CN113781113B - Chained information pushing system and method - Google Patents
Chained information pushing system and method Download PDFInfo
- Publication number
- CN113781113B CN113781113B CN202111065554.6A CN202111065554A CN113781113B CN 113781113 B CN113781113 B CN 113781113B CN 202111065554 A CN202111065554 A CN 202111065554A CN 113781113 B CN113781113 B CN 113781113B
- Authority
- CN
- China
- Prior art keywords
- interface
- carrier
- interfaces
- analysis
- marking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- 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
-
- 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
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Abstract
The invention discloses a chain-type information pushing system and a chain-type information pushing method, relates to the technical field of information pushing, and is used for solving the problems that the traditional information pushing is too hard to push, the user experience is reduced, the information pushing link is too short, and the opportunity of further pushing effective information is lost; the system comprises an information pushing module, a data processing module and a recommendation module, wherein the information pushing module is used for receiving correlation parameters between carrier interfaces, acquiring a carrier interface where a user is currently located, identifying the carrier interface where the user is currently located and a browsing interface of the user in a database to obtain a next carrier interface of the user, and sending recommendation information to a user side of the user; the method analyzes and outputs the correlation parameters among the appointed carrier interfaces by using the trained artificial intelligence model, then obtains the next carrier interface of the user according to the interface data browsed by the user before, and obtains the recommendation information of the carrier interface to send to the user, thereby forming the chain information push.
Description
Technical Field
The invention relates to the technical field of information pushing, in particular to a chained information pushing system and a chained information pushing method.
Background
The traditional information pushing method is based on the analysis of user labels, and directly pushes information with high correlation degree, particularly advertisement information, to users, but the method has the defect that the method is too hard and the user experience is probably reduced. And the link of information push is too short, the next operation of the user receiving the push often breaks away from the original information carrier interface, and the opportunity of further pushing effective information is lost. In the actual browsing environment of the user, progressive continuous searching or clicking is often adopted to browse the content which is interested by the user, and the browsed content affects the following browsing tendency.
Therefore, a linked information pushing system and a linked information pushing method are needed, which effectively improve the browsing experience of the user, enable the user to continuously explore the interested information, and obviously prolong the staying time of the user in the link of the pushed information, thereby generating more obvious advertising effect for the advertiser.
Disclosure of Invention
The invention aims to provide a linked information pushing system and a linked information pushing method to solve the problems that the traditional information pushing is too hard to push, the user experience is reduced, the link of the information pushing is too short, and the opportunity of further pushing effective information is lost.
The purpose of the invention can be realized by the following technical scheme: a linkage type information pushing system comprises a server and at least one user side which is in communication connection with the server, wherein the server comprises a feature extraction module, a model analysis module and an information pushing module;
the characteristic extraction module is used for acquiring interface data of the carrier interface, extracting characteristic keywords of the carrier interface and sending the extracted characteristic keywords of the carrier interface to the model analysis module;
the model analysis module is used for acquiring the trained artificial intelligence model in the database, analyzing the received characteristic keywords of the carrier interfaces through the artificial intelligence model to output the correlation parameters between the carrier interfaces, and then sending the correlation parameters between the carrier interfaces to the information push module;
the information pushing module is used for receiving the association parameters between the carrier interfaces, acquiring the current carrier interface of the user, identifying the current carrier interface of the user and the browsing interface of the user in the database to obtain the next carrier interface of the user, acquiring the recommendation information corresponding to the carrier interface, and sending the recommendation information to the user side of the user;
as a preferred embodiment of the present invention, the specific process of the model analysis module performing analysis through the artificial intelligence model is as follows:
marking any carrier interface as an analysis interface, and marking the rest carrier interfaces as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number;
setting a plurality of part-of-speech categories; each part of speech category corresponds to a plurality of characteristic keywords, and the special keywords of the analysis interface are matched with the plurality of characteristic keywords corresponding to all the part of speech categories to obtain the part of speech categories to which the analysis interface belongs; matching the feature keywords of the comparison interface with a plurality of feature keywords corresponding to all part-of-speech categories to obtain the part-of-speech categories to which the comparison interface belongs; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight;
normalizing the word weight number and the sex weight number, taking the normalized values of the word weight number and the sex weight number, and obtaining a correlation value GL of the first interface or the second interface by using a formula GL of 0.783(MC1 multiplied by mu 1+ MX2 multiplied by mu 2+ 1.26); wherein MC1 is the numeric value after word weight number normalization, and μ 1 is the weight preset factor of the numeric value after word weight number normalization; MC2 is the value after normalization of the number of sex weights, and μ 2 is the weight preset factor of the value after normalization of the number of sex weights;
marking the first interface and the second interface as correlation interfaces of the analysis interface, marking the correlation values of the correlation interfaces as correlation parameters, and outputting the analysis interface, the corresponding correlation interfaces and the correlation parameters;
as a preferred embodiment of the present invention, the server further includes a data acquisition module, where the data acquisition module is configured to acquire browsing data of the user side and send the browsing data to the database, where the browsing data includes browsing interface information and operation information of the received recommendation information; the browsed interface information comprises the content of a browsing interface and the content of retrieval recommendation information; the operation information comprises the time when the recommendation information is not opened and is directly closed, the time when the recommendation information is opened and checked and the time when the recommendation information is closed;
as a preferred embodiment of the present invention, the server further includes a push optimization module, where the push optimization module is configured to optimize a recommended information push path, and the specific steps are as follows:
generating an optimization signaling of the recommendation information and sending the optimization signaling to a database, and after receiving the optimization signaling of the recommendation information, executing operation corresponding to the optimization signaling to obtain browsing data corresponding to the recommendation information and feeding back the browsing data;
analyzing the browsing data of the recommendation information after receiving the browsing data, wherein the specific analysis process is as follows: acquiring all carrier interfaces pushed by recommendation information correspondingly and marking as primary selection interfaces, and counting accessed contents of the recommendation information in the primary selection interfaces, wherein the accessed contents comprise the number of the recommendation information, the position of the recommendation information on the primary selection interfaces, the number of the accessed recommendation information, the access initial time, the access ending time and access operation, and the access operation comprises adjusting the font display size of the recommendation information; calculating the time difference between the access initial time and the access finishing time of the recommendation information to obtain access duration; setting all carriers to correspond to a font display group, wherein the font display group comprises a plurality of display font numbers, each display font number corresponds to a preset display value, the display font number is in inverse proportion to the preset display value, and the smaller the numerical value corresponding to the display font number is, the larger the preset display value is; matching the font display size of the recommendation information with the numbers of all display fonts in a font display group to which the corresponding carrier interface belongs to obtain a corresponding preset display value and marking the preset display value as an access display value; dividing a carrier interface into a plurality of display areas, wherein each display area corresponds to one display merit value; identifying the position of the recommendation information on the primary selection interface to obtain a display area where the recommendation information is located and a display optimal value corresponding to the display area; normalizing the visit duration, the visit display value and the display merit value, taking the normalized values of the visit duration, the visit display value and the display merit value, and respectively marking the normalized values of the visit duration, the visit display value and the display merit value as TSQ, FSQ and XSQ;
substitution formula
Obtaining a click-to-order value ED of the recommendation information on the primary selection interface; wherein λ 1, λ 2 and λ 3 are all preset weighting factors; summing all the single visit values of the recommendation information on the primary selection interface and dividing the sum by the total visit number of the primary selection interface to obtainThe visit rate of the recommended information on the primary selection interface is obtained; the total number of the access of the primary selection interface is the number of the user sides corresponding to the access primary selection interface in a time range between the time when the recommendation information is pushed to the primary selection interface and the current time;
sorting the primary selection interfaces from large to small according to the visiting rate, sequentially selecting a preset number of primary selection interfaces from front to back, and marking the selected primary selection interfaces as recommendation interfaces of recommendation information; the recommendation information is sequentially sent to a recommendation interface according to the size of the access rate, and the sequence of sending the recommendation information to the recommendation interface and the recommendation interface are marked as a push path after optimization of the recommendation information;
as a preferred embodiment of the present invention, the server further includes an interface analysis module, where the interface analysis module is configured to analyze browsing data of the user side, and specifically: acquiring data of a corresponding browsing interface of a user side and jumping to another browsing interface, and marking the other browsing interface as a jumping interface; counting the jumping times of the jumping interface, the initial time and the ending time of the jumping, and calculating the time difference between the initial time and the ending time of the jumping to obtain the browsing duration of the jumping interface; then extracting numerical values of the jumping times and the browsing duration of the jumping interface, and respectively marking the two numerical values as TZ1 and LL 1; obtaining a jump merit value XY1 of the jump interface by using a formula XY1 ═ TZ1 × d1+ LL1 × d 2; wherein d1 and d2 are both preset weight coefficients; marking the jump interface with the maximum jump goodness value as the next carrier interface corresponding to the user browsing interface and sending the next carrier interface to the database;
as a preferred embodiment of the present invention, the information push module is further configured to perform feature extraction on a marketing material in advertisement push information provided by an advertiser through an NLP technology to obtain feature keywords of the marketing material, and match the feature keywords of the marketing material with the feature keywords of a carrier interface to obtain a corresponding carrier interface, where the matched carrier interface is a carrier interface in which the number of coincided feature keywords and feature keywords of the marketing material is greater than one; counting the number of coincidences matched to the carrier interface and marking as CH 1; then acquiring the number of the associated interfaces matched with the carrier interface and the corresponding association value; marking the number of the associated interfaces as CH2, summing the association degree values of all the associated interfaces matched with the carrier interface to obtain an associated total value, and marking the associated total value as CH 3; extracting numerical values of the coincidence quantity CH1, the quantity CH2 and the total correlation value CH3 which are matched with the carrier interface, and acquiring corresponding preset weight coefficient fixed values b1, b2 and b 3; obtaining a broad-spectrum value GT of the matched carrier interface by using a formula GT of CH1 × b1+ CH2 × b2+ CH3 × b 3; sending the advertisement push information to a carrier interface with a push value larger than a preset push threshold value;
a chained information pushing method comprises the following steps:
s1: acquiring interface data of a carrier interface and extracting characteristic keywords of the carrier interface;
s2: the trained artificial intelligence model analyzes the characteristic keywords of the carrier interfaces to obtain the associated parameters between the carrier interfaces, and the specific analysis process is as follows:
s21: marking any carrier interface as an analysis interface, and marking the rest carrier interfaces as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number;
s22: setting a plurality of part-of-speech categories; each part of speech category corresponds to a plurality of characteristic keywords, and the special keywords of the analysis interface are matched with the characteristic keywords corresponding to all the part of speech categories to obtain the part of speech category of the analysis interface; matching the characteristic keywords of the comparison interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the comparison interface; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight;
s23: normalizing the word weight number and the sex weight number, taking the normalized values of the word weight number and the sex weight number, and obtaining a correlation value GL of the first interface or the second interface by using a formula GL of 0.783(MC1 multiplied by mu 1+ MX2 multiplied by mu 2+ 1.26); wherein MC1 is the numeric value after word weight number normalization, and μ 1 is the weight preset factor of the numeric value after word weight number normalization; MC2 is a numerical value after the weight normalization, and mu 2 is a weight preset factor of the numerical value after the weight normalization;
s24: marking the first interface and the second interface as correlation interfaces of the analysis interface, marking the correlation values of the correlation interfaces as correlation parameters, and outputting the analysis interface, the corresponding correlation interfaces and the correlation parameters;
s3: the method comprises the steps of obtaining a current carrier interface of a user, identifying the current carrier interface of the user and a browsing interface of the user in a database to obtain a next carrier interface of the user, obtaining recommendation information corresponding to the carrier interface, and sending the recommendation information to a user side of the user.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the method, the characteristic keywords of each carrier interface are extracted, the trained artificial intelligence model is used for analyzing and outputting the correlation parameters among the appointed carrier interfaces, then the next carrier interface of the user is obtained according to interface data browsed by the user before, the recommendation information of the carrier interface is obtained and sent to the user, and therefore chain type information pushing is formed.
2. The intelligent pushing path optimization is carried out on pushing through the pushing optimization module, and the pushing path is optimized by recombining the carrier interface appearance sequence of the initial pushing path so as to solve the problem that the pushing path formed by the interface initially set by the system is probably not the most suitable;
3. the invention extracts the characteristics of the marketing materials provided by the advertiser, performs matching analysis on the carrier interface and the pushing path, and mounts the advertisement pushing information of the advertiser to the most suitable carrier interface and the pushing path, so that the residence time of the user in the link of the pushing information is obviously prolonged, and the advertisement effect is more obvious for the advertiser.
Drawings
In order to facilitate understanding for those skilled in the art, the present invention will be further described with reference to the accompanying drawings.
Fig. 1 is a schematic block diagram of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a chain information push system includes a server and a plurality of clients communicatively connected to the server;
the server comprises a feature extraction module, a model analysis module, an information pushing module, a data acquisition module, a pushing optimization module and an interface analysis module;
the data acquisition module acquires browsing data of a user side and sends the browsing data to the database, wherein the browsing data comprises browsing interface information and operation information of received recommendation information; the browsed interface information comprises the content of a browsing interface and the content of retrieval recommendation information; the operation information comprises the time when the recommendation information is not opened and is directly closed, the time when the recommendation information is opened and checked and the time when the recommendation information is closed;
the characteristic extraction module collects interface data of the carrier interface, extracts characteristic keywords of the carrier interface and sends the extracted characteristic keywords of the carrier interface to the model analysis module; the carrier interface is a user side interface; the user side comprises intelligent equipment such as a smart phone, a tablet, a computer and the like;
the model analysis module obtains the trained artificial intelligence model in the database, and analyzes the characteristic keywords of the carrier interfaces through the artificial intelligence model to output the associated parameters between the carrier interfaces, and specifically comprises the following steps: marking any carrier interface as an analysis interface, and marking the rest carrier interfaces as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number;
setting a plurality of part-of-speech categories; each part of speech category corresponds to a plurality of characteristic keywords, and the special keywords of the analysis interface are matched with the characteristic keywords corresponding to all the part of speech categories to obtain the part of speech category of the analysis interface; matching the characteristic keywords of the comparison interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the comparison interface; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight;
normalizing the word weight number and the character weight number, taking the normalized values of the word weight number and the character weight number, and obtaining a correlation value GL of the first interface or the second interface by using a formula GL of 0.783(MC1 multiplied by mu 1+ MX2 multiplied by mu 2+ 1.26); wherein, MC1 is the value after the normalization of the word weight number, and μ 1 is the weight preset factor of the value after the normalization of the word weight number; MC2 is the value after normalization of the number of sex weights, and μ 2 is the weight preset factor of the value after normalization of the number of sex weights; the values of mu 1 and mu 2 are 0.72 and 0.28;
marking the first interface and the second interface as correlation interfaces of the analysis interface, marking the correlation values of the correlation interfaces as correlation parameters, outputting the analysis interface, the corresponding correlation interfaces and the correlation parameters, and finally sending the correlation parameters between the carrier interfaces to the information push module;
the information pushing module receives the association parameters between the carrier interfaces, acquires the current carrier interface of the user, identifies the current carrier interface of the user and the browsing interface of the user in the database to obtain the next carrier interface of the user, acquires the recommendation information corresponding to the carrier interface, and then sends the recommendation information to the user side of the user;
according to the method, the characteristic keywords of each carrier interface are extracted, the trained artificial intelligence model is used for analyzing and outputting the correlation parameters among the appointed carrier interfaces, then the next carrier interface of a user is obtained according to interface data browsed by the user before, the recommendation information of the carrier interface is obtained and sent to the user, and therefore chain type information pushing is formed;
the information pushing module extracts the characteristics of the marketing materials in the advertisement pushing information provided by the advertiser through an NLP technology to obtain the characteristic keywords of the marketing materials, and matches the characteristic keywords of the marketing materials with the characteristic keywords of the carrier interface to obtain a corresponding carrier interface, wherein the matched carrier interface is the carrier interface with the number of the coincidence of the characteristic keywords and the characteristic keywords of the marketing materials larger than one; counting the number of coincidences matched to the carrier interface and marking as CH 1; then acquiring the number of the associated interfaces matched with the carrier interface and the corresponding association value; marking the number of the associated interfaces as CH2, summing the association degree values of all the associated interfaces matched with the carrier interface to obtain an associated total value, and marking the associated total value as CH 3; extracting numerical values of the coincidence quantity CH1, the quantity CH2 and the total correlation value CH3 which are matched with the carrier interface, and acquiring corresponding preset weight coefficient fixed values b1, b2 and b 3; obtaining a broad-spectrum value GT of the matched carrier interface by using a formula GT of CH1 × b1+ CH2 × b2+ CH3 × b 3; sending the advertisement push information to a carrier interface with a push value larger than a preset push threshold value;
the invention utilizes NLP technology to extract the characteristics of the marketing materials provided by the advertiser, carries out matching analysis on the carrier interface and the pushing path, and mounts the advertisement pushing information of the advertiser to the most suitable carrier interface and the pushing path, so that the residence time of the user in the link of the pushing information is obviously prolonged, and the advertiser can generate more obvious advertisement effect;
the push optimization module optimizes a recommended information push path, and the specific optimization steps are as follows:
generating an optimization signaling of the recommendation information and sending the optimization signaling to a database, after receiving the optimization signaling of the recommendation information, analyzing the optimization signaling to obtain the recommendation information, matching browsing data corresponding to the recommendation information, compressing and packaging the browsing data, and feeding the browsing data back to a pushing optimization module;
decompressing and analyzing the compressed and packaged data packet of the browsing data of the recommendation information after receiving the compressed and packaged data packet, wherein the specific analysis process comprises the following steps:
acquiring all carrier interfaces pushed by recommendation information correspondingly and marking as primary selection interfaces, and counting accessed contents of the recommendation information in the primary selection interfaces, wherein the accessed contents comprise the number of the recommendation information, the position of the recommendation information on the primary selection interfaces, the number of the accessed recommendation information, the access initial time, the access ending time and access operation, and the access operation comprises adjusting the font display size of the recommendation information; calculating the time difference between the access initial time and the access finishing time of the recommendation information to obtain access duration;
setting all carriers to correspond to a font display group, wherein the font display group comprises a plurality of display font numbers, each display font number corresponds to a preset display value, the display font number is in inverse proportion to the preset display value, and the smaller the numerical value corresponding to the display font number is, the larger the preset display value is; if the preset display value corresponding to the character size 14 is smaller than the preset display value corresponding to the character size 10; matching the font display size of the recommendation information with the numbers of all display fonts in a font display group to which the corresponding carrier interface belongs to obtain a corresponding preset display value and marking the preset display value as an access display value; dividing a carrier interface into a plurality of display areas, wherein each display area corresponds to one display merit value; identifying the position of the recommendation information on the primary selection interface to obtain a display area where the recommendation information is located and a display optimal value corresponding to the display area; normalizing the visit duration, the visit display value and the display merit value, taking the normalized values of the visit duration, the visit display value and the display merit value, and respectively marking the normalized values of the visit duration, the visit display value and the display merit value as TSQ, FSQ and XSQ;
substituting into formula
Obtaining a click-to-order value ED of the recommendation information on the primary selection interface; wherein, λ 1, λ 2 and λ 3 are allIs a preset weight factor; the values of lambda 1, lambda 2 and lambda 3 are 0.37, 0.64 and 1.9;
summing all the single visit values of the recommendation information on the primary selection interface, and dividing the sum by the total visit number of the primary selection interface to obtain the visit rate of the recommendation information on the primary selection interface; the total number of the access of the primary selection interface is the number of the user terminals corresponding to the access primary selection interface in the time range between the time when the recommendation information is pushed to the primary selection interface and the current time;
sorting the primary selection interfaces from large to small according to the visiting rate, sequentially selecting a preset number of primary selection interfaces from front to back, and marking the selected primary selection interfaces as recommendation interfaces of recommendation information; the recommendation information is sequentially sent to a recommendation interface according to the size of the access rate, and the sequence of sending the recommendation information to the recommendation interface and the recommendation interface are marked as a push path after optimization of the recommendation information;
the intelligent pushing path optimization is carried out on pushing through the pushing optimization module so as to solve the problem that a pushing path formed by an interface initially set by a system is probably not the most suitable;
the interface analysis module analyzes the browsing data of the user side, acquires the browsing interface corresponding to the user side and data for jumping to another browsing interface, and marks the other browsing interface as a jumping interface; counting the jumping times of the jumping interface, the initial time and the ending time of the jumping, and calculating the time difference between the initial time and the ending time of the jumping to obtain the browsing duration of the jumping interface; then extracting numerical values of the jumping times and the browsing duration of the jumping interface, and respectively marking the two numerical values as TZ1 and LL 1; obtaining a jump merit value XY1 of the jump interface by using a formula XY1 ═ TZ1 × d1+ LL1 × d 2; wherein d1 and d2 are both preset weight coefficients; the values are 0.7 and 0.3 respectively; marking the jump interface with the maximum jump goodness value as the next carrier interface corresponding to the user browsing interface and sending the next carrier interface to the database;
when the system is used, the characteristic keywords of the carrier interface are extracted, the characteristic keywords of the carrier interface are analyzed through a trained artificial intelligence model, any one carrier interface is marked as an analysis interface, and the rest carrier interfaces are marked as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number; matching the special keywords of the analysis interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the analysis interface; matching the characteristic keywords of the comparison interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the comparison interface; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight; normalizing the word weight number and the sex weight number, taking the normalized values of the word weight number and the sex weight number, and obtaining a correlation value GL of the first interface or the second interface by using a formula GL of 0.783(MC1 multiplied by mu 1+ MX2 multiplied by mu 2+ 1.26); marking the first interface and the second interface as correlation interfaces of the analysis interface, marking the correlation values of the correlation interfaces as correlation parameters, and outputting the analysis interface, the corresponding correlation interfaces and the correlation parameters; the method comprises the steps of obtaining a current carrier interface of a user, identifying the current carrier interface of the user and a browsing interface of the user in a database to obtain a next carrier interface of the user, obtaining recommendation information corresponding to the carrier interface, and sending the recommendation information to a user side of the user.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise forms disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand the invention for and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims (7)
1. A linkage type information push system comprises a server and at least one user side which is in communication connection with the server, and is characterized in that the server comprises a feature extraction module, a model analysis module and an information push module;
the characteristic extraction module is used for acquiring interface data of the carrier interface, extracting characteristic keywords of the carrier interface and sending the extracted characteristic keywords of the carrier interface to the model analysis module;
the model analysis module is used for acquiring a trained artificial intelligence model in the database, analyzing the received characteristic keywords of the carrier interfaces through the artificial intelligence model to output correlation parameters among the carrier interfaces, and then sending the correlation parameters among the carrier interfaces to the information push module;
the information pushing module is used for receiving the association parameters between the carrier interfaces, acquiring the current carrier interface of the user, identifying the current carrier interface of the user and the browsing interface of the user in the database to obtain the next carrier interface of the user, acquiring the recommendation information corresponding to the carrier interface, and sending the recommendation information to the user side of the user.
2. The chained information push system according to claim 1, wherein the specific process of the model analysis module performing analysis through the artificial intelligence model is as follows:
marking any carrier interface as an analysis interface, and marking the rest carrier interfaces as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number;
setting a plurality of part-of-speech categories; each part of speech category corresponds to a plurality of characteristic keywords, and the special keywords of the analysis interface are matched with the plurality of characteristic keywords corresponding to all the part of speech categories to obtain the part of speech categories to which the analysis interface belongs; matching the characteristic keywords of the comparison interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the comparison interface; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight;
normalizing the word weight number and the sex weight number, taking the normalized values of the word weight number and the sex weight number, and obtaining a correlation value GL of the first interface or the second interface by using a formula GL of 0.783(MC1 multiplied by mu 1+ MX2 multiplied by mu 2+ 1.26); wherein MC1 is the numeric value after word weight number normalization, and μ 1 is the weight preset factor of the numeric value after word weight number normalization; MC2 is a numerical value after the weight normalization, and mu 2 is a weight preset factor of the numerical value after the weight normalization;
and marking the first interface and the second interface as association interfaces of the analysis interface, marking the association values of the association interfaces as association parameters, and outputting the analysis interface, the corresponding association interfaces and the corresponding association parameters.
3. The system of claim 1, wherein the server further comprises a data collection module, and the data collection module is configured to collect browsing data of the user side and send the browsing data to the database.
4. The system according to claim 3, wherein the server further comprises a push optimization module, the push optimization module is configured to optimize a recommended information push path, and the method specifically includes:
generating an optimization signaling of the recommendation information and sending the optimization signaling to a database, and after receiving the optimization signaling of the recommendation information, executing operation corresponding to the optimization signaling to obtain browsing data corresponding to the recommendation information and feeding back the browsing data;
analyzing the browsing data of the recommendation information after receiving the browsing data, wherein the specific analysis process is as follows: acquiring all carrier interfaces pushed by the recommendation information correspondingly and marking as a primary selection interface, counting accessed contents of the recommendation information in the primary selection interface, and calculating the time difference between the access initial time and the access finishing time of the recommendation information to obtain access duration; setting all carriers to correspond to a font display group, wherein the font display group comprises a plurality of display font numbers, and each display font number corresponds to a preset display value; matching the font display size of the recommendation information with all display font numbers in a font display group to which the corresponding carrier interface belongs to obtain a corresponding preset display value and marking the preset display value as an access display value; dividing a carrier interface into a plurality of display areas, wherein each display area corresponds to one display merit value; identifying the position of the recommendation information on the primary selection interface to obtain a display area where the recommendation information is located and a display optimal value corresponding to the display area; normalizing the visit duration, the visit display value and the display preference value, taking the normalized values of the visit duration, the visit display value and the display preference value, and analyzing the values to obtain a click-to-order visit value of the recommendation information on the primary selection interface; summing all the single visit values of the recommendation information on the primary selection interface, and dividing the sum by the total visit number of the primary selection interface to obtain the visit rate of the recommendation information on the primary selection interface;
sorting the primary selection interfaces from large to small according to the visiting rate, sequentially selecting a preset number of primary selection interfaces from front to back, and marking the selected primary selection interfaces as recommendation interfaces of recommendation information; and the recommendation information sequentially sends the recommendation information to a recommendation interface according to the size of the access rate, and the sequence of sending the recommendation information to the recommendation interface and the recommendation interface are marked as a push path after the recommendation information is optimized.
5. The system according to claim 3, wherein the server further comprises an interface analysis module, the interface analysis module is configured to analyze browsing data of the user side, and specifically, the interface analysis module is configured to: acquiring data of a corresponding browsing interface of a user side and jumping to another browsing interface, and marking the other browsing interface as a jumping interface; counting the jumping times of the jumping interface, the initial time and the ending time of the jumping, and calculating the time difference between the initial time and the ending time of the jumping to obtain the browsing duration of the jumping interface; then extracting values of the jumping times and the browsing duration of the jumping interface, and analyzing the two values to obtain a jumping optimal value of the jumping interface; and marking the jump interface with the maximum jump goodness value as the next carrier interface corresponding to the user browsing interface and sending the next carrier interface to the database.
6. The chain-type information push system according to claim 1, wherein the information push module is further configured to perform feature extraction on a marketing material in the advertisement push information provided by the advertiser through NLP technology to obtain feature keywords of the marketing material, and match the feature keywords of the marketing material with the feature keywords of the carrier interface to obtain a corresponding carrier interface, wherein the matched carrier interface is a carrier interface in which the number of coincidences of the feature keywords and the feature keywords of the marketing material is greater than one; counting the number of coincidences matched to the carrier interface; then acquiring the number of the associated interfaces matched with the carrier interface and the corresponding association degree value; then summing the correlation values of all the correlation interfaces matched with the carrier interface to obtain a total correlation value; extracting the numerical values of the number of coincidences, the number of the associated interfaces and the total associated value of the matched carrier interfaces, analyzing the numerical values to obtain the push value of the matched carrier interfaces, and sending the advertisement push information to the carrier interfaces with the push values larger than the preset push threshold value.
7. A chained information pushing method is characterized by comprising the following steps:
s1: acquiring interface data of a carrier interface and extracting characteristic keywords of the carrier interface;
s2: the trained artificial intelligence model analyzes the characteristic keywords of the carrier interfaces to obtain the associated parameters between the carrier interfaces, and the specific analysis process is as follows:
s21: marking any carrier interface as an analysis interface, and marking the rest carrier interfaces as comparison interfaces; comparing the feature keywords of the analysis interface with the feature keywords of the comparison interface, marking the comparison interface which is overlapped with the corresponding feature keywords of the analysis interface as a first interface, counting the number of the overlapped first interface and the special keywords of the analysis interface and marking the number as the word number;
s22: setting a plurality of part-of-speech categories; each part of speech category corresponds to a plurality of characteristic keywords, and the special keywords of the analysis interface are matched with the plurality of characteristic keywords corresponding to all the part of speech categories to obtain the part of speech categories to which the analysis interface belongs; matching the characteristic keywords of the comparison interface with a plurality of characteristic keywords corresponding to all the part-of-speech categories to obtain the part-of-speech categories of the comparison interface; comparing the part of speech category of the comparison interface with the part of speech category of the analysis interface; marking the comparison interface which is coincided with the part of speech category corresponding to the analysis interface as a second interface, counting the number of the second interface which is coincided with the part of speech category of the analysis interface and marking as the number of the part of speech weight;
s23: carrying out normalization processing on the word weight quantity and the character weight quantity, taking the values of the word weight quantity and the character weight quantity after normalization processing, and analyzing the values to obtain a correlation value of the first interface or the second interface;
s24: marking the first interface and the second interface as correlation interfaces of the analysis interface, marking the correlation values of the correlation interfaces as correlation parameters, and outputting the analysis interface, the corresponding correlation interfaces and the correlation parameters;
s3: the method comprises the steps of obtaining a current carrier interface of a user, identifying the current carrier interface of the user and a browsing interface of the user in a database to obtain a next carrier interface of the user, obtaining recommendation information corresponding to the carrier interface, and sending the recommendation information to a user side of the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111065554.6A CN113781113B (en) | 2021-09-09 | 2021-09-09 | Chained information pushing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111065554.6A CN113781113B (en) | 2021-09-09 | 2021-09-09 | Chained information pushing system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113781113A CN113781113A (en) | 2021-12-10 |
| CN113781113B true CN113781113B (en) | 2022-06-21 |
Family
ID=78843013
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111065554.6A Active CN113781113B (en) | 2021-09-09 | 2021-09-09 | Chained information pushing system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113781113B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116739675B (en) * | 2023-06-17 | 2024-02-02 | 壹禾文化控股集团(宁夏)有限公司 | Advertisement intelligent pushing method based on big data |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101141454A (en) * | 2007-08-20 | 2008-03-12 | 阳术国 | Video subscribing instrument system and method using the same |
| WO2017020451A1 (en) * | 2015-08-03 | 2017-02-09 | 百度在线网络技术（北京）有限公司 | Information push method and device |
| CN111507769A (en) * | 2020-04-20 | 2020-08-07 | 苏州跃盟信息科技有限公司 | Information pushing method and system, storage medium and processor |
| CN111881343A (en) * | 2020-07-07 | 2020-11-03 | Oppo广东移动通信有限公司 | Information pushing method and device, electronic equipment and computer readable storage medium |
| CN112288516A (en) * | 2020-10-21 | 2021-01-29 | 腾讯科技（深圳）有限公司 | Information recommendation method and device, storage medium and electronic equipment |
| CN112381578A (en) * | 2020-11-16 | 2021-02-19 | 南京邦峰智能科技有限公司 | Internet advertisement intelligent recommendation management system based on behavior characteristic recognition |
| CN112398947A (en) * | 2020-11-18 | 2021-02-23 | 腾讯科技（深圳）有限公司 | Information pushing method, device and equipment and computer readable storage medium |
| CN113239275A (en) * | 2021-05-28 | 2021-08-10 | 北京百度网讯科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150227972A1 (en) * | 2014-02-07 | 2015-08-13 | Huang (Joy) Tang | System and methods for identifying and promoting tagged commercial products |
| CN110286976B (en) * | 2019-05-24 | 2021-10-01 | 华为技术有限公司 | Interface display method, device, terminal and storage medium |
-
2021
- 2021-09-09 CN CN202111065554.6A patent/CN113781113B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101141454A (en) * | 2007-08-20 | 2008-03-12 | 阳术国 | Video subscribing instrument system and method using the same |
| WO2017020451A1 (en) * | 2015-08-03 | 2017-02-09 | 百度在线网络技术（北京）有限公司 | Information push method and device |
| CN111507769A (en) * | 2020-04-20 | 2020-08-07 | 苏州跃盟信息科技有限公司 | Information pushing method and system, storage medium and processor |
| CN111881343A (en) * | 2020-07-07 | 2020-11-03 | Oppo广东移动通信有限公司 | Information pushing method and device, electronic equipment and computer readable storage medium |
| CN112288516A (en) * | 2020-10-21 | 2021-01-29 | 腾讯科技（深圳）有限公司 | Information recommendation method and device, storage medium and electronic equipment |
| CN112381578A (en) * | 2020-11-16 | 2021-02-19 | 南京邦峰智能科技有限公司 | Internet advertisement intelligent recommendation management system based on behavior characteristic recognition |
| CN112398947A (en) * | 2020-11-18 | 2021-02-23 | 腾讯科技（深圳）有限公司 | Information pushing method, device and equipment and computer readable storage medium |
| CN113239275A (en) * | 2021-05-28 | 2021-08-10 | 北京百度网讯科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113781113A (en) | 2021-12-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110532451A (en) | Search method and device for policy text, storage medium, electronic device | |
| CN109767318A (en) | Loan product recommended method, device, equipment and storage medium | |
| CN105095187A (en) | Search intention identification method and device | |
| CN107943792B (en) | Statement analysis method and device, terminal device and storage medium | |
| CN103425691A (en) | Search method and search system | |
| CN112348602B (en) | Automatic advertisement putting management system based on big data | |
| CN103577432A (en) | Method and system for searching commodity information | |
| CN105989125B (en) | The searching method and system of tag recognition are carried out to no result word | |
| CN112258254B (en) | Internet advertisement risk monitoring method and system based on big data architecture | |
| CN112104642A (en) | Abnormal account number determination method and related device | |
| CN113781113B (en) | Chained information pushing system and method | |
| CN108446333B (en) | Big data text mining processing system and method thereof | |
| CN110134845A (en) | Project public sentiment monitoring method, device, computer equipment and storage medium | |
| CN116205222A (en) | Aspect-level emotion analysis system and method based on multichannel attention fusion | |
| CN114663132A (en) | Intelligent marketing method and device based on real-time user image | |
| CN112669113A (en) | Product recommendation method and device, storage medium and electronic device | |
| CN110162624A (en) | A kind of text handling method, device and relevant device | |
| CN112749330A (en) | Information pushing method and device, computer equipment and storage medium | |
| CN111475731A (en) | Data processing method, device, storage medium and equipment | |
| CN106933864A (en) | A kind of search engine system and its searching method | |
| CN111428007B (en) | Cross-platform based synchronous push feedback method | |
| CN113077320A (en) | Business opportunity recommendation method based on user behavior acquisition and analysis | |
| CN114881712B (en) | Intelligent advertisement putting method, device, equipment and storage medium | |
| CN107145600A (en) | Suitable for generic enterprise's public information classifying method | |
| CN112445985A (en) | Similar population acquisition method based on browsing behavior optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |