CN104463594A - Method and device for computing object influence - Google Patents
Method and device for computing object influence Download PDFInfo
- Publication number
- CN104463594A CN104463594A CN201310435049.5A CN201310435049A CN104463594A CN 104463594 A CN104463594 A CN 104463594A CN 201310435049 A CN201310435049 A CN 201310435049A CN 104463594 A CN104463594 A CN 104463594A
- Authority
- CN
- China
- Prior art keywords
- event
- power mark
- time effects
- time
- relevant
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and device for computing object influence. The method comprises the steps that at least one event is obtained, and the event comprises the information of event occurrence time, an event issuing object and a theme related to the event; according to the occurrence time, time influence grades are distributed to events related to the given theme respectively; and an event related to the objected is determined, and according to the time influence grade of the event, the influence grade of the object is computed. The computing complexity of object influence computing is lowered, and system overhead is lowered.
Description
Technical field
The application relates to Internet technical field, particularly relates to a kind of method and device thereof of calculating object influence power.
Background technology
In internet arena, the often influence power of needs assessment object (such as everyone, the businessman etc. of Internet provider product or service of Internet user, website).By the assessment to businessman's influence power, that those closely follow trend, deeply popular businessman can be found.In a commercial articles searching engine, can be expanded these probability is represented to promote Consumer's Experience to the dependent merchandise of the businessman of city's field sensitive.On a C2C website, can according to the assessment of user force being found out to those shopping intelligent (namely the people how selecting to be expert at very much in commodity).On a SNS website, also can find out according to the assessment of user force the user that those often deliver valuable suggestion.
Sequence of events that a certain object is associated (such as this object issues the behavioural information of different themes not having in the same time), can be used for assessing the influence power of this object.
Prior art, when the influence power of calculating object, needs the information transfering relation first extracted between object, calculates the strength of association between object, and then the influence power of calculating object.But in practice, along with passage of time, the number of objects in a network system can increase gradually, even if the quantity of object is linear increasing, the computing cost of object relationship also will become a square doubly growth.In addition, when some focus theme occurs, often forward the message that this focus theme is relevant between object insanely, this also can cause the computing cost of relation between analytic target rapidly to increase.
In sum, in known prior art, the technical scheme of calculating object influence power exists along with the time increases the defect that computing cost is the increase of geometry quantity, is therefore necessary that the technological means proposing to improve solves the problem.
Summary of the invention
The fundamental purpose of the application is the method and the device thereof that provide a kind of calculating object influence power, and the expense along with time increase calculating object influence power existed to solve prior art is the defect that geometry quantity increases.
For solving the problems of the technologies described above, according to an aspect of the application, provide a kind of method of calculating object influence power, it comprises: obtain at least one event, and described event comprises following information: Time To Event, the object issuing event, the theme that this event is relevant; Be that each event relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event; Determine the event relevant to object, and calculate the influence power mark of described object according to the time effects power mark of event.
Wherein, the described sequencing according to Time To Event is the step that each event relevant to given theme distributes time effects mark respectively, comprising: each ordering of events relevant to given theme is formed sequence of events by the sequencing according to Time To Event; Be that each event in described sequence of events distributes time effects power mark respectively according to descending function given in advance, wherein, the mark of the time effects power event ranked forefront in described sequence of events is higher than the time effects power mark of the event leant backward.
Wherein, described descending function comprises one of following: the geometric sequence of geometric sequence, arithmetic sequence, consideration time span, consider the geometric sequence of sequence of events population size.
Wherein, if only comprise an event in described sequence of events, then the time effects power mark distributed for this event is the constant value pre-set.
Wherein, describedly determine the event relevant to object, and the step of the influence power mark of described object is calculated according to the time effects power mark of event, comprise: determine the multiple events relevant to object, being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object.
Wherein, describedly determine the event relevant to object, and the step of the influence power mark of described object is calculated according to the time effects power mark of event, comprise: determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object.
Wherein, describedly determine the event relevant to object, and the step of the influence power mark of described object is calculated according to the time effects power mark of event, comprise: determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object.
Wherein, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following: the mean value of multiple topic weights; The minimum value of multiple topic weights; The maximal value of multiple topic weights; The median of multiple topic weights.
Wherein, describedly determine the event relevant to object, and the step of the influence power mark of described object is calculated according to the time effects power mark of event, comprise: the event relevant to described object is polymerized according to theme, calculate the time effects power mark of this object under each theme respectively; The influence power mark of this object is calculated according to the time effects power mark of this object under each theme.
Wherein, in the step of the event relevant to described object being carried out according to theme be polymerized, also comprise: the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme; Calculating in the step of influence power mark of this object according to the time effects power mark of this object under each theme, also comprise: the time effects power mark under each theme is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
According to the another aspect of the application, additionally provide a kind of device of calculating object influence power, it comprises: acquisition module, and for obtaining at least one event, described event comprises following information: Time To Event, the object issuing event, the theme that this event is relevant; Event score computing module, for being that each event relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event; Object influences power mark computing module, for determining the event relevant to object, and calculates the influence power mark of described object according to the time effects power mark of event.
Wherein, described event score computing module comprises: ordering of events module, for the sequencing according to Time To Event, each ordering of events relevant to given theme is formed sequence of events; Event score distribution module, for being that each event in described sequence of events distributes time effects power mark respectively according to descending function given in advance, wherein, the time effects power mark of the event ranked forefront in described sequence of events is higher than the time effects power mark of the event leant backward.
Wherein, described descending function comprises one of following: the geometric sequence of geometric sequence, arithmetic sequence, consideration time span, consider the geometric sequence of sequence of events population size.
Wherein, if only comprise an event in described sequence of events, then described event score distribution module is the time effects power mark that this event is distributed is the constant value pre-set.
Wherein, described object influences power mark computing module is further used for: determine the multiple events relevant to object, and being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object.
Wherein, described object influences power mark computing module is further used for: determine the multiple events relevant to object, and the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object.
Wherein, described object influences power mark computing module is further used for: determine the multiple events relevant to object, and the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object.
Wherein, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following: the mean value of multiple topic weights; The minimum value of multiple topic weights; The maximal value of multiple topic weights; The median of multiple topic weights.
Wherein, described object influences power mark computing module is further used for: the event relevant to described object be polymerized according to theme, calculate the time effects power mark of this object under each theme respectively, wherein, the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme; Calculate the influence power mark of this object according to the time effects power mark of this object under each theme, wherein, the time effects power mark under each theme is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
According to the technical scheme of the application, by the importance of this object of mark average evaluation of the event according to object publishing, propagation relation between not consideration event and between object, thus reduce computation complexity and decrease system overhead, thus can on large-scale data the influence power of calculating object quickly.
Accompanying drawing explanation
Accompanying drawing described herein is used to provide further understanding of the present application, and form a application's part, the schematic description and description of the application, for explaining the application, does not form the improper restriction to the application.In the accompanying drawings:
Fig. 1 is the process flow diagram of the method for calculating object influence power according to the embodiment of the present application;
Fig. 2 is the structured flowchart of the device of calculating object influence power according to the application's embodiment;
Fig. 3 is the structured flowchart of the device of calculating object influence power according to another embodiment of the application.
Embodiment
For making the object of the application, technical scheme and advantage clearly, below in conjunction with drawings and the specific embodiments, the application is described in further detail.
According to the embodiment of the application, provide a kind of method of calculating object influence power.Fig. 1 is the process flow diagram of the method for calculating object influence power according to the embodiment of the present application, as shown in Figure 1, and the method comprising the steps of S102-S106.
Step S102, obtains at least one event, and described event comprises following information: the theme that Time To Event, the object issuing this event, this event are relevant.
The method of the calculating object influence power of the application can be applied to the plurality of application scenes such as business website, commodity website or social networks (SNS).In different application scenarioss, the information such as event, object and theme can represent different implications, and in simple terms, object just refers to the user registered in network, and event refers to the concrete behavior of this object on network, and theme is then the core of this event.
The result of the sequencing arrangement of the time occurred according to event corresponding to the set of same subject event can form the sequence of events of this theme.The time effects power mark of event is denoted as here according to the evaluation of Time To Event to the importance of a certain event in the sequence of events belonging to this event.
The influence power of object is then the comprehensive of the time effects power mark of each event of this object publishing.
Illustrate some concepts that the application relates to below.
For a business website, event can be the operation behavior of user in this business website to a certain commodity, and the behavior that the behavior of merchandise news or buyer buy commodity occurs such as seller.In such example, object is the user (i.e. seller or buyer) making the behavior; Theme as commodity corresponding to the behavior; The time that event occurs is the time that this seller issues that time of these commodity or buyer buy these commodity; The sequence of events of certain theme is the result that the set of the commodity operation behavior by belonging to a certain commodity together is sorted according to the sequencing of issuing time; Time effects power mark is for evaluating the importance of a certain event in Time To Event dimension in certain sequence of events.Relevant according to the sequence of time of origin in the time effects power mark of event and this event sequence of events belonging to it.The time of origin of a certain event relative to other events in the sequence of events belonging to this event time of origin more early, then the time effects power mark of this event is higher.The influence power of object is the comprehensive of the time effects power mark of all events that this user issues.If the time effects power mark of each event that a certain seller issues is always higher, then the influence power of this seller is general comparatively large, illustrates that this seller is deeply popular.If the time effects power mark of each event that a certain buyer issues is always higher, the influence power of this buyer is general comparatively large, illustrates that the assurance ability of the fashion trend to commodity of this buyer is stronger.
For a SNS system, event is the behavior that user issues article (or model) in SNS system, and object is the registered user issuing this article, themes as the topic that this article is corresponding; The time of event generation is the time of object publishing article (or model); The sequence of events of certain theme is the result that the behavior set of the article by belonging to same topic is sorted according to the sequencing of issuing time; The time effects power mark of event for evaluating in the sequence of events of certain topic, the importance that a certain event is relevant to time of origin; The influence power of object is the comprehensive of the time effects mark of all articles that this user issues.If the time effects power mark of the article that this user issues is always higher, then the influence power of this user is general comparatively large, illustrates that this user often delivers valuable suggestion.
It should be noted that, the theme of above-mentioned business website or commodity website is described for " product ", and the application is not limited to this, can also represent with other dimension such as brand, material, style or standard product unit (SPU) etc.
Step S104 is that each event in the sequence of events relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event.
For given theme, the whole events containing this theme are selected, and according to the sequencing of Time To Event, the ordering of events selected is formed sequence of events.Then, according to descending function given in advance, for each event in sequence of events distributes a mark, namely obtain time effects power mark.Wherein, the time effects power mark of the event ranked forefront in sequence of events is always higher than the time effects power mark of the event leant backward in sequence of events.
The time effects power mark of event is for evaluating the importance of event, in above-mentioned processing procedure, the time effects power mark (importance) of the event of same subject is decayed in time, so just need not propagate relation between consideration event, thus reduce computation complexity.
Wherein, described descending function can have multiple implementation, includes but not limited to one of following:
(1) by geometric sequence F (n)=a^n, draw the time effects power mark of the n-th event in sequence of events, wherein, a is the constant between given in advance 0 to 1, the time effects power mark of the n-th event in F (n) i.e. sequence of events;
(2) by arithmetic sequence F (n)=-b*n, draw the time effects power mark of the n-th event in sequence of events, wherein, b is constant given in advance, the time effects power mark of the n-th event in F (n) i.e. sequence of events;
(3) the geometric sequence F (n of time span is considered by certain, t, t_last)=(a^n)/(t – t_last), draw the time effects power mark of the n-th event in sequence of events, wherein, the Time To Event of the n-th event is t, the Time To Event of (n-1)th event is t_last, the time effects power mark of the n-th event in F (n, t, t_last) i.e. sequence of events;
(4) considered geometric sequence F (n, N)=(a^ (n/N)) of sequence population size by certain, draw the time effects power mark of the n-th event in sequence of events, wherein, N is the sum of whole event in sequence of events; The time effects power mark of the n-th event in F (n, N) i.e. sequence of events.
In above process, when only comprising an event in sequence of events, the time effects power of this event must be divided into a constant value given in advance.It should be noted that, in actual applications, corresponding descending function should be selected to calculate according to actual conditions, the application does not limit this.
Step S106, determines the event relevant to object, and the influence power mark of time effects power mark calculating object according to event.For each object, according to the time effects power mark of each event under relevant each theme that this object is correlated with, calculate the influence power mark of this object.That is, first calculate the influence power mark of each event relevant to object, then COMPREHENSIVE CALCULATING obtains the time effects power mark of all events corresponding to this object.
Particularly, the account form of the influence power mark of the object of the application includes but not limited to one of under type (1)-(4).
(1) determine the multiple events relevant to object, being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object;
(2) determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object;
(3) determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object, wherein, weight can be set according to the different events that themes as;
Wherein, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following:
The mean value of multiple topic weights;
The minimum value of multiple topic weights;
The maximal value of multiple topic weights;
The median of multiple topic weights.
(4) first the event relevant to object is polymerized according to theme, calculate the time effects power mark of this object under each theme respectively, further, the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme; Then the time effects power mark of this object under all themes is calculated according to the time effects power mark of this object under each theme, further, time effects power mark under different themes is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
Above-mentioned (4) are than more preferably method, when an object relates to a large amount of event, and event is when relating to a large amount of theme, if adopt the method for (4), more complicated weighting logic can be realized eaily, and be convenient to realize in the distributed system of similar map-reduce form, speed up processing.
In practice, can select corresponding account form according to actual needs, the application does not limit.
According to the embodiment of the application, by the importance of this object of mark average evaluation of the event of object publishing (if object score of event of being correlated with is generally high, then the influence power of this object is inevitable high), do not consider the propagation relation between object, thus reduce computation complexity, thus realize the influence power of calculating object quickly on large-scale data.
The process of the calculating object influence power of the application is described below in conjunction with instantiation.
Example one, calculating the influence power of certain shopping platform seller, issuing the new product time as weighing his influence power in this market by seller.
In this example, event is that seller issues new product at certain shopping platform, and object is seller, themes as product.Concrete comprises following 8 events altogether: (1) seller's first was on 03 01st, 2013 release product A; (2) seller's second was on 03 13rd, 2013 release product B; (3) seller's second was on 03 15th, 2013 release product A; (4) seller's second was on 03 21st, 2013 release product C; (5) seller third was on 04 01st, 2013 release product C; (6) seller's first was on 04 02nd, 2013 release product B; (7) seller's first was on 04 05th, 2013 release product C; (8) retainer of a big family is sold on 04 16th, 2013 release product A.
(1) first, obtain the information of event, event summary relevant for identical product also in chronological sequence sequentially being sorted forms sequence of events:
Product A event 1-> event 3-> event 8
Product B event 2-> event 6
Products C event 4-> event 5-> event 7
(2) time-sequencing on certain event product belonging to it is denoted as n, n be more than or equal to 1 natural number.Geometric sequence is adopted to be that each event that described product is relevant distributes time effects power mark, by the time effects power mark formation geometric sequence of each event that described product is correlated with.Described geometric sequence can be expressed as S (x)=2^ (1-n), the time effects power mark that S (x) is event.According to the above-mentioned mode for event distribution time effects power mark, determine the score of each event that each product is correlated with.
Therefore, the time effects power of the event 1 that product A is associated must be divided into 1, and the time effects power of the event 3 that product A is associated must be divided into 1/2, and the time effects power of the event 8 that product A is associated must be divided into 1/4.
The time effects power of the event 2 that product B is associated must be divided into 1, and the time effects power of the event 6 that product B is associated must be divided into 1/2.
The time effects power of the event 4 that products C is associated must be divided into 1, and the time effects power of the event 5 that products C is associated must be divided into 1/2, and the time effects power of the event 7 that products C is associated must be divided into 1/4.
(3) influence power of seller must be divided into the average mark of the time effects power score of all events involved by this seller, then:
Seller's first relates to event 1, event 6 and event 7 totally 3 events, and the influence power of seller's first must be divided into the average mark of the time effects power score sum of above-mentioned 3 events, that is: (1+1/2+1/4)/3=0.583;
Seller's second relates to event 2, event 3 and event 4 totally 3 events, and the influence power of seller's second must be divided into the average mark of the time effects power score sum of above-mentioned 3 events, that is: (1+1/2+1)/3=0.833;
Seller third only relates to event 5, and the influence power score of seller third is the score of the time effects power of event 5: 0.5;
Sell the retainer of a big family and only relate to event 8, the influence power score of selling the retainer of a big family is the score of the time effects power of event 8: 0.25.
Calculated the influence power of seller with simple account form by said process.The highest from the influence power score of mark seller second, the actual conditions that this and data embody are consistent because on the whole the time of seller's second release product more Zao than other seller.
Example two, calculating the influence power of certain shopping platform seller, issuing the new product time as weighing his influence power in this market by seller.
In this example, event is for seller is at certain shopping platform issue different product (belonging to different classification in product classification tree), and object is seller, themes as the product classification belonging to product.Concrete comprises following 6 events altogether: (1) seller's first was on 04 21st, 2013 release products, and classification belonging to product is A1/B2; (2) seller's second was on 04 22nd, 2013 release products, and classification belonging to product is A1/B2; (3) seller's second was on 04 24th, 2013 release products, and classification belonging to product is A2/B6; (4) seller's first was on 04 25th, 2013 release products, and classification belonging to product is A2/B3; (5) seller's second was on 04 27th, 2013 release products, and classification belonging to product is A1/B4; (6) seller's second was on 04 29th, 2013 release products, and classification belonging to product is A2/B5.
(1) obtain the information of event, by event summary relevant for classification belonging to identical product and in chronological sequence order sequence forms sequence of events:
Product classification A1 event 1 (score 1) – > event 2 (score 1/2)-> event 5 (score 1/4)
Product classification A2 event 3 (score 1) – > event 4 (score 1/2)-> event 6 (score 1/4)
Product classification B2 event 1 (score 1) – > event 2 (score 1/2)
Product classification B3 event 4 (score 1)
Product classification B4 event 5 (score 1)
Product classification B5 event 6 (score 1)
Product classification B6 event 3 (score 1)
(2) certain event certain product classification belonging to it (time-sequencing be denoted as c) is denoted as n, n be more than or equal to 1 natural number.Geometric sequence event is adopted to be that each event that described product classification is relevant distributes time effects power mark, by the time effects power mark formation geometric sequence of each event that described product classification is correlated with.Described geometric sequence can be expressed as S (x, c)=2^ (1-n), the time effects power mark that S (x, c) is event.According to the above-mentioned mode for event distribution time effects power mark, determine the time effects power score of each event that each product classification is correlated with.Setting this seller in this example must be divided into this seller at the best result of this product class time effects power of involved event now in this product class time effects power now.
Therefore, the time effects power of the event 1 that product classification A1 is associated must be divided into 1, and the time effects power of the event 2 that product classification A1 is associated must be divided into 1/2, and the time effects power of the event 5 that product classification A1 is associated must be divided into 1/4.
The time effects power of the event 3 that product classification A2 is associated must be divided into 1, and the time effects power of the event 4 that product classification A2 is associated must be divided into 1/2, and the time effects power of the event 6 that product classification A2 is associated must be divided into 1/4.
The time effects power of the event 1 that product classification B2 is associated must be divided into 1, and the time effects power of the event 2 that product classification B2 is associated must be divided into 1/2.
The time effects power of the event 4 that product classification B3 is associated must be divided into 1.
The time effects power of the event 5 that product classification B4 is associated must be divided into 1.
The time effects power of the event 6 that product classification B5 is associated must be divided into 1.
The time effects power of the event 3 that product classification B6 is associated must be divided into 1.
(3) influence power of seller must be divided into the average mark of the score of the time effects power of all events involved by this seller, then:
Seller's first relates to event 1 and event 4 totally 2 events and A1, A2, B2 and B3 totally 4 product classifications, the influence power of seller's first must be divided into the average mark of the time effects power score sum of the event that above-mentioned 4 product classifications are corresponding, that is: (1+1/2+1+1)/7=0.5;
Seller's second relates to event 2, event 3, event 5 and event 6 totally 4 events and A1, A2, B2, B4, B5 and B6 totally 6 product classifications (only calculating the best result of this seller in this product class involved event now), the influence power of seller's second must be divided into the average mark of the time effects power score sum of the event that above-mentioned 6 product classifications are corresponding, i.e. (1/2+1+1/2+1+1+1)/7=0.714.
Calculated the influence power of seller with simple account form by said process.From influence power score higher than seller's first of the influence power score of mark seller second, the actual conditions that this and data embody are consistent, because several product classifications only have the product of seller's second.
Need to illustrate, above-mentioned example one and example two are only exemplary illustration, do not limit the protection domain of the application.
According to the embodiment of the application, additionally provide a kind of device of calculating object influence power, with reference to figure 2, described device at least comprises: acquisition module 22, event score computing module 24 and object influences power mark computing module 26, be described below in detail the 26S Proteasome Structure and Function of above-mentioned each module.
Acquisition module 22, for obtaining at least one event, described event comprises following information: Time To Event, the object issuing event, the theme that this event is relevant.
Event score computing module 24, for being that each event relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event.
With reference to figure 3, described event score computing module 24 comprises further: ordering of events module 242 and event score distribution module 244.The each ordering of events relevant to given theme is formed sequence of events for the sequencing according to Time To Event by ordering of events module 242.Event score distribution module 244 couples mutually with ordering of events module 242, for being that each event in described sequence of events distributes time effects power mark respectively according to descending function given in advance, wherein, the time effects power mark of the event ranked forefront in described sequence of events is higher than the time effects power mark of the event leant backward.Further, described descending function includes but not limited to one of following: the geometric sequence of geometric sequence, arithmetic sequence, consideration time span, consider the geometric sequence of sequence of events population size.
In addition, if only comprise an event in described sequence of events, then described event score distribution module is the time effects power mark that this event is distributed is the constant value pre-set.
Object influences power mark computing module 26 couples mutually with event score computing module 24, for determining the event relevant to object, and calculates the influence power mark of described object according to the time effects power mark of event.
In actual applications, object influences power mark computing module 26 can the influence power mark of calculating object in the following ways:
(1) determine the multiple events relevant to object, being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object;
(2) determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object;
(3) determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object, wherein, weight can be set according to the different events that themes as;
Further, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following:
The mean value of multiple topic weights;
The minimum value of multiple topic weights;
The maximal value of multiple topic weights;
The median of multiple topic weights.
(4) first the event relevant to object is polymerized according to theme, calculate the time effects power mark of this object under each theme respectively, further, the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme; Then the influence power mark of this object is calculated according to the time effects power mark of this object under each theme, further, time effects power mark under each theme is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
Above-mentioned (4) are than more preferably method, when an object relates to a large amount of event, and event is when relating to a large amount of theme, if adopt the method for (4), can more conveniently realize more complicated weighting logic, and be convenient to realize in the distributed system of similar map-reduce form, speed up processing.
The operation steps of the method for the application is corresponding with the architectural feature of device, can be cross-referenced, repeats no longer one by one.
In sum, according to the technique scheme of the application, by the importance of this object of mark average evaluation of the event according to object publishing, do not consider the propagation relation between object, thus reduce computation complexity and decrease system overhead, thus can on large-scale data the influence power of calculating object quickly.
In one typically configuration, computing equipment comprises one or more processor (CPU), input/output interface, network interface and internal memory.
Internal memory may comprise the volatile memory in computer-readable medium, and the forms such as random access memory (RAM) and/or Nonvolatile memory, as ROM (read-only memory) (ROM) or flash memory (flashRAM).Internal memory is the example of computer-readable medium.
Computer-readable medium comprises permanent and impermanency, removable and non-removable media can be stored to realize information by any method or technology.Information can be computer-readable instruction, data structure, the module of program or other data.The example of the storage medium of computing machine comprises, but be not limited to phase transition internal memory (PRAM), static RAM (SRAM), dynamic RAM (DRAM), the random access memory (RAM) of other types, ROM (read-only memory) (ROM), Electrically Erasable Read Only Memory (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc ROM (read-only memory) (CD-ROM), digital versatile disc (DVD) or other optical memory, magnetic magnetic tape cassette, tape magnetic rigid disk stores or other magnetic storage apparatus or any other non-transmitting medium, can be used for storing the information can accessed by computing equipment.According to defining herein, computer-readable medium does not comprise temporary computer readable media (transitory media), as data-signal and the carrier wave of modulation.
Also it should be noted that, term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability, thus make to comprise the process of a series of key element, method, commodity or equipment and not only comprise those key elements, but also comprise other key elements clearly do not listed, or also comprise by the intrinsic key element of this process, method, commodity or equipment.When not more restrictions, the key element limited by statement " comprising ... ", and be not precluded within process, method, commodity or the equipment comprising described key element and also there is other identical element.
It will be understood by those skilled in the art that the embodiment of the application can be provided as method, system or computer program.Therefore, the application can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the application can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code.
The foregoing is only the embodiment of the application, be not limited to the application.To those skilled in the art, the application can have various modifications and variations.Any amendment done within all spirit in the application and principle, equivalent replacement, improvement etc., within the right that all should be included in the application.
Claims (19)
1. a method for calculating object influence power, is characterized in that, comprising:
Obtain at least one event, described event comprises following information: Time To Event, the object issuing event, the theme that this event is relevant;
Be that each event relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event;
Determine the event relevant to object, and calculate the influence power mark of described object according to the time effects power mark of event.
2. method according to claim 1, is characterized in that, the described sequencing according to Time To Event is the step that each event relevant to given theme distributes time effects power mark respectively, comprising:
The each ordering of events relevant to given theme is formed sequence of events by the sequencing according to Time To Event;
Be that each event in described sequence of events distributes time effects power mark respectively according to descending function given in advance, wherein, the time effects power mark of the event ranked forefront in described sequence of events is higher than the time effects power mark of the event leant backward.
3. method according to claim 2, is characterized in that, described descending function comprises one of following: the geometric sequence of geometric sequence, arithmetic sequence, consideration time span, consider the geometric sequence of sequence of events population size.
4. method according to claim 2, is characterized in that, if only comprise an event in described sequence of events, then the time effects power mark distributed for this event is the constant value pre-set.
5. method according to claim 1, is characterized in that, describedly determines the event relevant to object, and calculates the step of the influence power mark of described object according to the time effects power mark of event, comprising:
Determine the multiple events relevant to object, being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object.
6. method according to claim 1, is characterized in that, describedly determines the event relevant to object, and calculates the step of the influence power mark of described object according to the time effects power mark of event, comprising:
Determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object.
7. method according to claim 1, is characterized in that, describedly determines the event relevant to object, and calculates the step of the influence power mark of described object according to the time effects power mark of event, comprising:
Determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object.
8. method according to claim 7, is characterized in that, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following:
The mean value of multiple topic weights;
The minimum value of multiple topic weights;
The maximal value of multiple topic weights;
The median of multiple topic weights.
9. method according to claim 1, is characterized in that, describedly determines the event relevant to object, and calculates the step of the influence power mark of described object according to the time effects power mark of event, comprising:
The event relevant to described object is polymerized according to theme, calculates the time effects power mark of this object under each theme respectively;
The influence power mark of this object is calculated according to the time effects power mark of this object under each theme.
10. method according to claim 9, is characterized in that,
In the step of the event relevant to described object being carried out according to theme be polymerized, also comprise: the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme;
Calculating in the step of influence power mark of this object according to the time effects power mark of this object under each theme, also comprise: the time effects power mark under each theme is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
The device of 11. 1 kinds of calculating object influence powers, is characterized in that, comprising:
Acquisition module, for obtaining at least one event, described event comprises following information: Time To Event, the object issuing event, the theme that this event is relevant;
Event score computing module, for being that each event relevant to given theme distributes time effects power mark respectively according to the sequencing of Time To Event;
Object influences power mark computing module, for determining the event relevant to object, and calculates the influence power mark of described object according to the time effects power mark of event.
12. devices according to claim 11, is characterized in that, described event score computing module comprises:
Ordering of events module, forms sequence of events for the sequencing according to Time To Event by each ordering of events relevant to given theme;
Event score distribution module, for being that each event in described sequence of events distributes time effects power mark respectively according to descending function given in advance, wherein, the time effects power mark of the event ranked forefront in described sequence of events is higher than the time effects power mark of the event leant backward.
13. devices according to claim 12, is characterized in that, described descending function comprises one of following: the geometric sequence of geometric sequence, arithmetic sequence, consideration time span, consider the geometric sequence of sequence of events population size.
14. devices according to claim 12, is characterized in that, if only comprise an event in described sequence of events, then described event score distribution module is the time effects power mark that this event is distributed is the constant value pre-set.
15. devices according to claim 11, is characterized in that, described object influences power mark computing module is further used for:
Determine the multiple events relevant to object, being averaged by the time effects power mark of the plurality of event calculates the influence power mark of this object.
16. devices according to claim 11, is characterized in that, described object influences power mark computing module is further used for:
Determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to Time To Event the influence power mark that average computation obtains this object.
17. devices according to claim 11, is characterized in that, described object influences power mark computing module is further used for:
Determine the multiple events relevant to object, the time effects power mark of the plurality of event is weighted according to event weights given in advance the influence power mark that average computation obtains this object.
18. devices according to claim 17, is characterized in that, if an event is relevant to multiple theme, then event weights given in advance is relevant to one of following:
The mean value of multiple topic weights;
The minimum value of multiple topic weights;
The maximal value of multiple topic weights;
The median of multiple topic weights.
19. devices according to claim 11, is characterized in that, described object influences power mark computing module is further used for:
The event relevant to described object is polymerized according to theme, calculate the time effects power mark of this object under each theme respectively, wherein, the event under same subject is weighted according to time factor and on average obtains the time effects power mark of this event under this theme;
Calculate the influence power mark of this object according to the time effects power mark of this object under each theme, wherein, the time effects power mark under each theme is weighted on average according to topic weights given in advance, calculates the influence power mark of this object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310435049.5A CN104463594B (en) | 2013-09-23 | 2013-09-23 | The method and its device of computing object influence power |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310435049.5A CN104463594B (en) | 2013-09-23 | 2013-09-23 | The method and its device of computing object influence power |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104463594A true CN104463594A (en) | 2015-03-25 |
| CN104463594B CN104463594B (en) | 2018-05-29 |
Family
ID=52909586
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310435049.5A Active CN104463594B (en) | 2013-09-23 | 2013-09-23 | The method and its device of computing object influence power |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104463594B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105760618A (en) * | 2016-03-08 | 2016-07-13 | 中国人民解放军总参谋部第五十四研究所 | Target situation display method based on GIS (Geographic Information System) facing virtual process |
| CN106611339A (en) * | 2015-10-21 | 2017-05-03 | 阿里巴巴集团控股有限公司 | Seed user screening method, and method and apparatus for evaluating user influence of product |
| CN108959249A (en) * | 2018-06-21 | 2018-12-07 | 中国农业科学院农田灌溉研究所 | A kind of evaluation method of periodical best paper |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251850A (en) * | 2008-01-04 | 2008-08-27 | 杨虡 | Internet topics ranking system and method based on user prestige |
| CN102426610A (en) * | 2012-01-13 | 2012-04-25 | 中国科学院计算技术研究所 | Microblog rank searching method and microblog searching engine |
| CN102945279A (en) * | 2012-11-14 | 2013-02-27 | 清华大学 | Evaluating method and device of influence effect of microblog users |
-
2013
- 2013-09-23 CN CN201310435049.5A patent/CN104463594B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251850A (en) * | 2008-01-04 | 2008-08-27 | 杨虡 | Internet topics ranking system and method based on user prestige |
| CN102426610A (en) * | 2012-01-13 | 2012-04-25 | 中国科学院计算技术研究所 | Microblog rank searching method and microblog searching engine |
| CN102945279A (en) * | 2012-11-14 | 2013-02-27 | 清华大学 | Evaluating method and device of influence effect of microblog users |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611339A (en) * | 2015-10-21 | 2017-05-03 | 阿里巴巴集团控股有限公司 | Seed user screening method, and method and apparatus for evaluating user influence of product |
| CN106611339B (en) * | 2015-10-21 | 2020-12-29 | 创新先进技术有限公司 | Seed user screening method, and product user influence evaluation method and device |
| CN105760618A (en) * | 2016-03-08 | 2016-07-13 | 中国人民解放军总参谋部第五十四研究所 | Target situation display method based on GIS (Geographic Information System) facing virtual process |
| CN105760618B (en) * | 2016-03-08 | 2019-02-19 | 中国人民解放军总参谋部第五十四研究所 | Target situation methods of exhibiting based on Virtual process geography GIS |
| CN108959249A (en) * | 2018-06-21 | 2018-12-07 | 中国农业科学院农田灌溉研究所 | A kind of evaluation method of periodical best paper |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104463594B (en) | 2018-05-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tewari | Generating items recommendations by fusing content and user-item based collaborative filtering | |
| CN105678580A (en) | E-commerce platform and method for issuing coupons | |
| CN103309894B (en) | Based on search implementation method and the system of user property | |
| CN108108380A (en) | Search ordering method, searching order device, searching method and searcher | |
| CN104915423B (en) | The method and apparatus for obtaining target user | |
| CN105335409A (en) | Target user determination method and device and network server | |
| CN104281664B (en) | Distributed figure computing system data segmentation method and system | |
| CN110941762B (en) | Method for improving sequencing result of personalized recommendation system | |
| CN104574126A (en) | User characteristic identification method and device | |
| CN105989146B (en) | Object methods of exhibiting and device | |
| CN107093081A (en) | Service strategy formulating method and device | |
| CN104463594A (en) | Method and device for computing object influence | |
| CN111340522B (en) | Resource recommendation method, device, server and storage medium | |
| CN105809379A (en) | Logistics branch evaluation method, device and electronic device | |
| CN103337028A (en) | Recommendation method and device | |
| Sabar et al. | A variable local search based memetic algorithm for the load balancing problem in cloud computing | |
| CN102737038B (en) | Degree of association defining method and device, information providing method and device | |
| CN106257449A (en) | A kind of information determines method and apparatus | |
| TW201737128A (en) | Data control method and system | |
| Prasasti et al. | Customer lifetime value and defection possibility prediction model using machine learning: An application to a cloud-based software company | |
| CN109902178A (en) | A kind of multistage file classification method and system | |
| CN110008415A (en) | A kind of data object variation tendency determines method, apparatus and server | |
| CN113794727B (en) | Threat information feature library generation method, threat information feature library generation device, storage medium and processor | |
| CN110008403B (en) | Target information sorting method, sorting system, recommendation method and recommendation system | |
| CN104731819A (en) | Advertisement keyword extraction method and advertisement keyword extraction device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |