CN109040852A - A kind of barrage screening technique and device - Google Patents
A kind of barrage screening technique and device Download PDFInfo
- Publication number
- CN109040852A CN109040852A CN201810769274.5A CN201810769274A CN109040852A CN 109040852 A CN109040852 A CN 109040852A CN 201810769274 A CN201810769274 A CN 201810769274A CN 109040852 A CN109040852 A CN 109040852A
- Authority
- CN
- China
- Prior art keywords
- barrage
- preset
- accounting
- degree
- feature
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/488—Data services, e.g. news ticker
- H04N21/4884—Data services, e.g. news ticker for displaying subtitles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application provides a kind of barrage screening technique and devices, extract the feature of the multi-angle of barrage, and it is ranked up based on feature, the barrage for being unsatisfactory for preset condition is deleted from ranking results again, the barrage filtered out, compared with the mode of existing random selection barrage, the barrage of high quality can be obtained.
Description
Technical field
This application involves electronic information field more particularly to a kind of barrage screening techniques and device.
Background technique
With the development of video website and Video Applications, the quantity using the user of video website and Video Applications increasingly increases
It is more.The barrage that user delivers during watching video, is increased with explosive speed.
Currently, video website selects some barrages from the barrage of magnanimity at random, it is shown in video playing.It is this
The higher barrage of quality targetedly can not be pushed to user by mode.
Summary of the invention
Applicant has found that the barrage of high quality has at least one following characteristics: being conducive to improve in the course of the study
Interactivity, the emotion of user is positive, semantic content is high, freshness is high and the excellent degree of evaluation plot is high.
Based on above-mentioned discovery, this application provides a kind of barrage screening technique and devices, it is therefore intended that solution can not be by matter
Measure the problem of higher barrage is pushed to user.
To achieve the goals above, this application provides following technical schemes:
A kind of barrage screening technique includes:
Extract the feature of barrage, the feature include user's interaction category feature, positive emotion degree, semantic content characteristics and
At least one of in freshness, excellent degree;User's interaction category feature is for characterizing the barrage to user's level of interaction
Contribution, the forward direction emotion degree are used to characterize the degree of the positive emotion of the expression of the barrage, the semanteme content characteristics
For characterizing the specific gravity of effective content in the barrage, what the freshness was used to characterize the barrage delivered day apart from the same day
Duration, the excellent degree of video of the excellent degree for characterizing the barrage evaluation;
Using the feature of the barrage, the barrage is ranked up, ranking results are obtained;
Delete the barrage that preset condition is unsatisfactory in the ranking results, the barrage after being screened.
Optionally, user's interaction category feature includes:
Role's barrage degree, role's barrage degree of any one barrage are as follows: the ratio of molecule and denominator, described point
Mother is the total quantity of role's barrage in N barrage, and in the case that this barrage is role's barrage, the molecule is the first numerical value,
Otherwise, the molecule is second value, and first numerical value is greater than the second value.
Optionally, user's interaction category feature further include:
Degree is thumbed up, any one the described of barrage thumbs up degree are as follows: what this barrage obtained thumbs up number and N barrage
In every barrage obtain the ratio for thumbing up the sum of number.
Optionally, the positive emotion degree includes:
The positive viewpoint word content of positive viewpoint word content, any one barrage is pre- according to include in this barrage
If positive viewpoint word and preset negative sense viewpoint word quantity difference determine.
Optionally, the positive viewpoint word content of any one barrage is seen according to the forward direction for including in this barrage
The difference determination of quantity of point word and negative sense viewpoint word includes:
The positive viewpoint word content of any one barrage are as follows:
Optionally, the semantic content characteristics include at least one of the following:
Text size accounting, Chinese character accounting, noun accounting, verb accounting;
The text size accounting of any one barrage are as follows: the number for the character for including in this barrage and N barrage
The ratio of the sum of the character number for including;The Chinese character accounting of any one barrage are as follows: in including in this barrage
The ratio of the sum of the number of all characters that the number and this barrage of Chinese character include;The noun of any one barrage accounts for
Than are as follows: the ratio of the sum of number of all participles for including in the number and this barrage of the noun for including in this barrage;Appoint
It anticipates the noun accounting an of barrage are as follows: all points for including in the number and this barrage of the verb for including in this barrage
The ratio of the sum of the number of word.
Optionally, in each day it's overdue and N barrage that the freshness of any one barrage is delivered according to this barrage
The ratio of the sum of each day it's overdue of every barrage determine that described each day it's overdue is that barrage delivers number of days of the day apart from the same day.
Optionally, the freshness of any one barrage are as follows:
Optionally, the excellent degree includes:
The excellent degree of the plot of barrage evaluation, the excellent degree of the plot are as follows: i-th second barrage amount and preset time
The ratio of barrage amount in range, the preset time range include described i-th second.
Optionally, the preset condition includes that barrage length meets preset length, Chinese character accounting meets preset account for
Than, not with other barrages repeat at least one of;
Described to delete the barrage that preset condition is unsatisfactory in the barrage sequence, the barrage after being screened includes:
It is deleted from the ranking results and is unsatisfactory for the barrage of the first preset condition, obtain the first the selection result, described the
One preset condition include: barrage length meet preset length, Chinese character accounting meet in preset accounting at least one of;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is similar
Degree is greater than a barrage in two barrages of preset similarity threshold.
A kind of barrage screening technique, comprising:
With the preset period, the feature of storage barrage is obtained using method above-mentioned, and use the spy of the storage barrage
Sign, obtains the first ranking results of the storage barrage;The storage barrage is stored history barrage;
In the case where receiving newly-increased barrage, the feature of the newly-increased barrage, and foundation are obtained using method above-mentioned
The feature of the newly-increased barrage obtains the second ranking results of the newly-increased barrage;
First ranking results and the barrage in second ranking results are ranked up according to the feature, are obtained
Third ranking results;
Delete the barrage that preset condition is unsatisfactory in the third ranking results, the barrage after being screened.
Optionally, the preset condition includes that barrage length meets preset length, Chinese character accounting meets preset account for
Than, not with other barrages repeat at least one of;
Described to delete the barrage that preset condition is unsatisfactory in the third ranking results, the barrage after being screened includes:
The barrage for being unsatisfactory for the first preset condition is deleted from the third ranking results, obtains the first the selection result, institute
State the first preset condition include: barrage length meet preset length, Chinese character accounting meet in preset accounting at least one
?;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is similar
Degree is greater than a barrage in two barrages of preset similarity threshold.
A kind of barrage screening plant, comprising:
Characteristic extracting module, for extracting the feature of barrage, the feature includes user's interaction category feature, positive emotion journey
At least one of in degree, semantic content characteristics and freshness, excellent degree;User's interaction category feature is for characterizing the barrage
Contribution to user's level of interaction, the forward direction emotion degree are used to characterize the degree of the positive emotion of the expression of the barrage,
The semanteme content characteristics are used to characterize the specific gravity of effective content in the barrage, and the freshness is for characterizing the barrage
Deliver the duration of time gap current time, the excellent degree of video of the excellent degree for characterizing the barrage evaluation;
Sorting module is ranked up the barrage, obtains ranking results for using the feature of the barrage;
Removing module, the barrage for deleting the barrage for being unsatisfactory for preset condition in the ranking results, after being screened.
Optionally, user's interaction category feature includes:
Role's barrage degree, role's barrage degree of any one barrage are as follows: the ratio of molecule and denominator, described point
Mother is the total quantity of role's barrage in N barrage, and in the case that this barrage is role's barrage, the molecule is the first numerical value,
Otherwise, the molecule is second value, and first numerical value is greater than the second value.
Optionally, user's interaction category feature further include:
Degree is thumbed up, any one the described of barrage thumbs up degree are as follows: what this barrage obtained thumbs up number and N barrage
In every barrage obtain the ratio for thumbing up the sum of number.
Optionally, the positive emotion degree includes:
The positive viewpoint word content of positive viewpoint word content, any one barrage is pre- according to include in this barrage
If positive viewpoint word and preset negative sense viewpoint word quantity difference determine.
Optionally, the positive viewpoint word content of any one barrage is seen according to the forward direction for including in this barrage
The difference determination of quantity of point word and negative sense viewpoint word includes:
The positive viewpoint word content of any one barrage are as follows:
Optionally, the semantic content characteristics include at least one of the following:
Text size accounting, Chinese character accounting, noun accounting, verb accounting;
The text size accounting of any one barrage are as follows: the number for the character for including in this barrage and N barrage
The ratio of the sum of the character number for including;The Chinese character accounting of any one barrage are as follows: in including in this barrage
The ratio of the sum of the number of all characters that the number and this barrage of Chinese character include;The noun of any one barrage accounts for
Than are as follows: the ratio of the sum of number of all participles for including in the number and this barrage of the noun for including in this barrage;Appoint
It anticipates the noun accounting an of barrage are as follows: all points for including in the number and this barrage of the verb for including in this barrage
The ratio of the sum of the number of word.
Optionally, in each day it's overdue and N barrage that the freshness of any one barrage is delivered according to this barrage
The ratio of the sum of each day it's overdue of every barrage determine that described each day it's overdue is that barrage delivers number of days of the day apart from the same day.
Optionally, the freshness of any one barrage are as follows:
Optionally, the excellent degree includes:
The excellent degree of the plot of barrage evaluation, the excellent degree of the plot are as follows: i-th second barrage amount and preset time
The ratio of barrage amount in range, the preset time range include described i-th second.
Optionally, the preset condition includes that barrage length meets preset length, Chinese character accounting meets preset account for
Than, not with other barrages repeat at least one of;
The removing module is for deleting the barrage for being unsatisfactory for preset condition in the barrage sequence, the bullet after being screened
Curtain include:
The removing module is specifically used for, and the barrage for being unsatisfactory for the first preset condition is deleted from the ranking results, is obtained
To the first the selection result, first preset condition includes: that barrage length meets preset length, Chinese character accounting satisfaction is preset
Accounting at least one of;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is similar
Degree is greater than a barrage in two barrages of preset similarity threshold.
A kind of barrage screening plant, comprising:
Screening module, the screening module are device above-mentioned;
Control module, for obtaining the feature of storage barrage using the screening module, and use institute with the preset period
The feature for stating storage barrage obtains the first ranking results of the storage barrage;The storage barrage is stored history bullet
Curtain;And in the case where receiving newly-increased barrage, the feature of the newly-increased barrage is obtained using the screening module, and according to institute
The feature for stating newly-increased barrage obtains the second ranking results of the newly-increased barrage;By first ranking results and described second
Barrage in ranking results is ranked up according to the feature, obtains third ranking results;It deletes in the third ranking results
It is unsatisfactory for the barrage of preset condition, the barrage after being screened.
Optionally, the preset condition includes that barrage length meets preset length, Chinese character accounting meets preset account for
Than, not with other barrages repeat at least one of;
The control module is for deleting the barrage for being unsatisfactory for preset condition in the third ranking results, after obtaining screening
Barrage include:
The control module is specifically used for, and the bullet for being unsatisfactory for the first preset condition is deleted from the third ranking results
Curtain, obtains the first the selection result, and first preset condition includes: that barrage length meets preset length, Chinese character accounting expires
At least one of in the preset accounting of foot;The repetition barrage in first the selection result is deleted, the second the selection result, institute are obtained
Stating repetition barrage is the barrage that similarity is greater than in two barrages of preset similarity threshold.
Barrage screening technique and device described herein are extracted the feature of the multi-angle of barrage, and are carried out based on feature
Sequence, then the barrage for being unsatisfactory for preset condition, the barrage filtered out, with existing random selection are deleted from ranking results
The mode of barrage is compared, and the barrage of high quality can be obtained.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of flow chart of barrage screening technique disclosed in the embodiment of the present application;
Fig. 2 is the flow chart of another barrage screening technique disclosed in the embodiment of the present application;
Fig. 3 is a kind of structural schematic diagram of barrage screening plant disclosed in the embodiment of the present application;
Fig. 4 is a kind of structural schematic diagram of barrage screening plant disclosed in the embodiment of the present application.
Specific embodiment
Barrage screening technique disclosed in the embodiment of the present application can be applied on the server of video website, further,
It can apply on the barrage server of video website, for screening the higher barrage of mass from the barrage of magnanimity.
In the present embodiment the embodiment described, the higher barrage of quality refers to the barrage for meeting following any point: having
Conducive to improving, the interactivity of user, emotion forward direction, semantic content is high, freshness is high and the excellent degree height of the plot of evaluation.
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on
Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall in the protection scope of this application.
Fig. 1 is a kind of barrage screening technique disclosed in the embodiment of the present application, comprising the following steps:
S101: barrage is pre-processed, and removal is unsatisfactory for the barrage of preset condition.
Preset condition includes that non-advertisement barrage and/or smoothness degree are greater than threshold value.Wherein, clear and coherent degree refers to the clear and coherent journey of sentence
Degree, in general, the clear and coherent degree of Chinese can be expressed as following values according to the clear and coherent degree of sentence: 1,2,3,4,5,1 indicates very not
Smoothness, 5 indicate very clear and coherent, and smoothness degree gradually increases from 1 to 5.
Based on above-mentioned preset condition, specifically, pretreatment includes that purposes of commercial detection and smoothness degree detect, wherein purposes of commercial detection
Purpose be, detect advertisement barrage.The purpose of clear and coherent degree detection is, determines the clear and coherent degree of barrage.The detection of clear and coherent degree,
The numerical value for the clear and coherent degree that can be provided according to user for sample, training pattern, so that the model after training has automatic evaluation
The function of clear and coherent degree.Threshold value can be empirically arranged, for example, threshold value is 3.
Pretreated purpose is that removal advertisement barrage is not more than the barrage of threshold value with smoothness degree.The threshold value of clear and coherent degree can be with
It presets.
In other words, S101 is equivalent to the anti-spam processing of barrage.Therefore, existing anti-spam service remote can be called
Invocation of procedure agreement (Remote Procedure Call Protocol, RPC) interface carries out detection and the bullet of advertisement barrage
The detection of curtain smoothness degree.After obtaining testing result, removal advertisement barrage is not more than the barrage of threshold value with smoothness degree.
It should be noted that S101 is optional step.The detection of specific advertisement barrage and the detection of barrage smoothness degree
Algorithm (setting principle of the threshold value including clear and coherent degree), may refer to the prior art, which is not described herein again.
S102: the feature of pretreated barrage is extracted.
In the present embodiment, 9 features are proposed, it is therefore intended that more comprehensively evaluate the quality of barrage.Assuming that for extracting
The barrage of feature is the N barrage extracted from one section of video at interval of 5 minutes, then the feature for the barrage for needing to extract includes:
1, role's barrage degree:
Wherein, for any one barrage in N barrage, if it is role's barrage, then molecule takes the first numerical value,
Such as 1, otherwise, molecule takes second value, such as 0, and denominator is the total quantity of role's barrage in N barrage.
In the present embodiment, role's barrage refers to, is directed to the barrage of the certain role input in video.Specifically, user
In the playing process of video, some personage's (or other targets) occurred in video can be clicked, for the personage of click
Then the barrage is delivered for the particular persons (or other targets) in video to (perhaps other targets) input barrage, therefore,
Belong to role's barrage.
As it can be seen that compared to common barrage, deliver role's barrage, user and video website need stronger level of interaction,
That is role's barrage is larger to the contribution of user's level of interaction.
2, degree is thumbed up:
Wherein, for any one barrage in N barrage, molecule thumbs up number for what this barrage obtained, point
Mother is in N barrage, what every barrage obtained thumbs up the sum of number.
For example, the number that thumbs up that number is the acquisition of the 1, the 2nd article of barrage that thumbs up that the 1st article of barrage obtains is the 2, the 3rd article of barrage
The number that thumbs up obtained is 2, then the 1st article of barrage thumbs up degree
Therefore the interaction that barrage needs user and video website is thumbed up, so, degree is thumbed up for reflecting the user of barrage
Level of interaction, thumbing up for barrage acquisition is several more, and the interaction times for illustrating that user is done based on barrage are more, and barrage hands over user
The contribution of mutual degree is bigger.
Role's barrage degree and the degree that thumbs up can be divided into user's interaction category feature.In the present embodiment, user's interaction
Category feature can only include role's barrage degree, and the degree that thumbs up is optional feature.Certainly, role's barrage degree and to thumb up degree equal
It extracts, it is more comprehensive for evaluation of the barrage for the contribution of user's level of interaction.
3, positive viewpoint word content:
Wherein, for any one barrage in N barrage, the positive viewpoint word quantity in denominator is this barrage
In include positive viewpoint word quantity, negative sense viewpoint word quantity be this barrage in include negative sense viewpoint word quantity.
It should be noted that positive viewpoint word and negative sense viewpoint word can be preset, that is, preset positive viewpoint word
Set and negative sense viewpoint set of words belong to the word of positive viewpoint set of words for positive viewpoint word, belong to negative sense viewpoint set of words
Word is to belong to negative sense viewpoint word.Specifically, during determining the positive viewpoint word content of certain barrage, first by this barrage
In word respectively compared with the word in positive viewpoint set of words and negative sense viewpoint set of words, if a word and positive viewpoint word set
A word in conjunction is identical, then the word is positive viewpoint word, if a word is identical as a word in negative sense viewpoint set of words,
Then the word is negative sense viewpoint word.
Positive viewpoint word degree reflects the positive emotion degree of barrage, and positive viewpoint word degree value is bigger, illustrates barrage
The positive emotion degree of reflection is bigger.
4, text size accounting:
Wherein, for any one barrage in N barrage, molecule is for the character for including in this barrage
Number, denominator are the sum of the number of character for including in N barrage in every barrage.
For example, the number that the number for the character for including in the 1st article of barrage is the character for including in the 10, the 2nd article of barrage is 20,
The number for the character for including in 3rd article of barrage is 20, then the text size accounting of the 1st article of barrage is
5, Chinese character accounting:
Wherein, for any one barrage in N barrage, molecule is the Chinese character for including in this barrage
Number, a Chinese character are considered as a Chinese character.Denominator is the sum of the number for all characters for including, institute in this barrage
Having character includes: Chinese character and outer Chinese character etc..The type of character can be preset.
6, noun accounting:
Wherein, for any one barrage in N barrage, molecule is for the noun for including in this barrage
Number, denominator are the sum of the number for all participles for including in this barrage, and participle includes noun, verb and adjective.
Specifically, word segmentation processing first can be carried out to this barrage, word segmentation result is obtained, word segmentation result includes forming this
All participles of barrage.Word segmentation processing algorithm may refer to the prior art, and which is not described herein again.
7, verb accounting:
Compared with noun accounting, molecule is the number of verb, and which is not described herein again.
Above-mentioned text size accounting, Chinese character accounting, noun accounting, verb accounting can be used as the semantic content of barrage
Feature, for characterizing the specific gravity of effective content in barrage.Above-mentioned semanteme content characteristics can choose at least one, the semanteme of selection
Content characteristics are more, then more comprehensive for the evaluation of the specific gravity of effective content.
8, freshness:
Wherein, for any one barrage in N barrage, molecule delivered day apart from the same day for this barrage
Number of days, denominator deliver number of days the sum of of the day apart from the same day for every barrage in N barrage.
For example, the 1st article is delivered for barrage 2018-3-1 days by user, the 2nd article is delivered for barrage 2018-3-4 days by user, and is counted
The same day for calculating freshness is 2018-3-4, then the freshness of the 1st article of barrage
As can be seen that expired time is longer, then freshness is lower.
It should be noted that calculated by taking number of days as an example in the calculation formula of the above freshness, other than number of days,
Barrage such as delivers the moment, delivers the moon at the other times unit, can substitute number of days, as the chronomere for calculating freshness.Generally
For including, freshness is used to characterize the duration for delivering time gap current time of barrage.
9, the excellent degree of plot:
The excellent degree of the plot are as follows: the ratio of the barrage amount in i-th second barrage amount and preset time range is preset
Time range includes described i-th second.The excellent degree of plot is used to indicate the density of barrage, and the density of barrage is bigger, illustrates play
The barrage that feelings receive when playing is more, then it represents that the plot of barrage evaluation is more excellent.Time range, which is here preset at, with 300 seconds is
Example, but be not limited to 300 seconds.
In conclusion for one aspect, user's interaction category feature for extracting barrage in above-mentioned 9 features respectively, just
It is more comprehensive to emotion degree, semantic content characteristics and freshness and excellent degree feature, the evaluation angle of feature.
For another party, in above-mentioned 9 features, feature 3,5,6 and 7 is based on a barrage and extracts, and is considered as absolutely
To feature, feature 1,2,4 and 8,9 be barrage (or a unit time barrage) and a plurality of barrage comparison result, can be with
Regard relative characteristic as.Either absolute feature or relative characteristic can reflect the quality of barrage.It include absolute in feature
Feature includes relative characteristic again, can more fully reflect the quality of barrage.
S103: using the feature of the barrage extracted, the mass fraction of barrage is calculated.
In the present embodiment, classical Rogers's base of a fruit function is deformed to obtain such as minor function:
The domain of deformed function is (0 ,+∞), and codomain is (0,1).
Wherein, the weighted sum of all features is to be added again after the feature of the barrage of said extracted and respective multiplied by weight
Obtained result.
The weight of different characteristic can be preset, and the weight of a feature is bigger, illustrate the contribution for mass fraction
It is bigger.
S104: barrage is screened.
The foundation of screening can be the mass fraction of barrage.Specifically, can according to barrage mass fraction to barrage into
Row sequence, by the preceding preset quantity barrage of ranking results, as the selection result.Alternatively, mass fraction is greater than preset threshold
Barrage, as the selection result.
Alternatively, in addition to the mass fraction of barrage, the foundation of screening can also include: barrage length, Chinese character accounting and
Multiplicity.Specifically, after being ranked up according to mass fraction, then delete barrage length from ranking results and be unsatisfactory for preset length
Barrage and Chinese character accounting be unsatisfactory for the barrage of preset accounting.
Wherein, barrage length refers to the quantity of the significant character in barrage, and significant character is the intention that can express user
Character, significant character can be preset, for example, using at least five Chinese character as the standard of barrage length, even if this barrage
The enough height of mass fraction, or sequence is enough forward, also to be deleted, cannot function as the barrage being selected.
The foregoing feature 7 of Chinese character accounting, here as individual screening foundation.
Further, after being screened according to barrage length, Chinese character accounting and mass fraction, weight can also be used
Multiplicity is screened.
Multiplicity refers to, the similarity degree with other barrages, and cosine phase can be used in the similarity degree between two barrages
It is calculated like degree, reference can be made to the prior art, which is not described herein again.
If the similarity of two barrages so far is greater than preset similarity threshold, to deleting a wherein barrage.Into
One step, deleting foundation can be any one in mass fraction, barrage length and Chinese character accounting.For example, deleting matter
Measure the lower barrage of score.
It can be seen that the feature for extracting barrage, and the quality of the feature calculation barrage according to extraction from process shown in FIG. 1
Score, then be ranked up according to mass fraction, barrage is finally filtered out from sequence.With the side of existing random selection barrage
Formula has higher readability, that is, has higher quality according to the barrage that process shown in FIG. 1 filters out.
In order to further increase for the screening effect of barrage, it can be based on process shown in FIG. 1, for already present bullet
Curtain (the hereinafter referred to as storage barrage) process flow different with the barrage newly increased (hereinafter referred to as newly-increased barrage) progress.
Fig. 2 is another barrage screening technique disclosed in the embodiment of the present application, comprising the following steps:
S201: with the preset period, storage barrage is pre-processed, feature extraction, calculates mass fraction, and according to matter
Amount score is ranked up, and obtains the first ranking results.
Wherein, pretreatment, feature extraction and the concrete mode for calculating mass fraction, may refer to step shown in FIG. 1,
Which is not described herein again.
In the present embodiment, the preset period can be arranged as unit of the moon, for example, one month or two months etc..
S202: in the case where receiving newly-increased barrage, the execution of newly-increased barrage is pre-processed, feature extraction, calculating
Mass fraction, and be ranked up according to mass fraction, obtain the second ranking results.
Newly-increased barrage is inputted by user.
S203: by the first ranking results and the second ranking results barrage according to based on feature obtain mass fraction into
Row sequence, obtains third ranking results.
That is, after the first ranking results are mixed with the barrage in the second ranking results, then sort, sequence
According to the mass fraction for being barrage.
S204: according to barrage length, Chinese character accounting and multiplicity, barrage is screened from third ranking results.
Specifically, from third ranking results or in third ranking results before sequence in preset quantity barrage, selection
Barrage length is greater than preset length out and Chinese character accounting is greater than the barrage of preset ratio, and removes weight in the barrage selected
Multiplicity is greater than the lower barrage of mass fraction in the barrage of preset threshold, thus the barrage filtered out.
For example, being the period with one month, storage barrage is ranked up, obtain and stores the first sequence as a result, right in real time
Newly-increased barrage is ranked up, and obtains the second ranking results.After obtaining the second ranking results, stored first sequence knot is obtained
Barrage in first ranking results and the second ranking results is mixed and is ranked up, obtains third ranking results, then to third by fruit
Ranking results are screened.
During shown in Fig. 2, the sequence of different frequency is carried out for storage barrage and newly-increased barrage, is realizing newly-increased bullet
In the case where the real-time screening of curtain, it can also guarantee the screening for storage barrage, and storage barrage and newly-increased barrage is common
It is further screened, so that the quality of the barrage filtered out is higher, after barrage push, is able to ascend the friendship of user
Mutual inductance.
Fig. 3 be the embodiment of the present application disclosed in a kind of barrage screening plant, comprising: characteristic extracting module, sorting module and
Removing module.
Wherein, characteristic extracting module is used to extract the feature of barrage, and the feature of extraction may refer to above method embodiment,
Which is not described herein again.Sorting module is used for the feature using the barrage, is ranked up to the barrage, obtains ranking results.
For removing module for deleting the barrage for being unsatisfactory for preset condition in the ranking results, the barrage after being screened is described default
Condition includes that barrage length meets preset length, Chinese character accounting meets preset accounting, not and in the repetition of other barrages
At least one of.
The specific implementation of the function of the above modules, may refer to above method embodiment, and which is not described herein again.
Barrage screening plant shown in Fig. 3, the feature of the barrage based on different angle are ranked up barrage, then to row
Sequence result is screened again, compared with the mode of existing random selection barrage, can filter out the barrage of high quality.
Fig. 4 is a kind of barrage screening plant disclosed in the embodiment of the present application, comprising: screening module and control module.
Wherein, screening module is barrage screening plant shown in Fig. 3.
Control module was used for the preset period, obtained the feature of storage barrage using the screening module, and use institute
The feature for stating storage barrage obtains the first ranking results of the storage barrage;The storage barrage is stored history bullet
Curtain;And in the case where receiving newly-increased barrage, the feature of the newly-increased barrage is obtained using the screening module, and according to institute
The feature for stating newly-increased barrage obtains the second ranking results of the newly-increased barrage;By first ranking results and described second
Barrage in ranking results is ranked up according to the feature, obtains third ranking results;It deletes in the third ranking results
Be unsatisfactory for the barrage of preset condition, the barrage after being screened, the preset condition include barrage length meet preset length, in
Chinese character accounting meet preset accounting, not with other barrages repeat at least one of.
Specifically, screening module may include characteristic extracting module and sequence mould in barrage screening plant shown in Fig. 3
Block, control module may include the deletion mould in clock controller, hybrid-sorting module and barrage screening plant shown in Fig. 3
Block.Wherein, clock controller extracts feature and sequence to storage barrage with period 1 triggering screening module, obtains the first sequence
As a result, and in the case where receiving newly-increased barrage, triggering screening module extracts feature and sequence to newly-increased barrage, obtains second
Ranking results.Hybrid-sorting module arranges the barrage in the first ranking results and the second ranking results according to the feature
Sequence obtains third ranking results.Removing module deletes the barrage that preset condition is unsatisfactory in the third ranking results.
Barrage screening plant shown in Fig. 4 can be ranked up storage barrage and newly-increased barrage with different frequencies, and
Barrage after comprehensive sieve series sequence, to obtain the barrage of high quality.
If function described in the embodiment of the present application method is realized in the form of SFU software functional unit and as independent production
Product when selling or using, can store in a storage medium readable by a compute device.Based on this understanding, the application is real
The part for applying a part that contributes to existing technology or the technical solution can be embodied in the form of software products,
The software product is stored in a storage medium, including some instructions are used so that a calculating equipment (can be personal meter
Calculation machine, server, mobile computing device or network equipment etc.) execute each embodiment the method for the application whole or portion
Step by step.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), with
Machine accesses various Jie that can store program code such as memory (RAM, Random Access Memory), magnetic or disk
Matter.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other
The difference of embodiment, same or similar part may refer to each other between each embodiment.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the application
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (24)
1. a kind of barrage screening technique characterized by comprising
Extract the feature of barrage, the feature includes user's interaction category feature, positive emotion degree, semantic content characteristics and fresh
At least one of in degree, excellent degree;User's interaction category feature is used to characterize contribution of the barrage to user's level of interaction,
The forward direction emotion degree is used to characterize the degree of the positive emotion of the expression of the barrage, and the semanteme content characteristics are used for table
Levy the specific gravity of effective content in the barrage, what the freshness was used to characterizing the barrage delivers time gap current time
Duration, the excellent degree of video of the excellent degree for characterizing the barrage evaluation;
Using the feature of the barrage, the barrage is ranked up, ranking results are obtained;
Delete the barrage that preset condition is unsatisfactory in the ranking results, the barrage after being screened.
2. the method according to claim 1, wherein user interaction category feature includes:
Role's barrage degree, role's barrage degree of any one barrage are as follows: the ratio of molecule and denominator, the denominator are
The total quantity of role's barrage in N barrage, in the case that this barrage is role's barrage, the molecule is the first numerical value, otherwise,
The molecule is second value, and first numerical value is greater than the second value.
3. method according to claim 1 or 2, which is characterized in that user's interaction category feature further include:
Degree is thumbed up, any one the described of barrage thumbs up degree are as follows: what this barrage obtained thumbs up every in number and N barrage
The ratio for thumbing up the sum of number that barrage obtains.
4. the method according to claim 1, wherein the forward direction emotion degree includes:
The positive viewpoint word content of positive viewpoint word content, any one barrage is preset according to include in this barrage
The difference of the quantity of positive viewpoint word and preset negative sense viewpoint word determines.
5. according to the method described in claim 4, it is characterized in that, the positive viewpoint word content of any one barrage
The difference determination of quantity according to the positive viewpoint word and negative sense viewpoint word that include in this barrage includes:
The positive viewpoint word content of any one barrage are as follows:
6. the method according to claim 1, wherein the semanteme content characteristics include at least one of the following:
Text size accounting, Chinese character accounting, noun accounting, verb accounting;
The text size accounting of any one barrage are as follows: the number for the character for including in this barrage includes with N barrage
The sum of character number ratio;The Chinese character accounting of any one barrage are as follows: the middle text for including in this barrage
The ratio of the sum of the number of all characters that the number and this barrage of symbol include;The noun accounting of any one barrage
Are as follows: the ratio of the sum of number of all participles for including in the number and this barrage of the noun for including in this barrage;Arbitrarily
The noun accounting of one barrage are as follows: all participles for including in the number and this barrage of the verb for including in this barrage
The sum of number ratio.
7. the method according to claim 1, wherein the freshness of any one barrage is according to this barrage
The ratio of the sum of each day it's overdue of every barrage in each day it's overdue and N barrage delivered determines that described each day it's overdue is bullet
Curtain delivers number of days of the day apart from the same day.
8. the method according to the description of claim 7 is characterized in that the freshness of any one barrage are as follows:
9. the method according to claim 1, wherein the excellent degree includes:
The excellent degree of the plot of barrage evaluation, the excellent degree of the plot are as follows: i-th second barrage amount and preset time range
The ratio of interior barrage amount, the preset time range include described i-th second.
10. according to claim 1, method described in 2,4,6,7,8 or 9, which is characterized in that the preset condition includes that barrage is long
Degree meet preset length, Chinese character accounting meets preset accounting, not with other barrages repeat at least one of;
Described to delete the barrage that preset condition is unsatisfactory in the barrage sequence, the barrage after being screened includes:
The barrage for being unsatisfactory for the first preset condition is deleted from the ranking results, obtains the first the selection result, and described first is pre-
If condition include: barrage length meet preset length, Chinese character accounting meet in preset accounting at least one of;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is that similarity is big
A barrage in two barrages of preset similarity threshold.
11. a kind of barrage screening technique characterized by comprising
With the preset period, the feature of storage barrage is obtained using the described in any item methods of claim 1-10, and use institute
The feature for stating storage barrage obtains the first ranking results of the storage barrage;The storage barrage is stored history bullet
Curtain;
In the case where receiving newly-increased barrage, the newly-increased barrage is obtained using the described in any item methods of claim 1-10
Feature, and the feature according to the newly-increased barrage obtains the second ranking results of the newly-increased barrage;
First ranking results and the barrage in second ranking results are ranked up according to the feature, obtain third
Ranking results;
Delete the barrage that preset condition is unsatisfactory in the third ranking results, the barrage after being screened.
12. according to the method for claim 11, which is characterized in that the preset condition includes that barrage length meets default length
Degree, Chinese character accounting meet preset accounting, not with other barrages repeat at least one of;
Described to delete the barrage that preset condition is unsatisfactory in the third ranking results, the barrage after being screened includes:
It is deleted from the third ranking results and is unsatisfactory for the barrage of the first preset condition, obtain the first the selection result, described the
One preset condition include: barrage length meet preset length, Chinese character accounting meet in preset accounting at least one of;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is that similarity is big
A barrage in two barrages of preset similarity threshold.
13. a kind of barrage screening plant characterized by comprising
Characteristic extracting module, for extracting the feature of barrage, the feature include user's interaction category feature, positive emotion degree,
At least one of in semantic content characteristics and freshness, excellent degree;User's interaction category feature is for characterizing the barrage pair
The contribution of user's level of interaction, the forward direction emotion degree are used to characterize the degree of the positive emotion of the expression of the barrage, institute
Predicate justice content characteristics are used to characterize the specific gravity of effective content in the barrage, and the freshness is used to characterize the hair of the barrage
The duration of table time gap current time, the excellent degree of video of the excellent degree for characterizing the barrage evaluation;
Sorting module is ranked up the barrage, obtains ranking results for using the feature of the barrage;
Removing module, the barrage for deleting the barrage for being unsatisfactory for preset condition in the ranking results, after being screened.
14. device according to claim 13, which is characterized in that the user interacts category feature and includes:
Role's barrage degree, role's barrage degree of any one barrage are as follows: the ratio of molecule and denominator, the denominator are
The total quantity of role's barrage in N barrage, in the case that this barrage is role's barrage, the molecule is the first numerical value, otherwise,
The molecule is second value, and first numerical value is greater than the second value.
15. device described in 3 or 14 according to claim 1, which is characterized in that user's interaction category feature further include:
Degree is thumbed up, any one the described of barrage thumbs up degree are as follows: what this barrage obtained thumbs up every in number and N barrage
The ratio for thumbing up the sum of number that barrage obtains.
16. device according to claim 13, which is characterized in that it is described forward direction emotion degree include:
The positive viewpoint word content of positive viewpoint word content, any one barrage is preset according to include in this barrage
The difference of the quantity of positive viewpoint word and preset negative sense viewpoint word determines.
17. device according to claim 16, which is characterized in that the positive viewpoint word of any one barrage contains
It measures and includes: according to the difference determination of the quantity of the positive viewpoint word and negative sense viewpoint word that include in this barrage
The positive viewpoint word content of any one barrage are as follows:
18. device according to claim 13, which is characterized in that the semanteme content characteristics include at least one of the following:
Text size accounting, Chinese character accounting, noun accounting, verb accounting;
The text size accounting of any one barrage are as follows: the number for the character for including in this barrage includes with N barrage
The sum of character number ratio;The Chinese character accounting of any one barrage are as follows: the middle text for including in this barrage
The ratio of the sum of the number of all characters that the number and this barrage of symbol include;The noun accounting of any one barrage
Are as follows: the ratio of the sum of number of all participles for including in the number and this barrage of the noun for including in this barrage;Arbitrarily
The noun accounting of one barrage are as follows: all participles for including in the number and this barrage of the verb for including in this barrage
The sum of number ratio.
19. device according to claim 13, which is characterized in that the freshness of any one barrage is according to this bar bullet
Each day it's overdue for delivering of curtain and the ratio of the sum of each day it's overdue of every barrage in N barrage determines, described each day it's overdue is
Barrage delivers number of days of the day apart from the same day.
20. device according to claim 19, which is characterized in that the freshness of any one barrage are as follows:
21. device according to claim 13, which is characterized in that it is described it is excellent degree include:
The excellent degree of the plot of barrage evaluation, the excellent degree of the plot are as follows: i-th second barrage amount and preset time range
The ratio of interior barrage amount, the preset time range include described i-th second.
22. device described in 3,14,16,18,19,20 or 21 according to claim 1, which is characterized in that the preset condition packet
Include barrage length meets preset length, Chinese character accounting meets preset accounting, not with other barrages repeat at least one
?;
The removing module is for deleting the barrage for being unsatisfactory for preset condition in the barrage sequence, the barrage packet after being screened
It includes:
The removing module is specifically used for, and deletes from the ranking results and is unsatisfactory for the barrage of the first preset condition, obtains the
One the selection result, first preset condition include: that barrage length meets preset length, Chinese character accounting meets preset account for
At least one of than in;
The repetition barrage in first the selection result is deleted, obtains the second the selection result, the barrage that repeats is that similarity is big
A barrage in two barrages of preset similarity threshold.
23. a kind of barrage screening plant characterized by comprising
Screening module, the screening module are the described in any item devices of claim 13-22;
Control module, for obtaining the feature of storage barrage using the screening module, and deposit using described with the preset period
The feature for measuring barrage, obtains the first ranking results of the storage barrage;The storage barrage is stored history barrage;And
In the case where receiving newly-increased barrage, the feature of the newly-increased barrage is obtained using the screening module, and according to described new
The feature for increasing barrage, obtains the second ranking results of the newly-increased barrage;By first ranking results and second sequence
As a result the barrage in is ranked up according to the feature, obtains third ranking results;It deletes in the third ranking results and is discontented with
The barrage of sufficient preset condition, the barrage after being screened.
24. according to claim 23 described device, which is characterized in that the preset condition includes that barrage length meets default length
Degree, Chinese character accounting meet preset accounting, not with other barrages repeat at least one of;
The control module is for deleting the barrage for being unsatisfactory for preset condition in the third ranking results, the bullet after being screened
Curtain include:
The control module is specifically used for, and the barrage for being unsatisfactory for the first preset condition is deleted from the third ranking results, is obtained
To the first the selection result, first preset condition includes: that barrage length meets preset length, Chinese character accounting satisfaction is preset
Accounting at least one of;The repetition barrage in first the selection result is deleted, the second the selection result, the repetition are obtained
Barrage is the barrage that similarity is greater than in two barrages of preset similarity threshold.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810769274.5A CN109040852A (en) | 2018-07-13 | 2018-07-13 | A kind of barrage screening technique and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810769274.5A CN109040852A (en) | 2018-07-13 | 2018-07-13 | A kind of barrage screening technique and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109040852A true CN109040852A (en) | 2018-12-18 |
Family
ID=64641028
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810769274.5A Pending CN109040852A (en) | 2018-07-13 | 2018-07-13 | A kind of barrage screening technique and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109040852A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109413484A (en) * | 2018-12-29 | 2019-03-01 | 咪咕文化科技有限公司 | A kind of barrage methods of exhibiting, device and storage medium |
CN110113673A (en) * | 2019-04-30 | 2019-08-09 | 北京奇艺世纪科技有限公司 | A kind of barrage display methods, device and electronic equipment |
CN110166847A (en) * | 2019-06-27 | 2019-08-23 | 北京奇艺世纪科技有限公司 | Barrage treating method and apparatus |
CN111212328A (en) * | 2019-12-31 | 2020-05-29 | 咪咕互动娱乐有限公司 | Bullet screen display method, bullet screen server and computer readable storage medium |
CN113691856A (en) * | 2021-08-17 | 2021-11-23 | 维沃移动通信(杭州)有限公司 | Bullet screen generation method and device, electronic equipment and readable storage medium |
CN114173198A (en) * | 2021-12-08 | 2022-03-11 | 湖南快乐阳光互动娱乐传媒有限公司 | Barrage rendering method and device, storage medium and electronic equipment |
CN114296604A (en) * | 2021-12-14 | 2022-04-08 | 咪咕音乐有限公司 | Bullet screen information display method, device, equipment and medium |
CN115022729A (en) * | 2022-06-02 | 2022-09-06 | 上海哔哩哔哩科技有限公司 | Bullet screen playing method and device |
CN115150663A (en) * | 2022-07-01 | 2022-10-04 | 北京奇艺世纪科技有限公司 | Heat curve generation method and device, electronic equipment and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1154358A2 (en) * | 2000-05-12 | 2001-11-14 | Applied Psychology Research Limited | Automatic text classification system |
CN105898375A (en) * | 2016-07-01 | 2016-08-24 | 上海幻电信息科技有限公司 | Barrage pushing system and method |
CN105916057A (en) * | 2016-04-18 | 2016-08-31 | 乐视控股(北京)有限公司 | Video barrage display method and device |
CN106096050A (en) * | 2016-06-29 | 2016-11-09 | 乐视控股(北京)有限公司 | A kind of method and apparatus of video contents search |
CN106101863A (en) * | 2016-06-24 | 2016-11-09 | 武汉斗鱼网络科技有限公司 | A kind of have barrage sending method and the device deleting protection mechanism |
CN106648283A (en) * | 2015-11-02 | 2017-05-10 | 腾讯科技(北京)有限公司 | Pop-up screen information processing method and server |
-
2018
- 2018-07-13 CN CN201810769274.5A patent/CN109040852A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1154358A2 (en) * | 2000-05-12 | 2001-11-14 | Applied Psychology Research Limited | Automatic text classification system |
CN106648283A (en) * | 2015-11-02 | 2017-05-10 | 腾讯科技(北京)有限公司 | Pop-up screen information processing method and server |
CN105916057A (en) * | 2016-04-18 | 2016-08-31 | 乐视控股(北京)有限公司 | Video barrage display method and device |
CN106101863A (en) * | 2016-06-24 | 2016-11-09 | 武汉斗鱼网络科技有限公司 | A kind of have barrage sending method and the device deleting protection mechanism |
CN106096050A (en) * | 2016-06-29 | 2016-11-09 | 乐视控股(北京)有限公司 | A kind of method and apparatus of video contents search |
CN105898375A (en) * | 2016-07-01 | 2016-08-24 | 上海幻电信息科技有限公司 | Barrage pushing system and method |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109413484A (en) * | 2018-12-29 | 2019-03-01 | 咪咕文化科技有限公司 | A kind of barrage methods of exhibiting, device and storage medium |
CN110113673A (en) * | 2019-04-30 | 2019-08-09 | 北京奇艺世纪科技有限公司 | A kind of barrage display methods, device and electronic equipment |
CN110166847A (en) * | 2019-06-27 | 2019-08-23 | 北京奇艺世纪科技有限公司 | Barrage treating method and apparatus |
CN110166847B (en) * | 2019-06-27 | 2021-07-23 | 北京奇艺世纪科技有限公司 | Bullet screen processing method and device |
CN111212328B (en) * | 2019-12-31 | 2022-03-25 | 咪咕互动娱乐有限公司 | Bullet screen display method, bullet screen server and computer readable storage medium |
CN111212328A (en) * | 2019-12-31 | 2020-05-29 | 咪咕互动娱乐有限公司 | Bullet screen display method, bullet screen server and computer readable storage medium |
CN113691856A (en) * | 2021-08-17 | 2021-11-23 | 维沃移动通信(杭州)有限公司 | Bullet screen generation method and device, electronic equipment and readable storage medium |
CN114173198A (en) * | 2021-12-08 | 2022-03-11 | 湖南快乐阳光互动娱乐传媒有限公司 | Barrage rendering method and device, storage medium and electronic equipment |
CN114173198B (en) * | 2021-12-08 | 2024-04-19 | 湖南快乐阳光互动娱乐传媒有限公司 | Barrage rendering method and device, storage medium and electronic equipment |
CN114296604A (en) * | 2021-12-14 | 2022-04-08 | 咪咕音乐有限公司 | Bullet screen information display method, device, equipment and medium |
CN115022729A (en) * | 2022-06-02 | 2022-09-06 | 上海哔哩哔哩科技有限公司 | Bullet screen playing method and device |
CN115150663A (en) * | 2022-07-01 | 2022-10-04 | 北京奇艺世纪科技有限公司 | Heat curve generation method and device, electronic equipment and storage medium |
CN115150663B (en) * | 2022-07-01 | 2023-12-15 | 北京奇艺世纪科技有限公司 | Heat curve generation method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109040852A (en) | A kind of barrage screening technique and device | |
CN107832437B (en) | Audio/video pushing method, device, equipment and storage medium | |
WO2021042826A1 (en) | Video playback completeness prediction method and apparatus | |
CN105138670B (en) | Audio file label generating method and system | |
CN105138558B (en) | The real time individual information collecting method of content is accessed based on user | |
CN106126524B (en) | Information pushing method and device | |
CN110019286B (en) | Expression recommendation method and device based on user social relationship | |
CN104317784A (en) | Cross-platform user identification method and cross-platform user identification system | |
CN105989074A (en) | Method and device for recommending cold start through mobile equipment information | |
CN106886518A (en) | A kind of method of microblog account classification | |
CN103377258A (en) | Method and device for classification display of microblog information | |
Redi et al. | Crowdsourcing for rating image aesthetic appeal: Better a paid or a volunteer crowd? | |
CN108415961A (en) | A kind of advertising pictures recommendation method and device | |
CN107578263A (en) | A kind of detection method, device and the electronic equipment of advertisement abnormal access | |
CN106601243A (en) | Video file identification method and device | |
CN103984741A (en) | Method and system for extracting user attribute information | |
CN110457566A (en) | Method, device, electronic equipment and storage medium | |
CN110737859A (en) | UP main matching method and device | |
CN108229994A (en) | A kind of information-pushing method and device | |
Sharifi et al. | Automatic summarization of twitter topics | |
CN105930532B (en) | A kind of method and apparatus from multimedia resource to user that recommending | |
CN112100221A (en) | Information recommendation method and device, recommendation server and storage medium | |
CN108021619A (en) | A kind of event description object recommendation method and device | |
CN104636470B (en) | A kind of method and device of recommendation business information | |
CN109635192A (en) | Magnanimity information temperature seniority among brothers and sisters update method and platform towards micro services |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |