram constructed is digraph, then also needs the association for including according to each books associated pair
Relationship determines the direction on each side, specifically, the direction on each side is set as referring to from books on the left of each books associated pair
Books to the right;It, can be directly by the associated weights of each books association pair during determining the side right weight values on each side
Value is determined as the side right weight values on each side, or can also carry out calculation process to the associated weights value of each books association pair,
Such as divided by a fixed value, then operation result is determined as to the side right weight values on each side.Wherein, the books of digraph form
Associated diagram can be as shown in Figure 3a.
If desired the books associated diagram constructed is non-directed graph, in the books associated diagram each side be do not have it is directive, then
During determining the side right weight values on each side, need for books association pair, according to the associated weights of books association pair
Value and the side right weight for determining its corresponding sides to the summation of the associated weights value of corresponding books association pair with books association
Value.For being associated with books to (books 1, books 3), the association of corresponding books is to being (books 3, books 1), if books association
Associated weights value to (books 1, books 3) is 500, and books association is 200 to the associated weights value for (books 3, books 1),
So need to determine the side right weight values on the side between books 1 and books 3 according to the summation of the two associated weights values.For example, can
The summation of the two associated weights values is directly determined as to the side right weight values on the side, i.e. the side right weight values on the side are 700;Or
Calculation process can be carried out to the summation of the two associated weights values, such as divided by a fixed value 100, then by operation result
It is determined as the side right weight values on the side, i.e. the side right weight values on the side are 7.Wherein, the books associated diagram of undirected diagram form can be such as Fig. 3 b
It is shown.
After construction obtains books associated diagram, so that it may carry out random walk calculating according to books associated diagram, obtain every
Books similarity matrix of the books relative to other books can specifically be realized by step S207 to step S210.
Step S207 is searched to have with the books from books associated diagram and be closed for every books in books associated diagram
Other books of connection relationship.
Wherein, other books with the books with incidence relation include: to have the straight of direct correlation relationship with the books
It connects association books and there are the indirect association books of indirect association relationship specifically to have with the books and directly close with the books
The direct correlation books of connection relationship refer to the books being connected directly in books associated diagram with the books, have with the books indirect
The indirect association books of incidence relation refer in books associated diagram in addition to the books, be directly linked books be connected directly and
The books being indirectly connected.
For the books associated diagram shown in Fig. 3 a, for books 1 in the books associated diagram, search with books 1
Other books with incidence relation include: books 2, books 3, books 5, books 7, books 10, books 11, books 12, books
20, books 21, books 22, books 23 and books 27, wherein books 2, books 3 and books 5 are to have directly with books 1
The direct correlation books of incidence relation, books 7, books 10, books 11, books 12, books 20, books 21, books 22, books 23
It is the indirect association books that there is indirect association relationship with books 1 with books 27.
Step S208 constructs the corresponding deep tree of the books according to other books with the books with incidence relation.
Specifically, using the books as the root layer node books of deep tree, by with the books have incidence relation other
Each level of child nodes books of the books as deep tree;According to the incidence relation between the books and other books, root layer section is determined
The hierarchical relationship of point books and each level of child nodes books, construction obtain the corresponding deep tree of the books.
Step S209 calculates the book according to the side right weight values on the side in books associated diagram between the books and other books
The migration probability on the side in the corresponding deep tree of nationality between each node layer books.
If the corresponding deep tree of the books includes: root layer node books and n-layer child node books, n is the nature greater than 0
Number.For the 1st level of child nodes books of each of the deep tree, by the root layer node books recorded in books associated diagram and this
Between 1 level of child nodes books while side right weight values divided by between root layer node books and each 1st level of child nodes books while
Side right weight values summation, obtain the migration probability on the side between root layer node books and the 1st level of child nodes books.Then from
T=2 starts, for each of deep tree t level of child nodes books, the t straton section that will be recorded in books associated diagram
The side right weight values on the side between point books connected t-1 level of child nodes books and the t level of child nodes books are divided by the t-1
The summation of the side right weight values on the side between the connected each t level of child nodes books of level of child nodes books, obtains the t-1 straton
The migration probability on the side between node books and the t level of child nodes books;T is assigned a value of t+1, repeats this step, directly
Terminate to t=n+1.
For the books 1 in Fig. 3 a, the corresponding deep tree of books 1 that constructs can be as shown in figure 4, as indicated at 4, books
1 is the root layer node books of the deep tree, and books 2, books 3 and books 5 are the 1st level of child nodes books of the deep tree, books
10, books 11, books 12, books 21, books 23 and books 27 are the 2nd level of child nodes books of the deep tree, books 7, books
20 and books 22 be the deep tree the 3rd level of child nodes books, wherein the trip on the side in the deep tree between each node layer books
Walking probability can be indicated with r1 to r12.
Assuming that the side right weight values on the side between books 1 and books 2 are 300, the side right weight on the side between books 1 and books 3
Value is 500, and the side right weight values on the side between books 1 and books 5 are 1200, the side right weight values on the side between books 2 and books 21
It is 100, the side right weight values on the side between books 2 and books 23 are 200, then the migration probability r1 on the side between books 1 and books 2
It is 0.15 for 300/ (300+500+1200), i.e. r1, the migration probability r2 on the side between books 1 and books 3 is 500/ (300+
500+1200), i.e. r2 is 0.25, and the migration probability r3 on the side between books 1 and books 5 is 1200/ (300+500+1200), i.e.,
R3 is 0.60, and the migration probability r4 on the side between books 2 and books 21 is 100/ (100+200), i.e. r4 is 0.33,2 He of books
The migration probability r5 on the side between books 23 is 200/ (100+200), i.e. r5 is 0.67.
Step S210 carries out random walk calculating to the corresponding deep tree of the books, obtains the books relative to other books
The books similarity matrix of nationality.
If the corresponding deep tree of the books includes: root layer node books and n-layer child node books, with root layer node books
For starting point, all sides by being connected with root layer node books reach the 1st level of child nodes book to the 1st level of child nodes books migration
Nationality.Then since t=1, using each t level of child nodes books as starting point, pass through the institute being connected with the t level of child nodes books
There is Bian Xiang t+1 level of child nodes books migration, reaches t+1 level of child nodes books;T is assigned a value of t+1, repeats this step
Suddenly, until t=n terminates, all migration paths between root layer node books and each level of child nodes books are obtained.Then according to each
The migration probability on the corresponding side in a migration path, counts the similarity between root layer node books and each level of child nodes books
It calculates, obtains books similarity matrix of the books relative to other books.
It specifically, can be by the migration for the migration path between root layer node books and each level of child nodes books
The migration probability on the corresponding all sides in path carries out multiplying, and obtained calculated result as root layer node books and is somebody's turn to do
Similarity between level of child nodes books.After obtaining the similarity between root layer node books and each level of child nodes books,
Obtained similarity is summarized, to obtain books similarity matrix of the books relative to other books.The present invention
Obtained books similarity matrix precisely, effectively can reflect a books and other any books from user perspective
Between similarity.
By taking the corresponding deep tree of books 1 shown in Fig. 4 as an example, books similarity matrix of the books 1 relative to other books
It is represented by following form:
Books 2:r1, books 3:r2, books 5:r3,
Books 21:r1 × r4, books 23:r1 × r5, books 10:r2 × r6,
Books 11:r2 × r7, books 12:r2 × r8, books 27:r3 × r9,
Books 22:r1 × r4 × r10, books 20:r2 × r6 × r11, books 7:r2 × r8 × r12 }.
Step S211, books similarity matrix and user mutual behavior number according to every books relative to other books
According to user's recommended book.
Since obtained books similarity matrix precisely, effectively can reflect a books and other from user perspective
Similarity between any books can first be known according to the user mutual behavior data of the user then being directed to a certain user
Which this books the books that the user reads recently are, book of the books read recently then according to the user relative to other books
Nationality similarity matrix is ranked up each books according to the sequence of similarity from high to low.It is arranged in ranking results forward
Books are the more similar books of books read recently with the user, then when carrying out books recommendation, it can be to the user
Recommend to arrange m forward books in ranking results, wherein m is the natural number greater than 0, and the books recommended and the user are most
The books closely read similarity with higher can meet the reading hobby of the user, then the user will have very much well
Recommended books may be read or be downloaded, to effectively improve the use rate of recommended books, are significantly improved
Recommendation effect.
In addition, can also books similarity matrix according to every books relative to other books, classify to books, will
The higher books of similarity are divided in the same book category, when carrying out books lookup according to book category so as to user, energy
It is enough easily to find similar books.
Using the books similarity calculating method provided in this embodiment based on random walk, according to user mutual behavior number
According to, can accurately and rapidly determine books association to and its associated weights value, and then be used for books associated diagram construction, institute's structure
The books associated diagram made can intuitively, fully reflect the incidence relation between books from user perspective;It is closed based on books
Connection figure can easily construct the corresponding deep tree of every books, by carrying out random walk calculating to deep tree, rapidly
To books similarity matrix, obtained books similarity matrix can precisely, effectively reflect between books from user perspective
Similarity, effectively improve the accuracy in computation of books similarity, optimize books similarity calculation mode;In addition, root
According to books similarity matrix and user mutual behavior data, the books read recently with user can be recommended to have to user higher
The books of similarity effectively improve the use rate of recommended books, significantly improve recommendation effect.
Embodiment three
The embodiment of the present invention three provides a kind of non-volatile memory medium, and storage medium is stored at least one executable finger
It enables, which can be performed the books similarity calculating method based on random walk in above-mentioned any means embodiment.
Executable instruction specifically can be used for so that processor executes following operation: obtain user's interaction row for books
For data;According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;It is corresponding according to each user
Interaction books sequence, construction obtain books associated diagram;Random walk calculating is carried out according to books associated diagram, obtains every books phase
For the books similarity matrix of other books.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: for each
User carries out data analysis to the corresponding user mutual behavior data of the user, determines the books and book that the user interacted
Nationality interaction sequences;It is arranged according to the books that books interaction sequences interacted the user, obtains the corresponding interaction of the user
Books sequence.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: for each
User carries out data analysis to the corresponding user mutual behavior data of the user, determines the corresponding interactive duration of every books;From
The books that interactive duration is less than preset duration are screened out in the corresponding interactive books sequence of the user.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: according to books
Interaction sequences split the corresponding interactive books sequence of each user, obtain multiple books associations pair, books association pair
Include wantonly two books and its incidence relation;The association of each books counts corresponding number of users, obtains each book
The associated weights value of nationality association pair;According to the association pair of each books, the side between each books is determined, and close according to each books
The associated weights value of connection pair, determines the side right weight values on each side, and construction obtains books associated diagram;Books associated diagram is specially oriented
Figure or non-directed graph.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: for books
Every books in associated diagram search other books for having incidence relation with the books from books associated diagram;According to this
Books have other books of incidence relation, construct the corresponding deep tree of the books;According to the books in books associated diagram and its
The side right weight values on the side between his books, the migration for calculating the side in the corresponding deep tree of the books between each node layer books are general
Rate;Random walk calculating is carried out to the corresponding deep tree of the books, obtains books similarity of the books relative to other books
Matrix.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: by the books
As the root layer node books of deep tree, have other books of incidence relation as each straton section of deep tree for the books
Point books;According to the incidence relation between the books and other books, root layer node books and each level of child nodes books are determined
Hierarchical relationship, construction obtain the corresponding deep tree of the books.
In a kind of optional embodiment, if the corresponding deep tree of the books includes: root layer node books and n-layer sub- section
Point books, executable instruction further make processor execute following operation: with root layer node books starting point, by with root layer section
The connected all sides of point books reach the 1st level of child nodes books to the 1st level of child nodes books migration;Since t=1, with each
T level of child nodes books are starting point, by all sides for being connected with the t level of child nodes books to t+1 level of child nodes books
Migration reaches t+1 level of child nodes books;T is assigned a value of t+1, repeats this step, until t=n terminates, obtains root layer
All migration paths between node books and each level of child nodes books;Migration according to the corresponding side in each migration path is general
Rate calculates the similarity between root layer node books and each level of child nodes books, obtains the books relative to other books
The books similarity matrix of nationality.
In a kind of optional embodiment, executable instruction further makes processor execute following operation: according to every
Books similarity matrix and user mutual behavior data of the books relative to other books, to user's recommended book.
Example IV
Fig. 5 shows the structural schematic diagram of according to embodiments of the present invention four a kind of electronic equipment, present invention specific implementation
Example does not limit the specific implementation of electronic equipment.
As shown in figure 5, the electronic equipment may include: processor (processor) 502, communication interface
(Communications Interface) 504, memory (memory) 506 and communication bus 508.
Wherein:
Processor 502, communication interface 504 and memory 506 complete mutual communication by communication bus 508.
Communication interface 504, for being communicated with the network element of other equipment such as client or other servers etc..
Processor 502 can specifically execute based on the above-mentioned books similarity by random walk for executing program 510
Calculate the correlation step in embodiment of the method.
Specifically, program 510 may include program code, which includes computer operation instruction.
Processor 502 may be central processor CPU or specific integrated circuit ASIC (Application
Specific Integrated Circuit), or be arranged to implement the integrated electricity of one or more of the embodiment of the present invention
Road.The one or more processors that electronic equipment includes can be same type of processor, such as one or more CPU;It can also
To be different types of processor, such as one or more CPU and one or more ASIC.
Memory 506, for storing program 510.Memory 506 may include high speed RAM memory, it is also possible to further include
Nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.
Program 510 specifically can be used for so that processor 502 executes following operation: obtain user's interaction row for books
For data;According to user mutual behavior data, the corresponding interactive books sequence of each user is determined;It is corresponding according to each user
Interaction books sequence, construction obtain books associated diagram;Random walk calculating is carried out according to books associated diagram, obtains every books phase
For the books similarity matrix of other books.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: for every
A user carries out data analysis to the corresponding user mutual behavior data of the user, determine books that the user interacted and
Books interaction sequences;It is arranged according to the books that books interaction sequences interacted the user, obtains the corresponding friendship of the user
Mutual books sequence.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: for every
A user carries out data analysis to the corresponding user mutual behavior data of the user, determines the corresponding interactive duration of every books;
The books that interactive duration is less than preset duration are screened out from the corresponding interactive books sequence of the user.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: according to book
Nationality interaction sequences split the corresponding interactive books sequence of each user, obtain multiple books associations pair, books association
To including wantonly two books and its incidence relation;The association of each books counts corresponding number of users, obtains each
The associated weights value of books association pair;According to the association pair of each books, the side between each books is determined, and according to each books
The associated weights value of association pair, determines the side right weight values on each side, and construction obtains books associated diagram;Books associated diagram is specially to have
To figure or non-directed graph.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: being directed to book
Every books in nationality associated diagram search other books for having incidence relation with the books from books associated diagram;According to
The books have other books of incidence relation, construct the corresponding deep tree of the books;According to the books in books associated diagram with
The side right weight values on the side between other books calculate the migration on the side in the corresponding deep tree of the books between each node layer books
Probability;Random walk calculating is carried out to the corresponding deep tree of the books, the books for obtaining the books relative to other books are similar
Spend matrix.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: by the book
Root layer node books of the nationality as deep tree have other books of incidence relation as each straton of deep tree for the books
Node books;According to the incidence relation between the books and other books, root layer node books and each level of child nodes books are determined
Hierarchical relationship, construction obtain the corresponding deep tree of the books.
In a kind of optional embodiment, if the corresponding deep tree of the books includes: root layer node books and n-layer sub- section
Point books, program 510 is further such that processor 502 executes following operation: using root layer node books as starting point, by with root layer
The connected all sides of node books reach the 1st level of child nodes books to the 1st level of child nodes books migration;Since t=1, with every
A t level of child nodes books are starting point, by all sides for being connected with the t level of child nodes books to t+1 level of child nodes book
Nationality migration reaches t+1 level of child nodes books;T is assigned a value of t+1, repeats this step, until t=n terminates, obtains root
All migration paths between node layer books and each level of child nodes books;Migration according to the corresponding side in each migration path is general
Rate calculates the similarity between root layer node books and each level of child nodes books, obtains the books relative to other books
The books similarity matrix of nationality.
In a kind of optional embodiment, program 510 is further such that processor 502 executes following operation: according to every
Books similarity matrix and user mutual behavior data of this books relative to other books, to user's recommended book.
The specific implementation of each step may refer to the above-mentioned books similarity calculation implementation based on random walk in program 510
The corresponding description of corresponding steps in example, this will not be repeated here.It is apparent to those skilled in the art that for description
It is convenienct and succinct, the specific work process of the equipment of foregoing description can refer to corresponding processes in the foregoing method embodiment
Description, details are not described herein.
The scheme provided through this embodiment, can be based on the user mutual behavior data for being directed to books, conveniently, easily
Books associated diagram is constructed, random walk calculating is carried out according to books associated diagram, quickly obtains every books relative to other books
The books similarity matrix of nationality, obtained books similarity matrix can precisely, effectively from user perspective reflection books it
Between similarity, effectively improve the accuracy in computation of books similarity, optimize books similarity calculation mode.
Algorithm and display are not inherently related to any particular computer, virtual system, or other device provided herein.
Various general-purpose systems can also be used together with teachings based herein.As described above, it constructs required by this kind of system
Structure be obvious.In addition, the present invention is also not directed to any particular programming language.It should be understood that can use various
Programming language realizes summary of the invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the instructions provided here, numerous specific details are set forth.It is to be appreciated, however, that implementation of the invention
Example can be practiced without these specific details.In some instances, well known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this specification.
Similarly, it should be understood that in order to simplify the disclosure and help to understand one or more of the various inventive aspects, In
Above in the description of exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the disclosed method should not be interpreted as reflecting the following intention: i.e. required to protect
Shield the present invention claims features more more than feature expressly recited in each claim.More precisely, such as right
As claim reflects, inventive aspect is all features less than single embodiment disclosed above.Therefore, it then follows tool
Thus claims of body embodiment are expressly incorporated in the specific embodiment, wherein each claim conduct itself
Separate embodiments of the invention.
Those skilled in the art will understand that can be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more devices different from this embodiment.It can be the module or list in embodiment
Member or component are combined into a module or unit or component, and furthermore they can be divided into multiple submodule or subelement or
Sub-component.Other than such feature and/or at least some of process or unit exclude each other, it can use any
Combination is to all features disclosed in this specification (including adjoint claim, abstract and attached drawing) and so disclosed
All process or units of what method or apparatus are combined.Unless expressly stated otherwise, this specification is (including adjoint power
Benefit require, abstract and attached drawing) disclosed in each feature can carry out generation with an alternative feature that provides the same, equivalent, or similar purpose
It replaces.
In addition, it will be appreciated by those of skill in the art that although some embodiments described herein include other embodiments
In included certain features rather than other feature, but the combination of the feature of different embodiments mean it is of the invention
Within the scope of and form different embodiments.For example, in detail in the claims, embodiment claimed it is one of any
Can in any combination mode come using.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and ability
Field technique personnel can be designed alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between parentheses should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" located in front of the element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real
It is existing.The use of word first, second, and third does not indicate any sequence.These words can be construed to title.
The invention discloses: a kind of books similarity calculating method based on random walk of A1., comprising:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books phase of the every books relative to other books
Like degree matrix.
A2. method according to a1, it is described according to the user mutual behavior data, determine the corresponding friendship of each user
Mutual books sequence further comprises:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines that the user hands over
The books and books interaction sequences mutually crossed;
It is arranged according to the books that the books interaction sequences interacted the user, obtains the corresponding interaction of the user
Books sequence.
A3. the method according to A2, it is described according to the user mutual behavior data, determine the corresponding friendship of each user
Mutual books sequence further comprises:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines every books
Corresponding interactive duration;
The books that interactive duration is less than preset duration are screened out from the corresponding interactive books sequence of the user.
A4. method according to a1, described according to the corresponding interactive books sequence of each user, construction obtains books pass
Connection figure further comprises:
The corresponding interactive books sequence of each user is split according to books interaction sequences, obtains multiple books associations
Right, books association is to including wantonly two books and its incidence relation;
The association of each books counts corresponding number of users, obtains the associated weights of each books association pair
Value;
According to the association pair of each books, the side between each books, and the association power according to the association pair of each books are determined
Weight values, determine the side right weight values on each side, and construction obtains books associated diagram;The books associated diagram is specially digraph or nothing
Xiang Tu.
A5. method according to a1, it is described to carry out random walk calculating according to the books associated diagram, obtain every book
Nationality further comprises relative to the books similarity matrix of other books:
For every books in the books associated diagram, lookup has with the books and is associated with from the books associated diagram
Other books of relationship;
According to other books with the books with incidence relation, the corresponding deep tree of the books is constructed;
According to the side right weight values on the side in the books associated diagram between the books and other books, it is corresponding to calculate the books
Deep tree in side between each node layer books migration probability;
Random walk calculating is carried out to the corresponding deep tree of the books, obtains books phase of the books relative to other books
Like degree matrix.
A6. method according to a5, the basis have other books of incidence relation with the books, construct the books
Corresponding deep tree further comprises:
Using the books as the root layer node books of the deep tree, will there are other books of incidence relation with the books
Each level of child nodes books as the deep tree;
According to the incidence relation between the books and other books, the root layer node books and each straton section are determined
The hierarchical relationship of point books, construction obtain the corresponding deep tree of the books.
A7. method according to a5, if the corresponding deep tree of the books includes: root layer node books and n-layer child node
Books, then the corresponding deep tree of the described pair of books carries out random walk calculating, obtains book of the books relative to other books
Nationality similarity matrix further comprises:
Using root layer node books as starting point, by all sides for being connected with root layer node books to the 1st level of child nodes books
Migration reaches the 1st level of child nodes books;
Since t=1, using each t level of child nodes books as starting point, pass through what is be connected with the t level of child nodes books
All sides reach t+1 level of child nodes books to t+1 level of child nodes books migration;T is assigned a value of t+1, repeats this step
Suddenly, until t=n terminates, all migration paths between root layer node books and each level of child nodes books are obtained;
According to the migration probability on the corresponding side in each migration path, between root layer node books and each level of child nodes books
Similarity calculated, obtain books similarity matrix of the books relative to other books.
A8. according to the described in any item methods of A1-A7, random walk meter is carried out according to the books associated diagram described
It calculates, after obtaining every books relative to the books similarity matrix of other books, the method also includes:
Books similarity matrix and the user mutual behavior data according to every books relative to other books, Xiang Yong
Family recommended book.
The invention also discloses: B9. a kind of electronic equipment, comprising: processor, memory, communication interface and communication bus,
The processor, the memory and the communication interface complete mutual communication by the communication bus;
For the memory for storing an at least executable instruction, it is following that the executable instruction executes the processor
Operation:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books phase of the every books relative to other books
Like degree matrix.
B10. the electronic equipment according to B9, the executable instruction further make the processor execute following behaviour
Make:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines that the user hands over
The books and books interaction sequences mutually crossed;
It is arranged according to the books that the books interaction sequences interacted the user, obtains the corresponding interaction of the user
Books sequence.
B11. electronic equipment according to b10, the executable instruction further make the processor execute following behaviour
Make:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines every books
Corresponding interactive duration;
The books that interactive duration is less than preset duration are screened out from the corresponding interactive books sequence of the user.
B12. the electronic equipment according to B9, the executable instruction further make the processor execute following behaviour
Make:
The corresponding interactive books sequence of each user is split according to books interaction sequences, obtains multiple books associations
Right, books association is to including wantonly two books and its incidence relation;
The association of each books counts corresponding number of users, obtains the associated weights of each books association pair
Value;
According to the association pair of each books, the side between each books, and the association power according to the association pair of each books are determined
Weight values, determine the side right weight values on each side, and construction obtains books associated diagram;The books associated diagram is specially digraph or nothing
Xiang Tu.
B13. the electronic equipment according to B9, the executable instruction further make the processor execute following behaviour
Make:
For every books in the books associated diagram, lookup has with the books and is associated with from the books associated diagram
Other books of relationship;
According to other books with the books with incidence relation, the corresponding deep tree of the books is constructed;
According to the side right weight values on the side in the books associated diagram between the books and other books, it is corresponding to calculate the books
Deep tree in side between each node layer books migration probability;
Random walk calculating is carried out to the corresponding deep tree of the books, obtains books phase of the books relative to other books
Like degree matrix.
B14. electronic equipment according to b13, the executable instruction further make the processor execute following behaviour
Make:
Using the books as the root layer node books of the deep tree, will there are other books of incidence relation with the books
Each level of child nodes books as the deep tree;
According to the incidence relation between the books and other books, the root layer node books and each straton section are determined
The hierarchical relationship of point books, construction obtain the corresponding deep tree of the books.
B15. electronic equipment according to b13, if the corresponding deep tree of the books includes: root layer node books and n-layer
Child node books, the executable instruction further make the processor execute following operation:
Using root layer node books as starting point, by all sides for being connected with root layer node books to the 1st level of child nodes books
Migration reaches the 1st level of child nodes books;
Since t=1, using each t level of child nodes books as starting point, pass through what is be connected with the t level of child nodes books
All sides reach t+1 level of child nodes books to t+1 level of child nodes books migration;T is assigned a value of t+1, repeats this step
Suddenly, until t=n terminates, all migration paths between root layer node books and each level of child nodes books are obtained;
According to the migration probability on the corresponding side in each migration path, between root layer node books and each level of child nodes books
Similarity calculated, obtain books similarity matrix of the books relative to other books.
B16. according to the described in any item electronic equipments of B9-B15, the executable instruction further holds the processor
The following operation of row:
Books similarity matrix and the user mutual behavior data according to every books relative to other books, Xiang Yong
Family recommended book.
The invention also discloses a kind of storage medium of C17., an at least executable instruction is stored in the storage medium,
The executable instruction makes processor execute following operation:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books phase of the every books relative to other books
Like degree matrix.
C18. the storage medium according to C17, the executable instruction further make the processor execute following behaviour
Make:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines that the user hands over
The books and books interaction sequences mutually crossed;
It is arranged according to the books that the books interaction sequences interacted the user, obtains the corresponding interaction of the user
Books sequence.
C19. the storage medium according to C18, the executable instruction further make the processor execute following behaviour
Make:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines every books
Corresponding interactive duration;
The books that interactive duration is less than preset duration are screened out from the corresponding interactive books sequence of the user.
C20. the storage medium according to C17, the executable instruction further make the processor execute following behaviour
Make:
The corresponding interactive books sequence of each user is split according to books interaction sequences, obtains multiple books associations
Right, books association is to including wantonly two books and its incidence relation;
The association of each books counts corresponding number of users, obtains the associated weights of each books association pair
Value;
According to the association pair of each books, the side between each books, and the association power according to the association pair of each books are determined
Weight values, determine the side right weight values on each side, and construction obtains books associated diagram;The books associated diagram is specially digraph or nothing
Xiang Tu.
C21. the storage medium according to C17, the executable instruction further make the processor execute following behaviour
Make:
For every books in the books associated diagram, lookup has with the books and is associated with from the books associated diagram
Other books of relationship;
According to other books with the books with incidence relation, the corresponding deep tree of the books is constructed;
According to the side right weight values on the side in the books associated diagram between the books and other books, it is corresponding to calculate the books
Deep tree in side between each node layer books migration probability;
Random walk calculating is carried out to the corresponding deep tree of the books, obtains books phase of the books relative to other books
Like degree matrix.
C22. the storage medium according to C21, the executable instruction further make the processor execute following behaviour
Make:
Using the books as the root layer node books of the deep tree, will there are other books of incidence relation with the books
Each level of child nodes books as the deep tree;
According to the incidence relation between the books and other books, the root layer node books and each straton section are determined
The hierarchical relationship of point books, construction obtain the corresponding deep tree of the books.
C23. the storage medium according to C21, if the corresponding deep tree of the books includes: root layer node books and n-layer
Child node books, the executable instruction further make the processor execute following operation:
Using root layer node books as starting point, by all sides for being connected with root layer node books to the 1st level of child nodes books
Migration reaches the 1st level of child nodes books;
Since t=1, using each t level of child nodes books as starting point, pass through what is be connected with the t level of child nodes books
All sides reach t+1 level of child nodes books to t+1 level of child nodes books migration;T is assigned a value of t+1, repeats this step
Suddenly, until t=n terminates, all migration paths between root layer node books and each level of child nodes books are obtained;
According to the migration probability on the corresponding side in each migration path, between root layer node books and each level of child nodes books
Similarity calculated, obtain books similarity matrix of the books relative to other books.
C24. according to the described in any item storage mediums of C17-C23, the executable instruction further makes the processor
Execute following operation:
Books similarity matrix and the user mutual behavior data according to every books relative to other books, Xiang Yong
Family recommended book.
Claims (10)
1. a kind of books similarity calculating method based on random walk, comprising:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
2. determining that each user is corresponding according to the method described in claim 1, described according to the user mutual behavior data
Interaction books sequence further comprises:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines that the user interacted
Books and books interaction sequences;
It is arranged according to the books that the books interaction sequences interacted the user, obtains the corresponding interactive books of the user
Sequence.
3. determining that each user is corresponding according to the method described in claim 2, described according to the user mutual behavior data
Interaction books sequence further comprises:
For each user, data analysis is carried out to the corresponding user mutual behavior data of the user, determines that every books are corresponding
Interaction duration;
The books that interactive duration is less than preset duration are screened out from the corresponding interactive books sequence of the user.
4. construction obtains books according to the method described in claim 1, described according to each user corresponding interactive books sequence
Associated diagram further comprises:
The corresponding interactive books sequence of each user is split according to books interaction sequences, obtains multiple books associations pair,
Books association is to including wantonly two books and its incidence relation;
The association of each books counts corresponding number of users, obtains the associated weights value of each books association pair;
According to the association pair of each books, the side between each books, and the associated weights value according to the association pair of each books are determined,
Determine the side right weight values on each side, construction obtains books associated diagram;The books associated diagram is specially digraph or non-directed graph.
5. obtaining every according to the method described in claim 1, described carry out random walk calculating according to the books associated diagram
Books further comprise relative to the books similarity matrix of other books:
For every books in the books associated diagram, searching from the books associated diagram has incidence relation with the books
Other books;
According to other books with the books with incidence relation, the corresponding deep tree of the books is constructed;
According to the side right weight values on the side in the books associated diagram between the books and other books, the corresponding depth of the books is calculated
The migration probability on the side in degree tree between each node layer books;
Random walk calculating is carried out to the corresponding deep tree of the books, obtains books similarity of the books relative to other books
Matrix.
6. constructing the book according to the method described in claim 5, the basis has other books of incidence relation with the books
The corresponding deep tree of nationality further comprises:
Using the books as the root layer node books of the deep tree, using with the books have other books of incidence relation as
Each level of child nodes books of the deep tree;
According to the incidence relation between the books and other books, the root layer node books and each level of child nodes book are determined
The hierarchical relationship of nationality, construction obtain the corresponding deep tree of the books.
7. according to the method described in claim 5, if the corresponding deep tree of the books includes: root layer node books and n-layer sub- section
Point books, then the corresponding deep tree of the described pair of books carries out random walk calculating, obtains the books relative to other books
Books similarity matrix further comprises:
Using root layer node books as starting point, by all sides for being connected with root layer node books to the 1st level of child nodes books migration,
Reach the 1st level of child nodes books;
It is all by what is be connected with the t level of child nodes books using each t level of child nodes books as starting point since t=1
Bian Xiang t+1 level of child nodes books migration, reaches t+1 level of child nodes books;T is assigned a value of t+1, repeats this step,
Until t=n terminates, all migration paths between root layer node books and each level of child nodes books are obtained;
According to the migration probability on the corresponding side in each migration path, to the phase between root layer node books and each level of child nodes books
It is calculated like degree, obtains books similarity matrix of the books relative to other books.
8. method according to claim 1-7 carries out random walk meter according to the books associated diagram described
It calculates, after obtaining every books relative to the books similarity matrix of other books, the method also includes:
Books similarity matrix and the user mutual behavior data according to every books relative to other books, push away to user
Recommend books.
9. a kind of electronic equipment, comprising: processor, memory, communication interface and communication bus, the processor, the storage
Device and the communication interface complete mutual communication by the communication bus;
The memory makes the processor execute following behaviour for storing an at least executable instruction, the executable instruction
Make:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
10. a kind of storage medium, it is stored with an at least executable instruction in the storage medium, the executable instruction makes to handle
Device executes following operation:
Obtain the user mutual behavior data for being directed to books;
According to the user mutual behavior data, the corresponding interactive books sequence of each user is determined;
According to the corresponding interactive books sequence of each user, construction obtains books associated diagram;
Random walk calculating is carried out according to the books associated diagram, obtains books similarity of the every books relative to other books
Matrix.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910798207.0A CN110532528B (en) | 2019-08-27 | 2019-08-27 | Book similarity calculation method based on random walk and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910798207.0A CN110532528B (en) | 2019-08-27 | 2019-08-27 | Book similarity calculation method based on random walk and electronic equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110532528A true CN110532528A (en) | 2019-12-03 |
CN110532528B CN110532528B (en) | 2023-05-26 |
Family
ID=68664554
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910798207.0A Active CN110532528B (en) | 2019-08-27 | 2019-08-27 | Book similarity calculation method based on random walk and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110532528B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111125543A (en) * | 2019-12-17 | 2020-05-08 | 掌阅科技股份有限公司 | Training method of book recommendation sequencing model, computing device and storage medium |
CN113420056A (en) * | 2021-05-14 | 2021-09-21 | 北京达佳互联信息技术有限公司 | Behavior data processing method and device, electronic equipment and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110320276A1 (en) * | 2010-06-28 | 2011-12-29 | International Business Machines Corporation | System and method for online media recommendations based on usage analysis |
JP2013210787A (en) * | 2012-03-30 | 2013-10-10 | Dainippon Printing Co Ltd | Book recommendation system, book recommendation method, book recommendation server, and program |
CN109214692A (en) * | 2018-09-14 | 2019-01-15 | 掌阅科技股份有限公司 | E-book methods of marking and electronic equipment based on user's timing behavior |
CN109739411A (en) * | 2018-12-24 | 2019-05-10 | 咪咕数字传媒有限公司 | A kind of information processing method and device, equipment and storage medium |
CN109902203A (en) * | 2019-01-25 | 2019-06-18 | 北京邮电大学 | The network representation learning method and device of random walk based on side |
CN109978657A (en) * | 2019-03-07 | 2019-07-05 | 北京工业大学 | A kind of improvement random walk chart-pattern proposed algorithm towards many intelligence platforms |
-
2019
- 2019-08-27 CN CN201910798207.0A patent/CN110532528B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110320276A1 (en) * | 2010-06-28 | 2011-12-29 | International Business Machines Corporation | System and method for online media recommendations based on usage analysis |
JP2013210787A (en) * | 2012-03-30 | 2013-10-10 | Dainippon Printing Co Ltd | Book recommendation system, book recommendation method, book recommendation server, and program |
CN109214692A (en) * | 2018-09-14 | 2019-01-15 | 掌阅科技股份有限公司 | E-book methods of marking and electronic equipment based on user's timing behavior |
CN109739411A (en) * | 2018-12-24 | 2019-05-10 | 咪咕数字传媒有限公司 | A kind of information processing method and device, equipment and storage medium |
CN109902203A (en) * | 2019-01-25 | 2019-06-18 | 北京邮电大学 | The network representation learning method and device of random walk based on side |
CN109978657A (en) * | 2019-03-07 | 2019-07-05 | 北京工业大学 | A kind of improvement random walk chart-pattern proposed algorithm towards many intelligence platforms |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111125543A (en) * | 2019-12-17 | 2020-05-08 | 掌阅科技股份有限公司 | Training method of book recommendation sequencing model, computing device and storage medium |
CN111125543B (en) * | 2019-12-17 | 2023-03-10 | 掌阅科技股份有限公司 | Training method of book recommendation sequencing model, computing device and storage medium |
CN113420056A (en) * | 2021-05-14 | 2021-09-21 | 北京达佳互联信息技术有限公司 | Behavior data processing method and device, electronic equipment and storage medium |
CN113420056B (en) * | 2021-05-14 | 2023-12-26 | 北京达佳互联信息技术有限公司 | Behavior data processing method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN110532528B (en) | 2023-05-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Fournier-Viger et al. | A survey of high utility itemset mining | |
CN106023015B (en) | Course learning path recommendation method and device | |
Ben-Ari et al. | Modeling data from computer experiments: an empirical comparison of kriging with MARS and projection pursuit regression | |
CN106251174A (en) | Information recommendation method and device | |
CN106250403A (en) | Customer loss Forecasting Methodology and device | |
CN109299994A (en) | Recommended method, device, equipment and readable storage medium storing program for executing | |
Bergamini et al. | Approximating betweenness centrality in fully dynamic networks | |
Arrigo et al. | Non-backtracking walk centrality for directed networks | |
CN110442788A (en) | A kind of information recommendation method and device | |
CN107861981A (en) | A kind of data processing method and device | |
CN110532528A (en) | Books similarity calculating method and electronic equipment based on random walk | |
US10839420B2 (en) | Constrained large-data markdown optimizations based upon markdown budget | |
CN109214692B (en) | E-book methods of marking and electronic equipment based on user's timing behavior | |
CN109902852A (en) | Grouping of commodities method, apparatus, electronic equipment and readable storage medium storing program for executing | |
CN110910201B (en) | Information recommendation control method and device, computer equipment and storage medium | |
CN108052670A (en) | A kind of recommendation method and device of camera special effect | |
CN106445954A (en) | Business object display method and apparatus | |
Barbierato et al. | A Performance Modeling Language For Big Data Architectures. | |
CN109685545A (en) | Virtual network resource predictor method, device and electronic equipment to be provided | |
US20180189403A1 (en) | Website domain specific search | |
Bevan | Measuring dust in core-collapse supernovae with a Bayesian approach to line profile modelling | |
CN106940639A (en) | Method, system and computer-readable medium for can configure code fingerprint | |
US20230077708A1 (en) | Microservice measurement and merging | |
Mota et al. | A compact timed state space approach for the analysis of manufacturing systems: key algorithmic improvements | |
CN109543107A (en) | Data processing method, medium, device and calculating equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |