CN116383480A - Method and device for recommending media in call scene, electronic equipment and storage medium - Google Patents
Method and device for recommending media in call scene, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116383480A CN116383480A CN202310054735.1A CN202310054735A CN116383480A CN 116383480 A CN116383480 A CN 116383480A CN 202310054735 A CN202310054735 A CN 202310054735A CN 116383480 A CN116383480 A CN 116383480A
- Authority
- CN
- China
- Prior art keywords
- user
- media
- data
- industry
- preprocessing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000007781 pre-processing Methods 0.000 claims abstract description 43
- 230000003068 static effect Effects 0.000 claims abstract description 31
- 238000003058 natural language processing Methods 0.000 claims abstract description 28
- 238000010801 machine learning Methods 0.000 claims abstract description 25
- 238000007621 cluster analysis Methods 0.000 claims abstract description 15
- 238000005516 engineering process Methods 0.000 claims abstract description 15
- 230000001960 triggered effect Effects 0.000 claims description 8
- 230000001360 synchronised effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000000737 periodic effect Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 3
- 230000006399 behavior Effects 0.000 description 16
- 230000001680 brushing effect Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 5
- 235000013305 food Nutrition 0.000 description 5
- 230000002093 peripheral effect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 241000219109 Citrullus Species 0.000 description 1
- 235000012828 Citrullus lanatus var citroides Nutrition 0.000 description 1
- 206010044565 Tremor Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Databases & Information Systems (AREA)
- Game Theory and Decision Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method, a device, electronic equipment and a storage medium for media recommendation in a call scene. The media recommendation method under the call scene comprises preprocessing and accurate matching, and specifically comprises the following steps: s11, preprocessing user data, wherein the preprocessing of the user data comprises setting of user static data, calculating of industry label scores of users by using an NLP (natural language processing) technology and setting of user dynamic data; s12, preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning; s21, accurately matching the pre-delivery media, and outputting an advertisement industry label weight result through NLP machine learning; s22, accurate matching of a user side is carried out, and a user industry label weight result is output; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played. According to the media recommendation method under the call scene, the advertisement media delivery precision and efficiency can be improved.
Description
Technical Field
The invention relates to the technical field of 5G wireless communication, big data AI and video early media negotiation, in particular to a method, a device, electronic equipment and a storage medium for displaying media recommendation in a call scene based on intelligent analysis of user data and video playing on a mobile phone when a user receives telephone ringing or calling.
Background
With the development of advertising industry, traditional advertisements such as televisions and signs are gradually unattractive due to the fact that audience numbers are reduced, effective effect monitoring is not available, large-area delivery is achieved, follow-up means are lacked, and the like; the media advertisement forms under the internet situation, such as WeChat, today's headline, microblog and the like, and the audio-video media delivery platforms such as tremble voice, fast hand, watermelon video and the like, can click to jump to own platform to change private due to the convenience of delivery, and have the characteristics of operation indexes and the like for quantitative analysis to be favored by more and more advertisers.
In the face of massive media channels, advertisers need to select proper popularization channels from the media channels according to their own budget and other requirements, just like a sea fishing needle. The prior method comprises the following steps: the method comprises the following steps: all the delivery is performed under the condition of sufficient budget; the second method is as follows: manually screening and screening according to the requirements of the user according to the category, the price, the reading quantity and the like; and a third method: repeatedly using well known media channels. Problems of the existing method are as follows: the delivery cost is high, malicious bill brushing, screen brushing software and automatic click software cause a plurality of invalid flows, and an advertiser is still required to buy the bill, so that additional propaganda cost is caused; a large amount of data are counterfeited, and through technical means, such as embedding an advertisement page into a user's necessary page, the browsing exposure is high, the subsequent operation analysis of an advertiser is influenced, and the analysis precision of enterprises is reduced.
The prior art has the following related patents: chinese patent application CN112752310a, name: a telephone video business card system based on media negotiation. The system can realize the service functions of calling triggering and called video media playing, and under the condition of insufficient coverage of the support terminal, the existing call service is not affected, and meanwhile, the transmission and the display of various calling number associated information can be realized. Chinese patent application CN111970408A, name: the method comprises the steps of sending a telephone call request to a video color ring server, wherein the telephone call request carries a called terminal identifier and playing configuration parameters of a peripheral terminal which is in communication connection with a calling terminal; receiving first video color ring data returned by the video color ring server according to the called terminal identification and the play configuration parameters; and sending the first video color ring data to the peripheral terminal for the peripheral terminal to play the first video color ring data. The method is applied to the telephone call process, and achieves the purpose of playing the video color ring on the peripheral terminal. Chinese patent application CN114266595a, name: the digital media elevator advertisement playing recommendation system based on the Internet comprises the steps of collecting passenger images in an elevator through an image collecting module, identifying whether passengers exist in the passenger images in the elevator, obtaining the number of people in the passenger images, and judging the recommendation type of the next advertisement of the elevator advertisement machine according to the number of people in the passenger images. Chinese patent application CN114254202a, name: the scheme is based on the construction of a labeling subsystem, a recommending subsystem and a storage subsystem, and by collecting user data and media asset library data, extracting labels from the user data and the media asset library data, performing similarity calculation on an extraction result by adopting a similarity algorithm model, generating recommended data according to the calculation result, sequencing the recommended data by adopting a sequencing algorithm, outputting a sequencing result and increasing the pertinence of recommendation.
The above patents still have the following problems: the advertisement media is not accurately put, a network user cannot locate an individual, whether the user information is accurate or not cannot be checked, and the uncertainty of subsequent operation analysis is increased; the coverage of the user group is insufficient; the media with cooperative experience is repeatedly used, the coverage of vermicelli is small, and the propagation influence is low; through internet channel delivery, the user population has limitations. Therefore, when the mobile phone terminal is used, a more targeted method for displaying the intelligent analysis of the central platform based on the user big data behaviors and targeted sending of ringing/color ring video media on the mobile phone terminal is needed to be developed.
Disclosure of Invention
The invention aims to solve the technical problem of precisely putting media on a mobile phone terminal of a user.
In order to solve the above technical problems, according to an aspect of the present invention, there is provided a method for recommending media in a call scenario, including preprocessing and exact matching, wherein the preprocessing is a setting and calculation performed in advance before the exact matching, and the preprocessing includes the following steps: s11, preprocessing user data, wherein the user data comprises: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user place, mobile phone model, consumption level and credit level data which can be obtained by an operator; the dynamic data comprises network search vocabulary carried out on the same day, number industry attribute of current conversation with the vocabulary, current conversation time, real-time position and new app type loaded on the same day; preprocessing user data comprises setting user static data, calculating industry label scores of users by using an NLP (natural language processing) technology, and setting user dynamic data; s12, preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning; wherein, the accurate matching includes the following steps: s21, accurately matching the pre-delivery media through NLP (non-linear projection) robotOutputting a label weight result of the advertising industry; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]，B＝[User i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; the front number of users is measured according to the number of media to be put required by the advertiser to put, and media 1 data is synchronized to a media library played by the next call of the users; s22, the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played.
According to an embodiment of the present invention, in step S11, the industry label score of the user is calculated, and the user may perform periodic processing by using NLP (natural language processing) technology, including the following steps: s111, acquiring a webpage recently browsed by the number by adopting a crawler mode or cooperation with an Internet company; word segmentation is carried out on the content acquired by the crawler; analyzing the field (Industry) to which the word frequency belongs through dictionary inquiry; calculating the industry label score Tab of each account i ，Tab i ＝[Ind i :x i ]The method comprises the steps of carrying out a first treatment on the surface of the S112, acquiring recently browsed media content of the number by adopting a crawler mode or cooperation with an Internet company, performing cluster analysis by adopting unsupervised machine learning, and calculating an industry label score Tab of each account according to the mode j ，Tab j ＝[Ind j :x j ]The method comprises the steps of carrying out a first treatment on the surface of the S113, weighting and averaging according to the sample quantity to obtain the industry label score Tab of each account p ，Tab p ＝[Ind p :x p ]. Wherein Ind represents industry domain, x represents score, and Tab represents corresponding setAnd (5) combining.
In step S12, in the process of pre-delivering media, the cluster analysis may be performed on the media content by using unsupervised machine learning, and the industry label score Tab of each account q ，Tab q ＝[Ind q :x q ]。
According to the embodiment of the present invention, in step S11, a dynamic data weight may be set to P (act); in step S22, when the condition based on the dynamic data triggers, the dynamic weight P (act) is compared with the industry weight x with the highest value max When the dynamic weight is greater than the industry weight, then playing the media advertisement triggered based on the dynamic data.
According to embodiments of the present invention, the recent period may be from 10 to 50 days, preferably from 20 to 40 days, and more preferably from 30 days.
According to an embodiment of the present invention, step S22 may further include: when the user clicks the media advertisement, a short message containing short chains is triggered when the user hangs up, so that the user can click the consultation or purchase the order. The information sent by the user after clicking the on-hook includes but is not limited to common text short messages, short messages containing H5 short chains, short messages containing micro-message small programs, intelligent short messages identified and upgraded by the terminal, multimedia short messages, terminal original video display and the like.
Further, step S22 may further include: and taking the clicking action of the user as the content of subsequent machine learning.
According to a second aspect of the present invention, there is provided an apparatus for media recommendation in a call scenario, comprising: the system comprises a preprocessing module and an accurate matching module, wherein the preprocessing module is used for preprocessing user data and pre-delivery media, the accurate matching module is used for accurately matching the pre-delivery media and a user side, the preprocessing is set and calculated in advance before the data are accurately matched, and the preprocessing comprises the following steps: preprocessing of user data, the user data comprising: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user place, mobile phone model, consumption level and credit level data which can be obtained by an operator; dynamic data includes a web for the daySearching words, and currently communicating with the words, namely, the attribute of the number industry, the current communication time, the real-time position and the new app type of the current day; preprocessing user data comprises setting user static data, calculating industry label scores of users by using an NLP (natural language processing) technology, and setting user dynamic data; preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning; wherein, the accurate matching includes: accurate matching of pre-delivery media is carried out, and an advertisement industry label weight result is output through NLP machine learning; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]，B＝[User i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; the front number of users is measured according to the number of media to be put required by the advertiser to put, and media 1 data is synchronized to a media library played by the next call of the users; the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played. Based on the call behavior, the method is not limited to the call behavior between terminals, and can be expanded to the communication behavior between internet apps, such as WeChat call, spike call and the like, as long as dynamic data such as call time, call objects, call positions and the like can be obtained.
According to a third aspect of the present invention, there is provided an electronic device comprising: the method comprises the steps of a memory, a processor and a media recommendation program in a call scene, wherein the media recommendation program is stored in the memory and can run on the processor, and the media recommendation method in the call scene is realized when the media recommendation program in the call scene is executed by the processor.
According to a fourth aspect of the present invention, there is provided a computer storage medium, wherein a media recommendation program under a call scene is stored on the computer storage medium, and when the media recommendation program under the call scene is executed by a processor, the steps of the media recommendation method under the call scene are implemented.
Compared with the prior art, the technical scheme provided by the embodiment of the invention at least has the following beneficial effects:
the invention provides a realization method for displaying a central platform on a mobile phone terminal, intelligently analyzing based on user big data behaviors and sending ringing/color ring video media in a targeted manner. According to the technical scheme of the method, not only can the AI intelligent customized video media be played and displayed on the supported network and terminal, but also the information resources which can be interacted with by the user can be provided, and the media resources are expanded. The method has the advantages of incapability of brushing, accurate matching, efficient recommendation and wide vermicelli coverage, and solves the problems in the prior art. The advertisement delivery object takes the mobile phone number as a unique identifier, and can be specific to the real property person or user of the number, and is not a network virtual account number any more.
The invention can analyze and accurately match the media advertisement based on the real user through the real data, thereby greatly reducing the invalid flow caused by screen brushing and flow brushing and lowering the delivery cost.
According to the invention, through the learning of multi-user data, the combination of conversation behavior and internet behavior is considered, and the advertisement media is put through comprehensively analyzing and precisely matching the static data, communication data, web searching and browsing data, mobile terminal app and other data of the user, so that the advertisement media putting precision and efficiency are improved.
The invention is based on the release of the conversation behavior, is very suitable for the business forms of the color ring and ringing, and has the universality for the receiving user group.
According to the scheme, not only is play and display of video media in a call scene considered, but also information resources which can be interacted with by a user are provided, and the released media resources are extended, including short messages, multimedia short messages, short chains, small programs, 5G messages and the like.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following brief description of the drawings of the embodiments will make it apparent that the drawings in the following description relate only to some embodiments of the present invention and are not limiting of the present invention.
Fig. 1 is a flowchart illustrating a media recommendation method in a call scenario according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without creative efforts, based on the described embodiments of the present invention fall within the protection scope of the present invention.
Unless defined otherwise, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this invention belongs. The terms "first," "second," and the like in the description and in the claims, are not used for any order, quantity, or importance, but are used for distinguishing between different elements. Likewise, the terms "a" or "an" and the like do not denote a limitation of quantity, but rather denote the presence of at least one.
Fig. 1 is a flowchart illustrating a media recommendation method in a call scenario according to an embodiment of the present invention.
As shown in fig. 1, the method for recommending media in a call scene comprises preprocessing and accurate matching.
The pretreatment is preset and calculated before the accurate matching. The pretreatment comprises the following steps:
s11, preprocessing user data, wherein the user data comprises: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user place, mobile phone model, consumption level and credit level data which can be obtained by an operator; the dynamic data comprises network search vocabulary carried out on the same day, number industry attribute of current conversation with the vocabulary, current conversation time, real-time position and new app type loaded on the same day; the preprocessing of the user data comprises setting of the user static data, calculating of the industry label score of the user by utilizing the NLP (natural language processing) technology and setting of the user dynamic data.
S12, preprocessing the pre-delivered media, extracting advertisement content, and performing cluster analysis on the media content to be delivered by adopting unsupervised machine learning.
The accurate matching comprises the following steps:
s21, accurately matching the pre-delivery media, and outputting an advertisement industry label weight result through NLP machine learning; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]Wherein b= [ User ] i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; and the front number of users is measured according to the number of media delivery required by the advertiser to deliver, and the media 1 data is synchronized to a media library played by the next call of the users.
S22, the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played.
The invention can analyze and accurately match the media advertisement based on the real user through the real data, thereby greatly reducing the invalid flow caused by screen brushing and flow brushing and lowering the delivery cost.
In step S11, according to one or some embodiments of the present invention, the industry label score of the user is calculated, and the periodic processing is performed using NLP (natural language processing) technology. Step S11 includes the steps of:
s111, acquiring a webpage recently browsed by the number by adopting a crawler mode or cooperation with an Internet company; word segmentation is carried out on the content acquired by the crawler; analyzing the field (Industry) to which the word frequency belongs through dictionary inquiry; calculating the industry label score Tab of each account i ，Tab i ＝[Ind i :x i ]。
S112, acquiring recently browsed media content of the number by adopting a crawler mode or cooperation with an Internet company, performing cluster analysis by adopting unsupervised machine learning, and calculating an industry label score Tab of each account according to the mode j ，Tab j ＝[Ind j :x j ]。
S113, weighting and averaging according to the sample quantity to obtain the industry label score Tab of each account p ，Tab p ＝[Ind p :x p ]。
Wherein Ind represents the industry domain, x represents the score, and Tab represents the corresponding set.
According to the invention, through the learning of multi-user data, the combination of conversation behavior and internet behavior is considered, and the advertisement media is put through comprehensively analyzing and accurately matching the static data, communication data, web searching and browsing data, mobile terminal app and other data of the user, so that the advertisement putting precision and efficiency are improved.
In step S12, in the process of pre-delivering media, the cluster analysis is performed on the media content by using the unsupervised machine learning, and the industry label score Tab of each account is performed according to one or more embodiments of the present invention q ，Tab q ＝[Ind q :x q ]。
According to one or some embodiments of the present invention, in step S11, a dynamic data weight is set to P (act); when the dynamic data-based bar is in step S22When the part is triggered, the dynamic weight P (act) is compared with the industry weight x with the highest value max When the dynamic weight is greater than the industry weight, then playing the media advertisement triggered based on the dynamic data.
According to one or some embodiments of the invention, the recent period may be from 10 to 50 days recently, preferably from 20 to 40 days recently, more preferably 30 days recently.
According to one or some embodiments of the present invention, step S22 further includes: when the user clicks the media advertisement, a short message containing short chains is triggered when the user hangs up, so that the user can click the consultation or purchase the order. The information sent by the user after clicking the on-hook includes but is not limited to common text short messages, short messages containing H5 short chains, short messages containing micro-message small programs, intelligent short messages identified and upgraded by the terminal, multimedia short messages, terminal original video display and the like. Further, step S22 further includes: and taking the clicking action of the user as the content of subsequent machine learning.
According to the scheme, not only is play and display of video media in a call scene considered, but also information resources which can be interacted with by a user are provided, and the released media resources are extended, including short messages, multimedia short messages, short chains, small programs, 5G messages and the like.
According to a second aspect of the present invention, there is provided an apparatus for media recommendation in a call scenario, comprising: the system comprises a preprocessing module and an accurate matching module, wherein the preprocessing module is used for preprocessing user data and pre-delivery media, the accurate matching module is used for accurately matching the pre-delivery media and a user side, the preprocessing is set and calculated in advance before the data are accurately matched, and the preprocessing comprises the following steps: preprocessing of user data, the user data comprising: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user place, mobile phone model, consumption level and credit level data which can be obtained by an operator; the dynamic data comprises network search vocabulary carried out on the same day, number industry attribute of current conversation with the vocabulary, current conversation time, real-time position and new app type loaded on the same day; preprocessing of user data includes setting of user static data, application of NLP (natural languageLanguage processing) to calculate the industry label score of the user and set the dynamic data of the user; preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning; wherein, the accurate matching includes: accurate matching of pre-delivery media is carried out, and an advertisement industry label weight result is output through NLP machine learning; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]，B＝[User i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; the front number of users is measured according to the number of media to be put required by the advertiser to put, and media 1 data is synchronized to a media library played by the next call of the users; the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played. Based on the call behavior, the method is not limited to the call behavior between terminals, and can be expanded to the communication behavior between internet apps, such as WeChat call, spike call and the like, as long as dynamic data such as call time, call objects, call positions and the like can be obtained.
The invention is based on the release of the conversation behavior, is very suitable for the business forms of the color ring and ringing, and has the universality for the receiving user group.
When in use, the user data is preprocessed, each telephone number is used as a unique identifier, and the user data is divided into static data, industry label scores and dynamic data according to the property rights of the numbers. Setting static data of a user: the static data of the user including the user age, sex, attribution province, user location, mobile phone model, consumption level, credit level, etc. can be obtained by the operator. And calculating the industry label score of the user. Periodic processing is performed using NLP (natural language processing) technology. The method comprises the following steps:
(1) A crawler mode or cooperation with an Internet company is adopted to acquire a webpage which is browsed recently (such as in the last 30 days) by the number; word segmentation is carried out on the content acquired by the crawler; through dictionary inquiry, the fields (industry labels of science and technology, automobiles, film and television, life, education, health, military, history, food and the like) of 100 words with highest word frequency are analyzed; calculating the industry label score of each account, such as { food 47%, automobile 24%, science and technology 14%, film and television 12%, education 8% … …, and other 0.5% };
(2) The media content recently browsed by the number (such as near 30 days) is acquired by adopting a crawler mode or cooperation with an Internet company, and cluster analysis is carried out by adopting unsupervised machine learning, and the industry label score of each account number is calculated according to the mode, such as { delicious food 47%, automobile 24%, science and technology 14%, film and television 12%, education 8% … … and other 0.5% }, wherein the method comprises the following steps of
(3) Weighted average according to sample size to obtain final industry label score of each account, such as { food 47%, automobile 24%, science and technology 14%, film and television 12%, education 8% … …, other 0.5% }
Setting dynamic data of a user. The dynamic data of the user comprises network search words carried out on the same day, number industry attributes of the current call with the user, current call time, real-time position, new app type of the same day and the like. Dynamic data weights, such as 99%, need to be set.
And preprocessing the pre-put media. The media content is subjected to cluster analysis by adopting unsupervised machine learning, and the industry label score of each account is calculated in the mode, such as { food 80%, science and technology 14%, life 4% … …, and other 0.5% }
The method for precisely matching the pre-delivery media comprises the following steps:
a) The pre-cast media 1 content is based on discrete user static data to generate a coarse cast resource pool set [ A ].
b) Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the coincidence values in descending order to obtain a number sequence, for example [ user 17 media 1-98.5%, user 2008 media 1-98.3%, user 323 media 1-98.2%, … … ]
c) According to the requirement of an advertiser, if the exposure is performed for 1 ten thousand times, the first 1 ten thousand users are taken and put in. And synchronizing the media 1 data to a media library played by the user in the next call.
The method for carrying out accurate matching on the user side comprises the following steps:
a) The media library played by the next call of the user 1 is arranged in a descending order according to the coincident values to obtain a plurality of columns, for example [ 8-98.5% of the user 1 media, 23-98.3% of the user 1 media, 299-98% of the user 1 media, … … ]
b) When the condition based on the dynamic data triggers, the dynamic weight is compared with the industry weight with the highest value of 98.5 percent, and when the dynamic weight is greater than the industry weight, the media advertisement based on the dynamic data triggers is played.
c) When the user clicks the media advertisement, a short message containing short chains is triggered when the user hangs up, so that the user can click the consultation or purchase the order.
d) The clicking behavior of the user serves as the content of the subsequent machine learning.
The invention provides a realization method for displaying a central platform on a mobile phone terminal, intelligently analyzing based on user big data behaviors and sending ringing/color ring video media in a targeted manner. According to the technical scheme of the method, not only can the AI intelligent customized video media be played and displayed on the supported network and terminal, but also the information resources which can be interacted with by the user can be provided, and the media resources are expanded. The method has the advantages of incapability of brushing, accurate matching, efficient recommendation and wide vermicelli coverage, and solves the problems in the prior art. The advertisement delivery object takes the mobile phone number as a unique identifier, and can be specific to the real property person or user of the number, and is not a network virtual account number any more.
According to still another aspect of the present invention, there is provided an apparatus for media recommendation in a call scenario, including: the method comprises the steps of a memory, a processor and a media recommendation program in a call scene, wherein the media recommendation program is stored in the memory and can run on the processor, and the media recommendation method in the call scene is realized when the media recommendation program in the call scene is executed by the processor.
There is also provided a computer storage medium according to the present invention.
The computer storage medium stores a media recommendation program under a call scene, and the media recommendation program under the call scene realizes the steps of the media recommendation method under the call scene when being executed by the processor.
The method implemented when the media recommendation program is executed in the call scenario running on the processor may refer to each embodiment of the media recommendation method in the call scenario of the present invention, which is not described herein again.
The invention also provides a computer program product.
The computer program product of the present invention comprises a media recommendation program in a talk scenario, which when executed by a processor implements the steps of the media recommendation method in a talk scenario as described above.
The method implemented when the media recommendation program is executed in the call scenario running on the processor may refer to each embodiment of the media recommendation method in the call scenario of the present invention, which is not described herein again.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above, comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing is merely exemplary embodiments of the present invention and is not intended to limit the scope of the invention, which is defined by the appended claims.
Claims (10)
1. A method for recommending media in a call scene comprises preprocessing and accurate matching,
the preprocessing is preset and calculated before accurate matching, and comprises the following steps:
s11, preprocessing user data, wherein the user data comprises: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user location, mobile phone model, consumption level and credit level data which can be obtained through an operator; the dynamic data comprises network search words carried out on the same day, number industry attributes of the current call with the words, current call time, real-time position and new app type of the same day; preprocessing user data comprises setting user static data, calculating industry label scores of users by using an NLP (natural language processing) technology, and setting user dynamic data;
s12, preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning;
wherein the exact matching comprises the steps of:
s21, accurately matching the pre-delivery media, and outputting an advertisement industry label weight result through NLP machine learning; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]，B＝[User i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; the user of the number ahead is selected according to the media delivery number required by the advertiser to deliver, and media 1 data is synchronized to the next call of the userA media library;
s22, the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played.
2. The method as claimed in claim 1, wherein in step S11, the industry label score of the user is calculated, and the periodic processing is performed by using NLP (natural language processing) technology, comprising the steps of:
s111, acquiring a webpage recently browsed by the number by adopting a crawler mode or cooperation with an Internet company; word segmentation is carried out on the content acquired by the crawler; analyzing the field (Industry) to which the word frequency belongs through dictionary inquiry; calculating the industry label score Tab of each account i ，Tab i ＝[Ind i :x i ]Wherein Ind i Represents the field of industry, x i Representing the score;
s112, acquiring recently browsed media content of the number by adopting a crawler mode or cooperation with an Internet company, performing cluster analysis by adopting unsupervised machine learning, and calculating an industry label score Tab of each account according to the mode j ，Tab j ＝[Ind j :x j ]Wherein Ind j Represents the field of industry, x j Representing the score, tab j Representing its corresponding set;
s113, weighting and averaging according to the sample quantity to obtain the industry label score Tab of each account p ，Tab p ＝[Ind p :x p ]Wherein Ind p Represents the field of industry, x p Representing the score, tab p Representing its corresponding set.
3. The method according to claim 2, wherein in step S12, the media is pre-delivered in the process of pre-delivering the mediaThe content adopts unsupervised machine learning to carry out cluster analysis, and the industry label score Tab of each account is obtained q ，Tab q ＝[Ind q :x q ]Wherein Ind q Represents the field of industry, x q Representing the score, tab q Representing its corresponding set.
4. The method according to claim 1, wherein the dynamic data weight is set to P (act) in step S11; in step S22, when the condition based on the dynamic data triggers, the dynamic weight P (act) is compared with the industry weight x with the highest value max When the dynamic weight is greater than the industry weight, then playing the media advertisement triggered based on the dynamic data.
5. The method of claim 2, wherein the recent period is from 10 to 50 days, preferably from 20 to 40 days, more preferably 30 days.
6. The method of claim 1, wherein step S22 further comprises:
when the user clicks the media advertisement, a short message containing short chains is triggered when the user hangs up, so that the user can click the consultation or purchase the order.
7. The method of claim 6, wherein step S22 further comprises:
and taking the clicking action of the user as the content of subsequent machine learning.
8. An apparatus for media recommendation in a talk scenario, comprising: a preprocessing module and an exact matching module, wherein the preprocessing module is used for preprocessing the user data and the pre-delivery media, the exact matching module is used for exactly matching the pre-delivery media and the user side,
the preprocessing is set and calculated in advance before the accurate matching of the data, and comprises the following steps:
preprocessing of user data, the user data comprising: static data, industry label scores, and dynamic data; the static data comprise user age, sex, attribution province, user location, mobile phone model, consumption level and credit level data which can be obtained through an operator; the dynamic data comprises network search words carried out on the same day, number industry attributes of the current call with the words, current call time, real-time position and new app type of the same day; preprocessing user data comprises setting user static data, calculating industry label scores of users by using an NLP (natural language processing) technology, and setting user dynamic data;
preprocessing the pre-put media, extracting advertisement content, and performing cluster analysis on the media content to be put by adopting unsupervised machine learning;
wherein the exact matching comprises:
accurate matching of pre-delivery media is carried out, and an advertisement industry label weight result is output through NLP machine learning; generating a coarse delivery resource pool set [ A ] for the pre-delivered media 1 content based on discrete user static data]The method comprises the steps of carrying out a first treatment on the surface of the Calculating the coincidence value of the industry label score of each user in the resource pool and the industry label score of the pre-cast media 1 content, and arranging the industry label scores in a descending order to generate a cast resource pool sequence [ B ]]，B＝[User i med 1 :x i ]Wherein User i med 1 Representing media 1 and each user, x i Representing coincident values; the front number of users is measured according to the number of media to be put required by the advertiser to put, and media 1 data is synchronized to a media library played by the next call of the users;
the accurate matching of the user side outputs the label weight result of the user industry, firstly, the media library played by the user 1 in the next call is arranged in descending order according to the coincident value to obtain a number sequence [ C ]]，C＝[Med i user 1 :x i ]Wherein Med i user 1 Representing user 1 and each media, x i Representing coincident values; when the condition based on the dynamic data triggers, then the media advertisement based on the dynamic data triggers is played.
9. An electronic device, comprising: memory, a processor and a talk scene media recommendation program stored on the memory and executable on the processor, which talk scene media recommendation program when executed by the processor implements the steps of the talk scene media recommendation method according to any of claims 1 to 7.
10. A computer storage medium having stored thereon a talk scene media recommendation program which when executed by a processor implements the steps of the talk scene media recommendation method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310054735.1A CN116383480A (en) | 2023-02-03 | 2023-02-03 | Method and device for recommending media in call scene, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310054735.1A CN116383480A (en) | 2023-02-03 | 2023-02-03 | Method and device for recommending media in call scene, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116383480A true CN116383480A (en) | 2023-07-04 |
Family
ID=86970032
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310054735.1A Pending CN116383480A (en) | 2023-02-03 | 2023-02-03 | Method and device for recommending media in call scene, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116383480A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117041229A (en) * | 2023-10-09 | 2023-11-10 | 吉林省气象服务中心(吉林省专业气象台、吉林省气象影视宣传中心) | Calling waiting weather multimedia playing system and method based on VoLTE technology |
| CN117590951A (en) * | 2024-01-18 | 2024-02-23 | 江西科技学院 | Multi-scene VR interaction method, system and storage medium |
-
2023
- 2023-02-03 CN CN202310054735.1A patent/CN116383480A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117041229A (en) * | 2023-10-09 | 2023-11-10 | 吉林省气象服务中心(吉林省专业气象台、吉林省气象影视宣传中心) | Calling waiting weather multimedia playing system and method based on VoLTE technology |
| CN117041229B (en) * | 2023-10-09 | 2023-12-15 | 吉林省气象服务中心(吉林省专业气象台、吉林省气象影视宣传中心) | Calling waiting weather multimedia playing system and method based on VoLTE technology |
| CN117590951A (en) * | 2024-01-18 | 2024-02-23 | 江西科技学院 | Multi-scene VR interaction method, system and storage medium |
| CN117590951B (en) * | 2024-01-18 | 2024-04-05 | 江西科技学院 | Multi-scene VR interaction method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108269128B (en) | Advertisement putting method, device, equipment and storage medium | |
| CN116383480A (en) | Method and device for recommending media in call scene, electronic equipment and storage medium | |
| JP5954730B2 (en) | Computer-implemented method, system, and program for implementing challenge-response test | |
| CN107464142B (en) | Method, the user's portrait system for system of being drawn a portrait based on marketing feedback result back feeding user | |
| US8060390B1 (en) | Computer based method for generating representative questions from an audience | |
| CN105450778B (en) | Information transmission system | |
| CN103582897A (en) | Displaying phone number on the landing page based on keywords | |
| US20230410144A1 (en) | Methods and systems for automatic call routing with no caller intervention using anonymous online user behavior | |
| WO2022252513A1 (en) | Target display method and apparatus | |
| US20200175545A1 (en) | Methods and systems for online to offline conversion tracking an optimization | |
| US11509610B2 (en) | Real-time messaging platform with enhanced privacy | |
| US8538958B2 (en) | System and method for context map generation | |
| CN106960033B (en) | Method and device for labeling information stream | |
| US20090265226A1 (en) | Methods and apparatus for interactive advertising | |
| RU105508U1 (en) | SOCIAL ADVERTISING SYSTEM | |
| US20230316106A1 (en) | Method and apparatus for training content recommendation model, device, and storage medium | |
| US11741364B2 (en) | Deep neural networks modeling | |
| CN116167803A (en) | Advertisement putting method and device based on signaling data | |
| CN111107439B (en) | Content distribution method, content distribution device, server and storage medium | |
| CN114442869A (en) | User distribution processing method and device, electronic equipment and storage medium | |
| CN114119030A (en) | Fraud prevention method and device, electronic equipment and storage medium | |
| CN112348570A (en) | Automatic advertisement configuration system and method | |
| CN111724182A (en) | Advertisement pushing method facing mobile terminal | |
| WO2023016085A1 (en) | Service data processing method and apparatus, electronic device, computer readable storage medium, and computer program product | |
| US11272259B1 (en) | Real-time bidding based system, method, and computer program for using in-video annotations to select relevant advertisements for distribution |
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 |