CN105095258B - A kind of media information sort method, device and media information recommending system - Google Patents
A kind of media information sort method, device and media information recommending system Download PDFInfo
- Publication number
- CN105095258B CN105095258B CN201410192707.7A CN201410192707A CN105095258B CN 105095258 B CN105095258 B CN 105095258B CN 201410192707 A CN201410192707 A CN 201410192707A CN 105095258 B CN105095258 B CN 105095258B
- Authority
- CN
- China
- Prior art keywords
- media information
- data
- value
- user
- media
- 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.)
- Active
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Embodiment of the present invention proposes a kind of media information sort method, device and media information recommending systems.Method includes: to receive based on positive weights trigger data caused by triggering positive weights component, and the useful evidence of media information is determined from the positive weights trigger data；It receives to reply based on triggering and replys trigger data caused by component, the reply data of the media information are determined from the reply trigger data；Determine the attribute data of the media information；The ranking value of media information is determined based on the useful evidence, reply data and media information attribute data of media information；Media information is ranked up based on ranking value.Embodiment of the present invention improves media information sequence accuracy rate, effectively prevent practising fraud, and can reduce Matthew effect.
Description
Technical field
Embodiment of the present invention is related to technical field of internet application, more particularly, to a kind of media information sequence side
Method, device and media information recommending system.
Background technique
In the current information age, various information equipments come into being: having the fixed-line telephone for Tone Via, movement
Phone；There are the server and PC for information resources share, processing；There are the various television sets shown for video data
Etc..These equipment are all to solve actual demand in specific area and generate.With E-consumer, computer, lead to
Believe the arrival of (3C) fusion, attention has been put into integrates to the information equipment of each different field more and more by people
In the research utilized, preferably serviced with making full use of existing resource equipment for people.
Social network refers to person-to-person relational network, and this website based on social networking relationships system thinking is just
It is social network website, refers exclusively to the Internet application service for being intended to help people to establish social network, also refers to that society is existing
Mature universal information carrier.Commentary Systems are a simple social intercourse systems.After having browsed theme, user is often
Provide various comments.Concerned higher comment can preferentially be shown, therefore relate to ask what comment was ranked up
Topic.
In the prior art, Commentary Systems are mostly ranked up comment according to the quantity on top, such sortord mistake
It in dullness, and is easy to be influenced by cheating, therefore accuracy of sorting is not high.Moreover, this sortord do not account for it is cold
But the addition of law, therefore also result in serious Matthew effect.
Summary of the invention
Embodiment of the present invention proposes a kind of media information sort methods, to improve sequence accuracy rate.
Embodiment of the present invention also proposed a kind of media information collator, to improve sequence accuracy rate.
Embodiment of the present invention also proposed a kind of media information recommending system, to improve information recommendation quality.
The concrete scheme of embodiment of the present invention is as follows:
A kind of media information sort method, this method comprises:
It receives based on positive weights trigger data caused by triggering positive weights component, from the positive weights trigger data really
Determine the useful evidence of media information；It receives to reply based on triggering and replys trigger data caused by component, trigger number from the reply
According to the reply data of the middle determination media information；Determine the attribute data of the media information；
The reply data of useful evidence, the media information based on the media information and the attribute number of the media information
According to the ranking value of the determination media information；
Media information is ranked up based on the ranking value.
A kind of media information collator, comprising:
Data determination unit, for receiving based on positive weights trigger data caused by triggering positive weights component, from described
The useful evidence of media information is determined in positive weights trigger data；It receives to reply based on triggering and replys triggering number caused by component
According to the reply data of the determining media information from the reply trigger data；Determine the attribute data of the media information；
Sorting values determination unit, for useful evidence, the media information based on the media information reply data and
The attribute data of the media information determines the ranking value of the media information；
Sequencing unit, for being ranked up based on the ranking value to media information.
A kind of media information recommending system, comprising:
Issue client terminal, for sending media information to server；
Server is received for issuing the media information based on positive weights triggering caused by triggering positive weights component
Data determine the useful evidence of media information from the positive weights trigger data；It receives and is replied caused by component based on triggering
Trigger data is replied, the reply data of the media information are determined from the reply trigger data；Determine the media information
Attribute data；The reply data of useful evidence, the media information based on the media information and the category of the media information
Property data determine the ranking value of the media information；Media information is ranked up based on the ranking value.
It can be seen from the above technical proposal that in embodiments of the present invention, receiving and being produced based on triggering positive weights component
Raw positive weights trigger data determines the useful evidence of media information from the positive weights trigger data；It receives and is based on triggering back
Trigger data is replied caused by multiple component, the reply data of the media information are determined from the reply trigger data；Really
The attribute data of the fixed media information；The reply data of useful evidence, the media information based on the media information and institute
The attribute data for stating media information determines the ranking value of the media information；Media information is arranged based on the ranking value
Sequence, it can be seen that, embodiment of the present invention is not ranked up the media informations such as comment according still further to the quantity on top, but base
The ranking value of media information is determined in the useful evidence of media information, reply data and media information attribute data synthesis, so as to
To effectively prevent practising fraud, sequence accuracy is improved.Moreover, can periodically determine superseded push away after using embodiment of the present invention
Media information is recommended, so that the influence of Matthew effect be effectively reduced.
In addition, after using embodiment of the present invention, not further according to individual subscribers such as the log duration of user, dispatch frequencies
Behavior determines the credit rating of user.On the contrary, being served as a fill-in according to the history at very useful family, very useful family and the credit rating at very useful family determines hair
The credit rating of the user of media information out has fully considered the behavior of other users, it is thus determined that user credit degree accuracy rate
It is high.In addition, embodiment of the present invention avoid according to be easy to happen in individual subscriber behavior decision user credit degree cheating lack
It falls into, to further improve user credit degree accuracy rate.
Detailed description of the invention
Fig. 1 is according to embodiment of the present invention media information sort method flow chart；
Fig. 2 is according to embodiment of the present invention top relation chain schematic diagram；
Fig. 3 is according to embodiment of the present invention media information collator schematic diagram；
Fig. 4 is according to embodiment of the present invention media information recommending system first structure figure；
Fig. 5 is according to embodiment of the present invention media information recommending the second structure chart of system.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, the present invention is made with reference to the accompanying drawing further
Detailed description.
It is succinct and intuitive in order to what is described, hereafter by describing several representative embodiments come to side of the invention
Case is illustrated.A large amount of details is only used for helping to understand the solution of the present invention in embodiment.However, it will be apparent that of the invention
Technical solution can be not limited to these details when realizing.In order to avoid unnecessarily having obscured the solution of the present invention, Yi Xieshi
It applies mode not described meticulously, but only gives frame.Hereinafter, " comprising " refers to " including but not limited to ", " root
According to ... " refer to " according at least to ..., but be not limited to according only to ... ".Due to the speech habits of Chinese, hereinafter without spy
When not pointing out the quantity of an ingredient, it is meant that the ingredient is either one or more, or can be regarded as at least one.
Fig. 1 is the media information sort method flow chart according to embodiment of the present invention.
As shown in Figure 1, this method comprises:
Step 101: receiving based on positive weights trigger data caused by triggering positive weights component, triggered from the positive weights
The useful evidence of media information is determined in data；It receives to reply based on triggering and replys trigger data caused by component, from described time
The reply data of the media information are determined in multiple trigger data；Determine the attribute data of the media information；
Step 102: media information is determined based on the useful evidence, reply data and media information attribute data of media information
Ranking value；
Step 103: media information being ranked up based on ranking value.
There is positive weights component in user client.Positive weights component may include various types of positive weights icons
(such as the button pushed up).User triggers positive weights icon when browsing release information, that is, can produce trigger data.Trigger data is used for
It expresses and the front of release information is praised, these release informations include media information.
Server receives user and triggers positive weights trigger data caused by positive weights component, and from positive weights trigger data
In determine user trigger positive weights icon caused by be directed to media information useful evidence.
Also have in user client and replys component.Reply component may include various types of reply icons (such as
The button of reply frame is popped up after triggering).Component is replied in triggering when user browses release information, that is, can produce reply frame, Yong Huke
To issue the reply trigger data for release information in reply frame.Trigger data is replied to comment release information for expressing
By these release informations include media information.
Server receives user and triggers reply trigger data caused by reply component, and determines from replying in trigger data
User triggers the reply data replied and be directed to media information caused by icon out.
In one embodiment, this method further comprises:
Time attenuation operations are executed to the ranking value of media information, the ranking value after acquisition time decaying；
Media information is ranked up are as follows:
Media information is ranked up based on the ranking value after time decaying.
In one embodiment, the useful evidence of media information includes pushing up the credit rating Pi of the user of media information, is replied
Data include replying the number of users P of the media information, and media information attribute data is the length value C of the media information；To media
It includes: with time attenuation function (t that the ranking value of information, which executes time attenuation operations,comm+2)gThe ranking value of decaying media information；
T (i) is the ranking value after time decaying, in which:
Work as tiWithBetween time difference be greater than predetermined value when, tcommFor the predetermined value；Work as tiWithBetween time difference not
When greater than the predetermined value, tcommFor tiWithDifference, wherein tiFor the issuing time of the media information；It is right for the media information
The issuing time average value of all media informations under the theme answered；N is the number of users for pushing up media information；G is pre-set
Parameter.For example, the predetermined value can be 24.
In the application, (tcomm+2)gIt is a time attenuation function, is similar to Newton's law of cooling；G is one variable
Parameter, this parameter can be changed according to different applications.
In one embodiment, this method further comprises:
According to ranking results selection predetermined number media information, using as recommend media information；
It shows and recommends media information.
In fact, ranking value calculation formula can bring Matthew effect.Matthew effect refers to that powerhouse is stronger, weak person is weaker and shows
As.For the sake of justice, in the ranking value of various Commentary Systems calculates, it should avoid Matthew effect.Be added the time decaying because
Element can reduce Matthew effect, but can not also avoid Matthew effect completely.
In order to better comment is put into the position of recommendation, embodiment of the present invention is introduced with a kind of record in real time and is eliminated
Mechanism the comment of recommendation is analyzed, eliminate the worse comment of effect.Eliminative mechanism is based on to draw a conclusion:
In predetermined number (such as 10) the comment position of recommendation, ballot quantity obtained by commenting on and position are unrelated.Based on this premise,
It can recorde in lower each same time interval, then polled data caused by all recommendation positions is analyzed according to data
It eliminates and feeds back bad comment
In one embodiment, this method further comprises:
It determines at preset time intervals, recommends the useful evidence of media information；
Recommend media information according to determining eliminate based on serving as a fill-in for recommendation media information；
Recommend media information according to ranking results selection candidate, and recommends media information replacement to eliminate using candidate and recommend matchmaker
Body information.
In one embodiment, include: according to determining recommendation media information of eliminating based on serving as a fill-in for recommendation media information
The superseded value t for recommending media information i is calculated, is superseded recommendation media when t be -1 determinations recommendation media information
Information；
Wherein vtiIt is served as a fill-in for media information i in the predetermined time interval；It is pre-
It fixes time and is spaced averagely being served as a fill-in for interior all recommendation media informations, in which:
N is the number for recommending media information.
In one embodiment,
Up (i) is to have all very useful families for pushing work to the user i media information issued；P (j) is issued to user i
Media information executes the credit rating for pushing the very useful family j of work；N (j) is that the history of very useful family j is served as a fill-in, and E is pre-set escapes
The factor.
In one embodiment, this method further comprises:
When pre-set iteration convergence condition is unsatisfactory for, the credit rating Pi of user i is iteratively updated；Wherein change
Include: for the condition of convergence front and back iteration result of the credit rating Pi of user i difference be less than pre-set first threshold value, and/
Or the number of iterations reaches pre-set second threshold value.
Media information can be the news that various news channels issue, and is also possible to the micro-blog information of personal sending, may be used also
To be Usenet, living forum, the various comment follow-up in circle of friends, embodiment of the present invention is to this and is not limited.
In conjunction with above-mentioned algorithm, specific example is given below.
For example, having comment comm1 under certain theme news (it is assumed that topic1), and have for comment comm1
There are A, B, C, the tetra- user tops D；The issuing time of comment comm1 are as follows: before 28 hours；And under theme news topic1
The average time of all comments is 3 hours, and wherein user A credit rating is 1, and the credit rating of user B is 2, and the credit rating of user C is
The credit rating 128 that 24, user D are, setting G value are 0.5, and comment comm1 length is 120, and the reply for comment comm1 is commented
It is 2 by number；It is assumed that when the time between the average time of all comments under the issuing time of comment and theme news topic1
When difference is greater than predetermined value (such as 24 hours), tcommFor the predetermined value, as 24 hours；When the issuing time and theme of comment are new
When hearing the time difference under topic1 between the average time of all comments no more than the predetermined value, tcommFor the issuing time of comment
With the time difference between the average time of all comments under theme news topic1.
The difference of the average time of all comments is (28-3) under the issuing time and theme news topic1 of comment comm1
=25 hours > 24 hours, therefore tcommIt is 24.
The then score ranking value T of comment comm1 are as follows:
For another example, for example, having comment comm2 under certain theme news (it is assumed that topic1), and it is directed to the comment
Comm2 has A, the tri- user tops B, C；The issuing time of comment comm1 are as follows: before 22 hours；And the theme news
The average time of all comments is 3 hours under topic1, and wherein user A credit rating is 1, and the credit rating of user B is 2, user C's
Credit rating is 24, and setting G value is 0.5, and comment comm2 length is 100, and the reply comment number for comment comm2 is 2；
It is assumed that making a reservation for when the time difference between the average time of all comments under the issuing time of comment and theme news topic1 is greater than
When being worth (such as 24 hours), tcommFor the predetermined value, as 24 hours；When under the issuing time of comment and theme news topic1
When time difference between the average time of all comments is not more than the predetermined value, tcommFor the issuing time and theme news of comment
Time difference under topic1 between the average time of all comments.
The difference of the average time of all comments is (22-3) under the issuing time and theme news topic1 of comment comm2
=19 hours < 24 hours, therefore tcommIt is 19.
The then score ranking value T of comment comm1 are as follows:
In the above-described example, it is related to pushing up the credit rating of the user of media information.It for ease of calculation, can also be by institute
There is the credit rating of user to be set as identical.
It is highly preferred that embodiment of the present invention is also based on the credit rating that certain algorithm calculates user.
For example, the credit rating at useful and each very useful family can determine each very useful family based on the history at each very useful family
Contribution factor to the user credit degree for issuing media information, wherein contribution factor is directly proportional to the credit rating at very useful family, and tribute
The history for offering the factor and very useful family, which is served as a fill-in, to be inversely proportional；By each very useful family to issue media information user credit degree contribution because
Son summation, to obtain the credit rating for the user for issuing media information.
Such as, it is assumed that user A has issued media information, and user B, C, D be after browsing media information, all make top to
The top of user A operates.And user B was pushed up altogether 1 time (comprising this top) in history；User C was pushed up 3 times altogether in history
(including this top)；User D was pushed up altogether 4 times (comprising this top) in history；Assuming that the credit rating of user A is PeopleRank
(A)；The credit rating of user B is PeopleRank (B)；The credit rating of user C is PeopleRank (C)；Then
In one embodiment, if being exactly one being served as a fill-in according to gathering for all media informations of whole system
A very big digraph.Contribution factor of each very useful family to the user credit degree for issuing media information is summed, to obtain hair
The user credit degree of media information includes: out
Determine the credit rating Pi for issuing the user i of media information；
Up (i) is to have all very useful families for pushing work to the user i media information issued；P (j) is to execute the top for pushing work
The credit rating of user j；N (j) is that the history of very useful family j is served as a fill-in, and E is pre-set escape factor.
Since media information continues updating, media information is served as a fill-in evidence also in continuous updating, therefore can root
According to the credit rating for being served as a fill-in the user according to successive ignition calculating sending media information of the media information of update.Believe when issuing media
After the credit rating convergence of the user of breath, as the PeopleRank value of user, this value are absolute values, when in use can be with
All users are carried out to user and are divided into N layer according to certain section, each level assigns certain value, facilitates from the background to user
It is managed.
In one embodiment,
When pre-set iteration convergence condition is unsatisfactory for, the credit rating P (i) of user i is iteratively updated；Wherein
Iteration convergence condition includes: that the difference of the front and back iteration result of the credit rating P (i) of user i is less than pre-set first threshold value,
And/or the number of iterations reaches pre-set second threshold value.
In one embodiment, it is served as a fill-in based on the history at very useful family, very useful family and the credit rating at very useful family determines hair
The credit rating of the user of media information includes: out
Determining transfer matrix is served as a fill-in based on the history at very useful family, very useful family；
All user credit degree vectors of preparatory assignment are multiplied with transfer matrix, to obtain all user credit degree squares
Battle array；
When pre-set iteration convergence condition is unsatisfactory for, all user credit degree matrixes are iteratively updated；Its
Middle iteration convergence condition includes: to have the difference of the front and back iteration result of at least one user to be less than in advance in all user credit degree matrixes
The third threshold value and/or the number of iterations being first arranged reach pre-set 4th threshold value.
User credit degree of the invention can be specifically calculated using numerous embodiments.For example, can be converted in calculating
The problem of Stationary Distribution to seek Markov chain, carry out problem analysis from the angle of markov chain.
For example, Fig. 2 is according to embodiment of the present invention top relation chain schematic diagram.
From Figure 2 it can be seen that user A has pushed up user B, user C and user D；User B has pushed up user C and user A；User C is pushed up
User D；User D has pushed up user A and user B.
The credit rating initial value that each user can be preset is 1, then transfer matrix M are as follows:
Wherein transfer matrix the first row first is classified as 0, indicates that the number of the very useful family A of user A is 0；Transfer matrix the first row
Second is classified as 1/2, indicates that the number of the very useful family A of user B is 1, and the total degree that user A is pushed up is 2；Transfer matrix the first row
Second is classified as 0, indicates that the number of the very useful family A of user C is 0；Transfer matrix the first row the 4th is classified as 1/2, indicates the very useful family user D
The number of A is 1, and the total degree that user A is pushed up is 2.
Similarly, the second row of transfer matrix first is classified as 1/3, indicates that the number of the very useful family B of user A is 1, and user B
It is 3 by top total degree；The second row of transfer matrix second is classified as 0, indicates that the number of the very useful family B of user B is 0；Transfer matrix second
Row third is classified as 0, indicates that the number of the very useful family B of user C is 0；The second row of transfer matrix the 4th is classified as 1/2, indicates that user D is very useful
The number of family B is 1, and the total degree that user B is pushed up is 2.
Similarly, transfer matrix the third line first is classified as 1/3, indicates that the number of the very useful family C of user A is 1, and user C
It is 3 by top total degree；Transfer matrix the third line second is classified as 1/2, indicates that the number of the very useful family C of user B is 1, and user's C quilt
The total degree on top is 2；Transfer matrix the third line third is classified as 0, indicates that the number of the very useful family C of user C is 0；Transfer matrix third
Row the 4th is classified as 0, indicates that the number of the very useful family C of user D is 0.
Similarly, transfer matrix fourth line first is classified as 1/3, indicates that the number of the very useful family D of user A is 1, and user D
It is 3 by top total degree；Transfer matrix fourth line second is classified as 0, indicates that the number of the very useful family D of user B is 0；Transfer matrix the 4th
Row third is classified as 1, indicates that the number of the very useful family D of user C is 1；Transfer matrix fourth line the 4th is classified as 0, indicates the very useful family user D
The number of D is 0.
For each user, the credit rating initial value that can preset each user is 1, then credit rating vector matrix is
Matrix M is multiplied by the credit rating vector matrix that the result of vector matrix V is exactly after this calculating, i.e.,
Therefore, the credit rating of user A is 1；The credit rating of user B is 5/6；The credit rating of user C is 5/6；The letter of user D
Expenditure is 4/3.
In iterative calculation, by vector matrix V again multiplied by transfer matrix M, as iteration result, it is
Then, persistently with vector matrix V again multiplied by transfer matrix M, until the credit rating of each user restrains.
In actual use, the data for accumulating certain time can make PeopleRank more efficient.Obtain each user's
After PeopleRank data, there is a PeopleRank value to each user, a mode can be changed and understood, user's
PeopleRank value represents value of the user in Commentary Systems, all behaviors of that user can be related with this value
System.For example, can use the assessment of the useful evidence of PeopleRank value replacement user: before, one withstands on introducing formula and carries out
When calculating one top only indicate a radix, after quoting PeopleRank, by the value on the top of user according to
PeopleRank's is replaced, and can more effectively recommend to comment warmly in this way.Actual in use, having been embodied, effect
Obviously.
In addition in terms of user management, the reaction the activity of the user of the credit rating of user, the assessment for high-quality user
There is good value with definition.
In one embodiment, this method further comprises:
Credit rating sequence based on each user, is ranked up each user；
According to ranking results, determine the user of predetermined number as recommended user, the comment information of recommended user to be made
To recommend comment information；And/or according to ranking results, determine the user of predetermined number user will be shielded as shielding user
Comment information as shielding comment information.
Based on above-mentioned detailed analysis, embodiment of the present invention also proposed a kind of user credit degree determining device.
Fig. 3 is according to embodiment of the present invention media information collator structure chart.
As shown in figure 3, media information collator includes:
Data determination unit 301, for receiving based on positive weights trigger data caused by triggering positive weights component, from institute
State the useful evidence that media information is determined in positive weights trigger data；It receives to reply based on triggering and replys triggering number caused by component
According to the reply data of the determining media information from the reply trigger data；Determine the attribute data of the media information；
Sorting values determination unit 302, for useful evidence, reply data and media information attribute data based on media information
Determine the ranking value of media information；
Sequencing unit 303, for being ranked up based on ranking value to media information.
In one embodiment, further comprise attenuation processing unit 304, executed for the ranking value to media information
Time attenuation operations, the ranking value after acquisition time decaying；
Sequencing unit 303 is ranked up media information for the ranking value after being decayed based on the time.
In one embodiment, the useful evidence of media information includes pushing up the credit rating Pi of the user of media information, is replied
Data include replying the number of users P of the media information, and media information attribute data is the length value C of the media information；To media
It includes: with time attenuation function (t that the ranking value of information, which executes time attenuation operations,comm+2)gThe ranking value of decaying media information；
T (i) is the ranking value after time decaying, in which:
Work as tiWithBetween time difference be greater than predetermined value when, tcommFor the predetermined value；Work as tiWithBetween time difference not
When greater than the predetermined value, tcommFor tiWithDifference, wherein tiFor the issuing time of the media information；It is right for the media information
The issuing time average value of all media informations under the theme answered；N is the number of users for pushing up media information；G is pre-set
Parameter.
It in one embodiment, further include recommending media information determination unit 305, it is pre- for being selected according to ranking results
Fixed number purpose media information, and to show recommendation media information as media information is recommended.
In one embodiment, recommend media information determination unit 305, be also used to determine interior at preset time intervals push away
Recommend the useful evidence of media information；Recommend media information according to determining eliminate based on serving as a fill-in for recommendation media information；According to ranking results
It selects candidate to recommend media information, and recommends media information replacement to eliminate using candidate and recommend media information.
In one embodiment, recommend media information determination unit 305, recommend eliminating for media information i for calculating
Value t is to eliminate to recommend media information when t is -1 determination recommendation media information；
Wherein vtiIt is served as a fill-in for media information i in the predetermined time interval；It is pre-
It fixes time and is spaced averagely being served as a fill-in for interior all recommendation media informations, in which:
N is the number for recommending media information.
Can by the present invention application with various media releasings application in.For example, can be applied to the new of news channel sending
It hears, is also possible to the micro-blog information of personal sending, can also be that Usenet, living forum, various in circle of friends are posted, this
Invention embodiment is to this and is not limited.
Fig. 4 is according to embodiment of the present invention media information recommending system first structure figure；
As shown in figure 4, the system includes:
Multiple issue client terminals 401, for sending media information to server 402；
Server 402 is used for publication medium information, receives and triggers number based on positive weights caused by triggering positive weights component
According to the useful evidence of determining media information from the positive weights trigger data；It receives to reply based on triggering and be returned caused by component
Multiple trigger data determines the reply data of the media information from the reply trigger data；Determine the media information
Attribute data；The ranking value of media information is determined based on the useful evidence, reply data and media information attribute data of media information；
Media information is ranked up based on ranking value.Fig. 4 show the schematic diagram of ranking results list.
Wherein, user can execute the processing for sending media information on various issue client terminals 401, these publication clients
End may include but be not limited to: functional mobile phone, smart phone, palm PC, PC (PC), tablet computer or individual
Digital assistants (PDA), etc..
Although enumerating the specific example of issue client terminal in detail above, those skilled in the art are it is to be appreciated that these are enumerated
It is only purposes of illustration, is not intended to limit the present invention the protection scope of embodiment.Moreover, these publishing sides can use it is clear
Implement body of looking at may include the Safari of Firefox, Apple of Internet Explorer of Microsoft, Mozilla, Opera,
The browsers such as Google Chrome, GreenBrowser.
Although listing some common browsers in detail above, it will be appreciated by those of skill in the art that the present invention is implemented
Mode is not limited to these browsers, but can be adapted for arbitrarily can be used for showing in web page server or archives economy
File simultaneously allows the application (App) that user interacts with file, these applications can be various browsers common at present, can also be with
It is other arbitrarily with the application program of web page browsing function.
In one embodiment, server 402 are also used to execute time attenuation operations to the ranking value of media information,
Ranking value after acquisition time decaying, and media information is ranked up based on the ranking value after time decaying.
In one embodiment 402, server is also used to select the media information of predetermined number according to ranking results,
Using as recommend media information；It shows and recommends media information.
Fig. 5 is according to embodiment of the present invention media information recommending the second structure chart of system；
As shown in figure 5, the system includes:
Multiple issue client terminals 401, for sending media information to server 402；
Server 402 is used for publication medium information, receives and triggers number based on positive weights caused by triggering positive weights component
According to the useful evidence of determining media information from the positive weights trigger data；It receives to reply based on triggering and be returned caused by component
Multiple trigger data determines the reply data of the media information from the reply trigger data；Determine the media information
Attribute data；The ranking value of media information is determined based on the useful evidence, reply data and media information attribute data of media information；
Media information is ranked up based on ranking value.
Wherein, user can execute the processing for sending media information on various issue client terminals 401, these publication clients
End may include but be not limited to: functional mobile phone, smart phone, palm PC, PC (PC), tablet computer or individual
Digital assistants (PDA), etc..
Although enumerating the specific example of issue client terminal in detail above, those skilled in the art are it is to be appreciated that these are enumerated
It is only purposes of illustration, is not intended to limit the present invention the protection scope of embodiment.Moreover, these publishing sides can use it is clear
Implement body of looking at may include the Safari of Firefox, Apple of Internet Explorer of Microsoft, Mozilla, Opera,
The browsers such as Google Chrome, GreenBrowser.
Although listing some common browsers in detail above, it will be appreciated by those of skill in the art that the present invention is implemented
Mode is not limited to these browsers, but can be adapted for arbitrarily can be used for showing in web page server or archives economy
File simultaneously allows the application (App) that user interacts with file, these applications can be various browsers common at present, can also be with
It is other arbitrarily with the application program of web page browsing function.
In one embodiment, server 502 are also used to determine the top for recommending media information at preset time intervals
Data；Recommend media information according to determining eliminate based on serving as a fill-in for recommendation media information；Recommend matchmaker according to ranking results selection candidate
Body information, and recommend media information to replace eliminating recommendation media information using candidate.
As shown in figure 5, server, which after determining multiple superseded recommendation media informations, can form to eliminate, recommends media
Information list, and recommend media information according to the media information ranking results selection candidate precalculated, and recommend using candidate
Media information, which replaces eliminating, recommends media information.It illustrates to eliminate in Fig. 5 and recommends media information list.
Indeed, it is possible to which the media information sequence side that embodiment of the present invention is proposed is embodied by diversified forms
Method, device and media information recommending system.
For example, the application programming interfaces centainly standardized can be followed, media information sort method is written as being installed to a
Plug-in card program in people's computer, mobile terminal etc. can also be encapsulated as application program so that user voluntarily downloads use.When
When being written as plug-in card program, a variety of card formats such as ocx, dll, cab can be implemented as.Can also by Flash plug-in unit,
The particular techniques such as RealPlayer plug-in unit, MMS plug-in unit, MIDI staff plug-in unit, ActiveX plug-in unit implement embodiment party of the present invention
The user credit degree that formula is proposed determines method.
It can be arranged by the media information that the storing mode of instruction or instruction set storage is proposed embodiment of the present invention
Sequence method is stored on various storage mediums.These storage mediums include but is not limited to: floppy disk, CD, DVD, hard disk, sudden strain of a muscle
It deposits, USB flash disk, CF card, SD card, mmc card, SM card, memory stick (Memory Stick), xD card etc..
Furthermore it is also possible to be applied to the media information sort method that embodiment of the present invention is proposed based on flash memory
In the storage medium of (Nand flash), such as USB flash disk, CF card, SD card, SDHC card, mmc card, SM card, memory stick, xD card etc..
To sum up, in embodiments of the present invention, it receives based on positive weights trigger data caused by triggering positive weights component,
The useful evidence of media information is determined from the positive weights trigger data；It receives to reply based on triggering and replys touching caused by component
Data are sent out, the reply data of the media information are determined from the reply trigger data；Determine the attribute of the media information
Data；The ranking value of media information is determined based on the useful evidence, reply data and media information attribute data of media information；It is based on
Ranking value is ranked up media information, it can be seen that, comment is not ranked up according still further to the quantity on top, but is based on media
The useful evidence of information replys data and the comprehensive ranking value for determining media information of media information attribute data, so as to effective
It prevents from practising fraud, improves sequence accuracy.Moreover, can periodically determine to eliminate and recommend media after using embodiment of the present invention
Information, so that the influence of Matthew effect be effectively reduced.
In addition, after using embodiment of the present invention, not further according to individual subscribers such as the log duration of user, dispatch frequencies
Behavior determines the credit rating of user.On the contrary, being served as a fill-in according to the history at very useful family, very useful family and the credit rating at very useful family determines hair
The credit rating of the user of media information out has fully considered the behavior of other users, it is thus determined that user credit degree accuracy rate
It is high.In addition, embodiment of the present invention avoid according to be easy to happen in individual subscriber behavior decision user credit degree cheating lack
It falls into, to further improve user credit degree accuracy rate.
More than, only presently preferred embodiments of the present invention is not intended to limit the scope of the present invention.It is all in this hair
Within bright spirit and principle, any modification, equivalent replacement, improvement and so on should be included in protection scope of the present invention
Within.
Claims (16)
1. a kind of media information sort method, which is characterized in that this method comprises:
It receives based on positive weights trigger data caused by triggering positive weights component, matchmaker is determined from the positive weights trigger data
The useful evidence of body information；It receives to reply based on triggering and replys trigger data caused by component, from the reply trigger data
Determine the reply data of the media information；Determine the attribute data of the media information；
The reply data of useful evidence, the media information based on the media information and the attribute data of the media information are true
The ranking value of the fixed media information, and media information is ranked up based on the ranking value；
According to the ranking results selection predetermined number media information, using as recommend media information；
It determines at preset time intervals, the useful evidence for recommending media information；
Serving as a fill-in for media information is recommended to recommend media information according to determining eliminate based on described；
Recommend media information according to ranking results selection candidate, and recommends media information to replace described wash in a pan using the candidate
Eliminate recommendation media information.
2. media information sort method according to claim 1, which is characterized in that this method further comprises:
Time attenuation operations are executed to the ranking value of the media information, the ranking value after acquisition time decaying；
It is described that media information is ranked up are as follows:
Media information is ranked up based on the ranking value after time decaying.
3. media information sort method according to claim 2, which is characterized in that the useful evidence of the media information includes
The credit rating Pi of the user of the media information is pushed up, the data of replying include replying the number of users P of the media information, the matchmaker
Body information attribute data are the length value C of the media information；The ranking value to the media information executes time decaying behaviour
Work includes: with time attenuation function (tcomm+2)gDecay the ranking value of the media information；
T (i) is the ranking value after time decaying, in which:
Work as tiWithBetween time difference be greater than predetermined value when, tcommFor the predetermined value；Work as tiWithBetween time difference be not more than
When the predetermined value, tcommFor tiWithDifference；tiFor the issuing time of the media information；For master corresponding to the media information
The issuing time average value of all media informations under topic；N is the number of users for pushing up the media information；G is pre-set ginseng
Number.
4. media information sort method according to claim 1, which is characterized in that this method further comprises:
Show the recommendation media information.
5. media information sort method according to claim 1, which is characterized in that described to be based on the recommendation media information
Serve as a fill-in according to determine eliminate recommend media information include:
The superseded value t for recommending media information i is calculated, is to eliminate to recommend media letter when t is -1 determination recommendation media information i
Breath；
Wherein vtiIt is served as a fill-in for media information i in the predetermined time interval；For the predetermined time
It is all in interval that the average of media information is recommended to be served as a fill-in, in which:
N is the number for recommending media information.
6. media information sort method according to claim 3, which is characterized in that described
Up (i) is to have all very useful families for pushing work to the user i media information issued；P (j) is the media issued to user i
Information executes the credit rating for pushing the very useful family j of work；N (j) be very useful family j history serve as a fill-in, E be it is pre-set escape because
Son.
7. media information sort method according to claim 6, which is characterized in that this method further comprises:
When pre-set iteration convergence condition is unsatisfactory for, the credit rating Pi of the user i is iteratively updated；Wherein institute
The difference for stating the front and back iteration result for the credit rating Pi that iteration convergence condition includes: the user i is less than pre-set first
Limit value and/or the number of iterations reach pre-set second threshold value.
8. a kind of media information collator characterized by comprising
Data determination unit, for receiving based on positive weights trigger data caused by triggering positive weights component, from the positive power
The useful evidence of media information is determined in weight trigger data；It receives to reply based on triggering and replys trigger data caused by component, from
The reply data of the media information are determined in the reply trigger data；Determine the attribute data of the media information；
Sorting values determination unit, for the reply data of useful evidence, the media information based on the media information and described
The attribute data of media information determines the ranking value of the media information；
Sequencing unit, for being ranked up based on the ranking value to media information；
Recommend media information determination unit, for the media information according to ranking results selection predetermined number, using as pushing away
Recommend media information；
The recommendation media information determination unit is also used to determine serving as a fill-in for the recommendation media information interior at preset time intervals
According to；Serving as a fill-in for media information is recommended to recommend media information according to determining eliminate based on described；Candidate is selected according to the ranking results
Recommend media information, and recommends media information to replace the superseded recommendation media information using the candidate.
9. media information collator according to claim 8, which is characterized in that it further comprise attenuation processing unit,
Time attenuation operations are executed for the ranking value to the media information, the ranking value after acquisition time decaying；
The sequencing unit is ranked up media information for the ranking value after being decayed based on the time.
10. media information collator according to claim 9, which is characterized in that the top data packet of the media information
Include the credit rating Pi for pushing up the user of the media information, the data of replying include replying the number of users P of the media information, described
Media information attribute data is the length value C of the media information；The ranking value to the media information executes time decaying
Operation includes: with time attenuation function (tcomm+2)gDecay the ranking value of the media information；
T (i) is the ranking value after time decaying, in which:
Work as tiWithBetween time difference be greater than predetermined value when, tcommFor the predetermined value；Work as tiWithBetween time difference be not more than
When the predetermined value, tcommFor tiWithDifference；Wherein tiFor the issuing time of the media information；For corresponding to the media information
The issuing time average value of all media informations under theme；N is the number of users for pushing up the media information；G is pre-set
Parameter.
11. media information collator according to claim 9, which is characterized in that the recommendation media information determines single
Member is also used to show the recommendation media information.
12. media information collator according to claim 8, which is characterized in that the recommendation media information determines single
Member is to eliminate recommendation media to believe when t be -1 determinations recommendation media information for calculating the superseded value t for recommending media information i
Breath；
Wherein vtiIt is served as a fill-in for media information i in the predetermined time interval；For the predetermined time
It is all in interval that the average of media information is recommended to be served as a fill-in, in which:
N is the number for recommending media information.
13. a kind of media information recommending system characterized by comprising
Issue client terminal, for sending media information to server；
Server receives for issuing the media information and is based on positive weights trigger data caused by triggering positive weights component,
The useful evidence of media information is determined from the positive weights trigger data；It receives to reply based on triggering and replys touching caused by component
Data are sent out, the reply data of the media information are determined from the reply trigger data；Determine the attribute of the media information
Data；The reply data of useful evidence, the media information based on the media information and the attribute data of the media information
Determine the ranking value of the media information；Media information is ranked up based on the ranking value；It is selected according to the ranking results
The media information for selecting predetermined number, using as recommend media information；It determines at preset time intervals, the recommendation media information
Useful evidence；Serving as a fill-in for media information is recommended to recommend media information according to determining eliminate based on described；It is selected according to the ranking results
It selects candidate and recommends media information, and recommend media information to replace the superseded recommendation media information using the candidate.
14. media information recommending system according to claim 13, which is characterized in that
Server is also used to execute the ranking value of the media information time attenuation operations, the sequence after acquisition time decaying
Value, and media information is ranked up based on the ranking value after time decaying.
15. media information recommending system according to claim 14, which is characterized in that
Server, be also used to according to the ranking results select predetermined number media information, using as recommend media information；Exhibition
Show the recommendation media information.
16. media information recommending system according to claim 15, which is characterized in that
Server is also used to determine the useful evidence for recommending media information at preset time intervals；Based on the recommendation matchmaker
Serving as a fill-in for body information recommends media information according to determining eliminate；Recommend media information, and benefit according to ranking results selection candidate
Media information is recommended to replace the superseded recommendation media information with the candidate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410192707.7A CN105095258B (en) | 2014-05-08 | 2014-05-08 | A kind of media information sort method, device and media information recommending system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410192707.7A CN105095258B (en) | 2014-05-08 | 2014-05-08 | A kind of media information sort method, device and media information recommending system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105095258A CN105095258A (en) | 2015-11-25 |
| CN105095258B true CN105095258B (en) | 2019-06-21 |
Family
ID=54575712
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410192707.7A Active CN105095258B (en) | 2014-05-08 | 2014-05-08 | A kind of media information sort method, device and media information recommending system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105095258B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105898336A (en) * | 2015-12-15 | 2016-08-24 | 乐视网信息技术（北京）股份有限公司 | Hot review determining method, hot review display system and server |
| CN105740473B (en) * | 2016-03-14 | 2021-03-02 | 腾讯科技（深圳）有限公司 | User generated content display method and device |
| CN107885808B (en) * | 2017-11-03 | 2021-03-30 | 四川文轩教育科技有限公司 | Shared resource file anti-cheating method |
| CN109213908A (en) * | 2018-08-01 | 2019-01-15 | 浙江工业大学 | A kind of academic meeting paper supplying system based on data mining |
| CN110087118B (en) * | 2019-04-26 | 2022-01-21 | 北京达佳互联信息技术有限公司 | Comment message processing method, comment message processing device, comment message processing terminal, comment message processing server and comment message processing medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002330245A (en) * | 2001-05-02 | 2002-11-15 | Dainippon Printing Co Ltd | Super facsimile system |
| CN102449658A (en) * | 2009-03-18 | 2012-05-09 | 踏途音乐公司 | Entertainment server and associated social networking services |
| CN102970352A (en) * | 2012-11-08 | 2013-03-13 | 北京小米科技有限责任公司 | Method and device for providing information |
| CN103034706A (en) * | 2012-12-07 | 2013-04-10 | 合一网络技术(北京)有限公司 | Video recommended list generating device and method based on information network |
| CN103164414A (en) * | 2011-12-09 | 2013-06-19 | 腾讯科技（深圳）有限公司 | Multimedia file comment display method and multimedia file comment display system |
| CN103678623A (en) * | 2013-12-18 | 2014-03-26 | 乐视网信息技术（北京）股份有限公司 | Method for displaying browser page and terminal for displaying browser page |
-
2014
- 2014-05-08 CN CN201410192707.7A patent/CN105095258B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002330245A (en) * | 2001-05-02 | 2002-11-15 | Dainippon Printing Co Ltd | Super facsimile system |
| CN102449658A (en) * | 2009-03-18 | 2012-05-09 | 踏途音乐公司 | Entertainment server and associated social networking services |
| CN103164414A (en) * | 2011-12-09 | 2013-06-19 | 腾讯科技（深圳）有限公司 | Multimedia file comment display method and multimedia file comment display system |
| CN102970352A (en) * | 2012-11-08 | 2013-03-13 | 北京小米科技有限责任公司 | Method and device for providing information |
| CN103034706A (en) * | 2012-12-07 | 2013-04-10 | 合一网络技术(北京)有限公司 | Video recommended list generating device and method based on information network |
| CN103678623A (en) * | 2013-12-18 | 2014-03-26 | 乐视网信息技术（北京）股份有限公司 | Method for displaying browser page and terminal for displaying browser page |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105095258A (en) | 2015-11-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108108821B (en) | Model training method and device | |
| CN105095258B (en) | A kind of media information sort method, device and media information recommending system | |
| CN105989074B (en) | A kind of method and apparatus recommend by mobile device information cold start-up | |
| CN106326391B (en) | Multimedia resource recommendation method and device | |
| CN106022505A (en) | Method and device of predicting user off-grid | |
| CN105912550A (en) | Method and device for information recommendation of mobile terminal | |
| CN104408640B (en) | Application software recommends method and device | |
| CN112183818A (en) | Recommendation probability prediction method and device, electronic equipment and storage medium | |
| CN103870553B (en) | A kind of input resource supplying method and system | |
| US20220261591A1 (en) | Data processing method and apparatus | |
| CN105512156B (en) | Click model generation method and device | |
| CN108062692B (en) | Recording recommendation method, device, equipment and computer readable storage medium | |
| CN103577511A (en) | Method, device and system for realizing media application | |
| CN106648688A (en) | Information display method and apparatus | |
| CN113268641B (en) | User data processing method based on big data and big data server | |
| CN108512883A (en) | A kind of information-pushing method, device and readable medium | |
| CN110598109A (en) | Information recommendation method, device, equipment and storage medium | |
| CN112100221A (en) | Information recommendation method and device, recommendation server and storage medium | |
| CN112446736A (en) | Click through rate CTR prediction method and device | |
| CN113869931A (en) | Advertisement putting strategy determining method and device, computer equipment and storage medium | |
| CN107526753B (en) | Recommendation method and device for application program | |
| CN107562432A (en) | Information processing method and related product | |
| CN110968802A (en) | User characteristic analysis method, analysis device and readable storage medium | |
| CN109684546A (en) | Recommended method, device, storage medium and terminal | |
| CN107547626B (en) | User portrait sharing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |