CN114385904A - Information recommendation method, device, equipment and storage medium - Google Patents

Information recommendation method, device, equipment and storage medium Download PDF

Info

Publication number
CN114385904A
CN114385904A CN202011142991.9A CN202011142991A CN114385904A CN 114385904 A CN114385904 A CN 114385904A CN 202011142991 A CN202011142991 A CN 202011142991A CN 114385904 A CN114385904 A CN 114385904A
Authority
CN
China
Prior art keywords
information
user
heat
click
score
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
Application number
CN202011142991.9A
Other languages
Chinese (zh)
Inventor
刘运明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Hongxiang Technical Service Co Ltd
Original Assignee
Beijing Hongxiang Technical Service Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Hongxiang Technical Service Co Ltd filed Critical Beijing Hongxiang Technical Service Co Ltd
Priority to CN202011142991.9A priority Critical patent/CN114385904A/en
Publication of CN114385904A publication Critical patent/CN114385904A/en
Pending legal-status Critical Current

Links

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

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention belongs to the technical field of information processing, and discloses an information recommendation method, an information recommendation device, information recommendation equipment and a storage medium. According to the method, when the current login user is a cold user, the click occupation ratio and the heat score occupation ratio of each piece of information to be recommended are obtained, and then the selection probability corresponding to each piece of information to be recommended is determined according to the click occupation ratio, the heat score occupation ratio and the historical information of the current login user; and selecting proper information to be recommended as recommendation information according to the selection probability corresponding to each information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user, wherein the selection probability of the information to be recommended is calculated based on the click rate occupation ratio, the heat score occupation ratio and the historical information of the current login user, so that the recommendation information selected according to the selection probability of the information to be recommended better conforms to the interest point of the current login user.

Description

Information recommendation method, device, equipment and storage medium
Technical Field
The present invention relates to the field of information processing technologies, and in particular, to an information recommendation method, apparatus, device, and storage medium.
Background
A cold user recommendation algorithm, also called a cold start strategy, is a common strategy for performing information recommendation for cold users (inactive users), and how to accurately perform information recommendation for the cold users is a common problem in recommendation algorithms, and the prior art generally recommends hot information to users directly or recommends according to obtained user figures, but the prior art has two disadvantages, namely, recommendation of the hot information is not accurate enough, and interest points of different people may be different; secondly, the difficulty of obtaining the user portrait is different in different scenes, and in some specific scenes, information useful for a cold user is difficult to obtain, namely, a usable user portrait is very difficult to form.
The above is only for the purpose of assisting understanding of the technical aspects of the present invention, and does not represent an admission that the above is prior art.
Disclosure of Invention
The invention mainly aims to provide an information recommendation method, an information recommendation device, information recommendation equipment and a storage medium, and aims to solve the technical problem that hot information meeting the interest points of users cannot be recommended to cold users in the prior art.
To achieve the above object, the present invention provides a method comprising the steps of:
when the current login user is a cold user, acquiring the click percentage and the heat percentage of each piece of information to be recommended, wherein the cold user is a user without the click behavior of the recommended information within a preset time range;
determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user.
Preferably, the obtaining of the click percentage and the heat score percentage of each piece of information to be recommended when the current login user is a cold user, wherein before the step of the cold user being a user without a click behavior of the recommended information within a preset time range, the method further includes:
determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate and the information correlation attributes of each heat information;
determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score;
and determining the information to be recommended according to the heat scores of the heat information.
Preferably, the step of determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score includes:
adjusting the weight of weighted summation according to the acquisition time of the minute-level heat fraction;
and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
Preferably, the step of determining the information to be recommended according to the heat scores of the pieces of heat information includes:
and sorting the heat information from large to small according to the heat scores, and selecting a preset number of heat information as information to be recommended according to a sorting result.
Preferably, the step of determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate, and the information association attribute of each heat information includes:
acquiring a first user click rate, a first information click rate and first information association attributes of each hot information within a first preset time range;
acquiring a second user click rate, a second information click rate and second information association attributes of each hot information within a second preset time range;
determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information correlation attribute;
and determining the minute-level heat score of each piece of heat information according to the second user click rate, the second information click rate and the second information correlation attribute.
Preferably, the step of determining the hour-level heat score of each piece of heat information according to the first user click rate, the first information click rate, and the first information correlation attribute includes:
classifying the first user click rate according to the first information correlation attribute to obtain a recall user click rate, an image user click rate and an image-free user click rate;
classifying the first information click rate according to the first information correlation attribute to obtain a recall information click rate, an image information click rate and a no-image information click rate;
calculating the hour-grade heat scores of the heat information according to the recall user click rate, the portrait user click rate, the no portrait user click rate, the recall information click rate, the portrait information click rate and the no portrait information click rate through a heat score formula;
the heat fraction formula is:
score＝a*(PVCNP+UVCNP)+b*(PVCWP+UVCWP)+c*(PVback+UVback)
in the formula: a. b and c are weight parameters, a + b + c is 1, UVCNPFor no image user click rate, PVCNPClick rate for no image information, UVCWPTo portray user click rate, PVCWPFor image information click-through rate, UVbackTo recall user click-through rates, PVbackClick-through rates for recall information.
Preferably, the step of determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the history information of the current login user includes:
determining click exposure gain according to the click occupation ratio and the heat fraction occupation ratio;
determining user exposure gain according to the heat fraction percentage and the historical information of the current login user;
and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
Preferably, the step of determining the click exposure gain according to the click occupation ratio and the heat score occupation ratio includes:
calculating click exposure gain through a click exposure gain formula according to the click occupation ratio and the heat fraction occupation ratio;
the click exposure gain formula is as follows:
pv_gianitem＝(ratescore*confidence-ratePV)/(ratescore*confidence)
in the formula: pv _ gianitemFor click exposure gain, confidence is a preset confidence parameter, ratePVTo click to rate, ratescoreIs the heat fraction ratio.
Preferably, the step of determining the user exposure gain according to the popularity score percentage and the history information of the currently logged-in user includes:
determining the exposure occupation ratio of the user within a preset time range according to the historical information of the current login user;
calculating user exposure gain through a user exposure gain formula according to the heat fraction occupation ratio and the user exposure occupation ratio;
the user exposure gain formula is:
pv_gianuser＝(ratescore*confidence-ratendays)/(ratescore*confidence)
in the formula: pv _ gianuserConfidence is a pre-set confidence parameter, rate, for the user exposure gainndaysExposure to the user, n is a predetermined value, ratescoreIs the heat fraction ratio.
Preferably, the step of determining the selection probability corresponding to each piece of information to be recommended according to the heat score percentage, the click exposure gain and the user exposure gain includes:
calculating the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain through a selection probability formula;
the selection probability formula is as follows:
Probshow＝ratescore*exp(pv_gianitem)*exp(pv_gianuser)
in the formula: probshowTo select the probability, ratescorePv _ gian, a percentage of heatitemTo clickExposure gain, pv _ gianuserFor user exposure gain, exp (pv _ gian)item) Exp (pv _ gian) as an exponential function with click exposure gain as a parameteruser) Is an exponential function with the user exposure gain as a parameter.
Preferably, the obtaining of the click percentage and the heat score percentage of each piece of information to be recommended when the current login user is a cold user, wherein before the step of the cold user being a user without a click behavior of the recommended information within a preset time range, the method further includes:
when the user login is detected, acquiring the history information of the current login user;
and judging whether the current login user is a cold user or not according to the historical information.
Preferably, the step of determining whether the current login user is a cold user according to the history information includes:
judging whether the current login user belongs to a new user or not according to the historical information;
when the current login user belongs to a new user, judging that the current login user is a cold user;
when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information;
and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
In addition, to achieve the above object, the present invention also provides an information recommendation apparatus, including:
the information acquisition module is used for acquiring the click percentage and the heat percentage of each piece of information to be recommended when the current login user is a cold user, wherein the cold user is a user without the click behavior of the recommendation information within a preset time range;
the probability calculation module is used for determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and the information recommendation module is used for selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended and pushing the recommendation information to the terminal equipment corresponding to the current login user.
Preferably, the information obtaining module is further configured to obtain historical information of a currently logged-in user when the user login is detected; and judging whether the current login user is a cold user or not according to the historical information.
Preferably, the information obtaining module is further configured to determine whether the current login user belongs to a new user according to the history information; when the current login user belongs to a new user, judging that the current login user is a cold user; when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information; and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
Preferably, the information acquisition module is further configured to determine the hour-level heat score and the minute-level heat score of each piece of heat information according to the user click rate, the information click rate, and the information association attribute of each piece of heat information; determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score; and determining the information to be recommended according to the heat scores of the heat information.
Preferably, the information acquisition module is further configured to adjust a weight of the weighted summation according to an acquisition time of the minute-level heat score; and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
Preferably, the probability calculation module is further configured to determine a click exposure gain according to the click occupation ratio and the heat score occupation ratio; determining user exposure gain according to the heat fraction percentage and the historical information of the current login user; and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
In addition, to achieve the above object, the present invention also provides an information recommendation apparatus, including: a memory, a processor and an information recommendation program stored on the memory and operable on the processor, the information recommendation program when executed by the processor implementing the steps of any of the information recommendation methods described above
In addition, to achieve the above object, the present invention further provides a computer-readable storage medium, on which an information recommendation program is stored, and when the information recommendation program is executed, the method of any one of the above information recommendation methods is implemented.
According to the method, when the current login user is a cold user, the click occupation ratio and the heat score occupation ratio of each piece of information to be recommended are obtained, and then the selection probability corresponding to each piece of information to be recommended is determined according to the click occupation ratio, the heat score occupation ratio and the historical information of the current login user; and selecting proper information to be recommended as recommendation information according to the selection probability corresponding to each information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user, wherein the selection probability of the information to be recommended is calculated based on the click rate occupation ratio, the heat score occupation ratio and the historical information of the current login user, so that the recommendation information selected according to the selection probability of the information to be recommended better conforms to the interest point of the current login user.
Drawings
Fig. 1 is a schematic structural diagram of an electronic device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first embodiment of an information recommendation method according to the present invention;
FIG. 3 is a flowchart illustrating a second embodiment of an information recommendation method according to the present invention;
FIG. 4 is a flowchart illustrating a third embodiment of an information recommendation method according to the present invention;
fig. 5 is a block diagram of an information recommendation apparatus according to a first embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of an information recommendation device of a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the electronic device may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a Random Access Memory (RAM) Memory, or may be a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the electronic device and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include therein an operating system, a network communication module, a user interface module, and an information recommendation program.
In the electronic apparatus shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 of the electronic device according to the present invention may be provided in an information recommendation device, and the electronic device calls the information recommendation program stored in the memory 1005 through the processor 1001 and executes the information recommendation method provided by the embodiment of the present invention.
An embodiment of the present invention provides an information recommendation method, and referring to fig. 2, fig. 2 is a flowchart illustrating a first embodiment of an information recommendation method according to the present invention.
In this embodiment, the information recommendation method includes the following steps:
step S10: when the current login user is a cold user, acquiring the click percentage and the heat percentage of each piece of information to be recommended, wherein the cold user is a user without the click behavior of the recommended information within a preset time range;
it should be noted that the execution subject of this embodiment is the information recommendation device, and the information recommendation device may be an electronic device such as a server, a cloud server, a virtual server, or other devices that can achieve the same or similar functions.
In practical use, a cold user is a user who has no recommended information click behavior within a preset time range, for example: and judging the users who have not clicked any recommendation information within 30 days as cold users.
It should be noted that the information to be recommended may be trending information to be recommended to the user, the trending information may be trending article information, trending news information, trending activity information, and the like, the click ratio may be a ratio of the click rate of a single piece of information to be recommended to the sum of the click rates of all pieces of information to be recommended, and the heat score ratio may be a ratio of the heat score of a single piece of information to be recommended to the sum of the heat scores of all pieces of information to be recommended.
For example: the current information to be recommended is A, B, C, the click rate of each piece of recommendation information is 10%, 20% and 30%, the heat score of each piece of recommendation information is 20, 30 and 40, the click rate ratio of each piece of recommendation information is 10/60, 20/60 and 30/60, and the heat score ratio of each piece of recommendation information is 20/90, 30/90 and 40/90.
Step S20: determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
the history information of the currently logged-in user may include information such as a user history browsing record and a user login time, for example: the last login time of the current login user is 9: 00, A, B are recommended information viewed by browsing.
It should be noted that the selection probability is a probability that information to be recommended is selected as recommendation information, and according to different actual implementation manners, the selection probability may be a percentage value or an integer value.
Further, in order to make the selection probability corresponding to each determined information to be recommended more meet the requirement of the current login user, step S20 of this embodiment may be:
determining click exposure gain according to the click occupation ratio and the heat fraction occupation ratio; determining user exposure gain according to the heat fraction percentage and the historical information of the current login user; and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
It should be noted that the selection probability is obtained by calculating using the click exposure gain, the user exposure gain, and the heat score ratio, and is used as a reference value for representing the possibility of meeting the interest point of the currently logged user.
Further, in order to calculate the click exposure gain, the step of determining the click exposure gain according to the click percentage and the heat score percentage may be:
calculating click exposure gain through a click exposure gain formula according to the click occupation ratio and the heat fraction occupation ratio;
the click exposure gain formula is as follows:
pv_gianitem＝(ratescore*confidence-ratePV)/(ratescore*confidence)
in the formula: pv _ gianitemFor click exposure gain, confidence is a preset confidence parameter, ratePVTo click to rate, ratescoreIs the heat fraction ratio.
In actual use, the preset confidence parameter confidence may be preset according to actual conditions, for example: assuming that the confidence level of the data is 90%, the preset confidence level parameter confidence may be set to 0.9, and the click exposure gain formula is pv _ gianitem＝(0.9*ratescore-ratePV)/(0.9*ratescore) And then substituting the heat fraction ratio and the click ratio corresponding to the information to be recommended into a formula, and calculating to obtain the click exposure gain.
Further, in order to calculate the user exposure gain, the step of determining the user exposure gain according to the heat score percentage and the history information of the currently logged-in user in this embodiment may be:
determining the exposure occupation ratio of the user within a preset time range according to the historical information of the current login user; calculating user exposure gain through a user exposure gain formula according to the heat fraction occupation ratio and the user exposure occupation ratio;
the user exposure gain formula is:
pv_gianuser＝(ratescore*confidence-ratendays)/(ratescore*confidence)
in the formula: pv _ gianuserConfidence is a pre-set confidence parameter, rate, for the user exposure gainndaysExposure to the user, n is a predetermined value, ratescoreIs the heat fraction ratio.
It should be noted that the user exposure occupation ratio is a ratio of the number of recommendations, which is currently calculated, of information to be recommended to the currently logged-in user in n days to the total number of recommendations of all recommendation information recommended to the currently logged-in user in n days, for example: the number of times that the information to be recommended in the calculation is recommended to the current login user within 7 days is 3,the total recommendation times of all recommendation information recommended to the current login user within 7 days is 50 times, and the user exposure ratio ratendays＝3/50。
In actual use, the preset confidence parameter confidence may be preset according to actual conditions, for example: assuming that the confidence level of the data is 90%, the preset confidence parameter confidence may be set to 0.9, and the user exposure gain formula is pv _ gianuser＝(0.9*ratescore-ratendays)/(0.9*ratescore) And then substituting the heat fraction ratio and the user exposure ratio into a user exposure gain formula for calculation to obtain the user exposure gain.
Further, in order to calculate the selection probability, the step of determining the selection probability corresponding to each piece of information to be recommended according to the heat score occupation ratio, the click exposure gain, and the user exposure gain may be:
calculating the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain through a selection probability formula;
the selection probability formula is as follows:
Probshow＝ratescore*exp(pv_gianitem)*exp(pv_gianuser)
in the formula: probshowTo select the probability, ratescorePv _ gian, a percentage of heatitemFor click exposure gain, pv _ gianuserFor user exposure gain, exp (pv _ gian)item) Exp (pv _ gian) as an exponential function with click exposure gain as a parameteruser) Is an exponential function with the user exposure gain as a parameter.
Exp (x) is an exponential function and is an important function in mathematics, and the corresponding value is exp (x) exAnd e is a natural constant, a constant in mathematics, an infinite acyclic decimal number, and an transcendental number, and has a value of about 2.718.
In practical use, the heat fraction ratio can be set as a basic selection probability, the basic selection probability is adjusted by calculating the user exposure gain index value and the click exposure gain index value and using the user exposure gain index value and the click exposure gain index value, and thus, the selection probability of each piece of information to be recommended can be obtained.
For example: the percentage of the heat fraction of the information to be recommended is 15%, and the user exposure gain index value exp (pv _ gian) is obtained through calculationuser) 0.7, click exposure gain exponent value exp (pv _ gian)item) 0.6, the probability Prob of selecting the information to be recommendedshow＝15％*0.7*0.6＝6.3％。
Step S30: and selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user.
In practical use, because of different ways of implementing the probabilities by computers, the selection probability may be a percentage value or a specific integer value, for example: when the selection probability is a percentage value, such as 6%, the judgment can be carried out by using a random number, and when the random number is less than or equal to 6%, the selection is determined; when the selection probability is a specific integer value, if there are A, B, C kinds of information to be recommended, if the selection probability is 20, 30, 40 respectively, an empty set can be established, 20A, 30B, 40C are put into the set, the elements in the set are subjected to disorder processing, so that the internal elements are uniformly distributed, the subscripts of the set are randomly extracted, the elements corresponding to the subscripts of the set are obtained, and the randomly selected information to be recommended is judged according to the obtained elements.
According to the method, when the current login user is a cold user, the click occupation ratio and the heat score occupation ratio of each piece of information to be recommended are obtained, and then the selection probability corresponding to each piece of information to be recommended is determined according to the click occupation ratio, the heat score occupation ratio and the historical information of the current login user; and selecting proper information to be recommended as recommendation information according to the selection probability corresponding to each information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user, wherein the selection probability of the information to be recommended is calculated based on the click rate occupation ratio, the heat score occupation ratio and the historical information of the current login user, so that the recommendation information selected according to the selection probability of the information to be recommended better conforms to the interest point of the current login user.
Referring to fig. 3, fig. 3 is a flowchart illustrating an information recommendation method according to a second embodiment of the present invention.
Based on the first embodiment, before the step S10, the information recommendation method of this embodiment further includes:
step S01: determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate and the information correlation attributes of each heat information;
it should be noted that, the user click rate of the heat information is a ratio of the number of users clicking the heat information to the total number of users clicking the heat information, the information click rate of the heat information is a ratio of the number of clicks of the heat information to the total number of clicks of all the heat information, the hour-level heat score may be a heat score periodically updated in hours, and the minute-level heat score may be a heat score periodically updated in minutes, for example: and the hour-level heat fraction is used for updating data in an hour period, and the minute-level heat fraction is used for updating data in a ten-minute period.
Further, in this embodiment, the step of determining the hour-level heat score and the minute-level heat score of each piece of heat information according to the user click rate, the information click rate, and the information association attribute of each piece of heat information may be:
acquiring a first user click rate, a first information click rate and first information association attributes of each hot information within a first preset time range; acquiring a second user click rate, a second information click rate and second information association attributes of each hot information within a second preset time range; determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information correlation attribute; and determining the minute-level heat score of each piece of heat information according to the second user click rate, the second information click rate and the second information correlation attribute.
The information association attribute may include various association attributes corresponding to the heat information, for example: the pushing channel when each click behavior occurs, whether a user portrait exists or not are related, and the like, the first preset time range may be that the calculation proceeding time is estimated from 24 hours to the calculation proceeding time, and the second preset time range may be that the calculation proceeding time is estimated from 10 minutes to the calculation proceeding time.
Further, in order to calculate the hour-level heat score, the step of determining the hour-level heat score of each piece of heat information according to the first user click rate, the first information click rate, and the first information correlation attribute in this embodiment may be:
classifying the first user click rate according to the first information correlation attribute to obtain a recall user click rate, an image user click rate and an image-free user click rate; classifying the first information click rate according to the first information correlation attribute to obtain a recall information click rate, an image information click rate and a no-image information click rate; calculating the hour-grade heat scores of the heat information according to the recall user click rate, the portrait user click rate, the no portrait user click rate, the recall information click rate, the portrait information click rate and the no portrait information click rate through a heat score formula;
the heat fraction formula is:
score＝a*(PVCNP+UVCNP)+b*(PVCWP+UVCWP)+c*(PVback+UVback)
in the formula: a. b and c are weight parameters, a + b + c is 1, UVCNPFor no image user click rate, PVCNPClick rate for no image information, UVCWPTo portray user click rate, PVCWPFor image information click-through rate, UVbackTo recall user click-through rates, PVbackClick-through rates for recall information.
It should be noted that a, b, and c in the heat score formula are all weight parameters, and may be adjusted according to actual situations, and it is only necessary to ensure that a + b + c is 1, which is not limited in this embodiment.
In actual use, a click channel of current heat information can be determined according to a first information correlation attribute, all information click rates and user click rates clicked through an information recommendation channel of information recommendation equipment are screened to obtain a recall information click rate and a recall user click rate, whether associated user portraits exist in the heat information or not is judged according to the first information correlation attribute, remaining information click rates and user click rates are classified to obtain a portray user click rate, a portray information click rate and a portray information click rate, and then an hour-level heat score of each heat information is calculated through a heat score calculation formula.
It should be noted that the minute-level heat score acquisition mode is basically the same as the hour-level heat score acquisition mode, and only the time range of data collection is different.
Step S02: determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score;
it should be noted that the actual interest points of the user may change with the time, the judgment of the heat degree by the hour-level heat degree score may result in a situation of untimely update, and the judgment of the heat degree by the minute and heat degree score may result in inaccurate judgment due to insufficient data samples, so that the heat degree score for judging the heat degree can be obtained by fusing the hour-level heat degree score and the minute-level heat degree score.
Further, in order to obtain the heat score for determining the heat, the step of determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score in this embodiment may be:
adjusting the weight of weighted summation according to the acquisition time of the minute-level heat fraction; and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
It should be noted that the actual interest point of the user changes with the lapse of time, and therefore, in the case of determining the heat degree, the longer the time interval between the minute-level heat degree score and the hour-level heat degree score is, the higher the confidence level of the minute-level heat degree score is, and therefore, the weight of the weighted sum needs to be adjusted according to the acquisition time of the specific minute-level heat degree score, and then the weighted sum is performed, so that it can be ensured that the obtained heat degree score of the heat degree information more conforms to the actual interest point of the user.
For example: updating the hour-level heat score by taking one hour as a period, updating the minute-level heat score by taking 10 minutes as a period, and setting the heat value of the hour-level heat score and the heat value of the minute-level heat score at the current moment as 9: 00, the hour-level heat score is updated, the weight of the hour-level heat score may be set to 1, the weight of the minute-level heat score may be set to 0, and the current time is 9: at time 10, the hour-level heat score is not updated, but the minute-level heat score is updated, and at this time, the weight of the hour-level heat score may be set to 0.5, the weight of the minute-level heat score may be set to 0.5, and the current time is 9: at 20 hours, the hour scale heat score is not updated, but the minute scale heat score is updated again, and at this time, the weight of the hour scale heat score can be set to 0.4 and the weight of the minute scale heat score can be set to 0.6.
Step S03: and determining the information to be recommended according to the heat scores of the heat information.
It should be noted that the heat scores represent the actual heat of each piece of heat information, and therefore, the information to be recommended may be determined by sorting according to the heat scores.
Further, in order to determine information to be recommended according to the popularity scores, the step of determining information to be recommended according to the popularity scores of the popularity information in this embodiment may be:
and sorting the heat information from large to small according to the heat scores, and selecting a preset number of heat information as information to be recommended according to a sorting result.
It should be noted that the preset number may be set according to actual requirements, and this embodiment is not limited to this.
In actual use, the preset number is set to be 10, all the heat information is sorted from large to small according to the heat scores, and after the sorting is completed, the top 10 heat information is selected from the sorting result as the information to be recommended.
According to the embodiment, the hour-level heat score and the minute-level heat score of each piece of heat information are determined by obtaining the user click rate, the information click rate and the information correlation attributes of each piece of heat information, the heat scores of each piece of heat information are determined by the small-level heat score and the minute-level heat score, the heat information is sorted from large to small according to the heat scores, the heat information with the preset number is selected as the information to be recommended, and the information to be recommended is guaranteed to be the heat information with the higher current heat.
Referring to fig. 4, fig. 4 is a flowchart illustrating an information recommendation method according to a third embodiment of the present invention.
Based on the first embodiment, before the step S10, the information recommendation method of this embodiment further includes:
step S01': when the user login is detected, acquiring the history information of the current login user;
it should be noted that, after the user logs in, the general system records the behavior information of the user, for example: browsing records, clicking records, purchasing records and other information, so that when the login of the user is detected, the recorded behavior information can be inquired to obtain the history information of the currently logged-in user.
Step S02': and judging whether the current login user is a cold user or not according to the historical information.
It can be understood that the analysis can be performed according to the history information, when the history information of the current login user meets the cold user determination condition, the current login user can be determined to be the cold user, and when the history information of the current login user does not meet the cold user determination condition, the current login user can be determined not to be the cold user.
Further, in order to determine whether the current login user is a cold user according to the history information, the step of determining whether the current login user is a cold user according to the history information may be:
judging whether the current login user belongs to a new user or not according to the historical information; when the current login user belongs to a new user, judging that the current login user is a cold user; when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information; and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
In actual use, the new user refers to a user who just completes registration and performs system login for the first time, and at this time, the history information of the current login user only includes login records, and no specific browsing record, click record and the like exist, so that whether the current login user is a new user can be easily judged according to the history information. When the current login user is not a new user, the browsing record, the click record and the like of the current login user can be acquired from the historical information, and the preset time range can be set according to actual needs, for example: and setting the preset time range to be within 30 days, and judging that the current login user is a cold user when no recommendation information click record exists in the current login user within 30 days.
According to the information recommendation method, whether the current login user is a cold user or not is judged by analyzing the historical information of the current login user, whether the current login user is a recommended audience of the information recommendation method is determined, and when the current login user is a cold user, information recommendation is performed by using the information recommendation method, so that the recommendation accuracy of the information recommendation method is further improved.
In addition, an embodiment of the present invention further provides a storage medium, where an information recommendation program is stored on the storage medium, and the information recommendation program, when executed by a processor, implements the steps of the information recommendation method described above.
Referring to fig. 5, fig. 5 is a block diagram illustrating a first embodiment of an information recommendation device according to the present invention.
As shown in fig. 5, an information recommendation apparatus according to an embodiment of the present invention includes:
the information obtaining module 501 is configured to obtain a click percentage and a hot score percentage of each piece of information to be recommended when the current login user is a cold user, where the cold user is a user who has no click behavior of recommendation information within a preset time range;
a probability calculation module 502, configured to determine, according to the click occupation ratio, the popularity score occupation ratio, and the history information of the current login user, a selection probability corresponding to each piece of information to be recommended;
the information recommendation module 503 is configured to select recommendation information according to the selection probability corresponding to each piece of information to be recommended, and push the recommendation information to the terminal device corresponding to the currently logged-in user.
According to the method, when the current login user is a cold user, the click occupation ratio and the heat score occupation ratio of each piece of information to be recommended are obtained, and then the selection probability corresponding to each piece of information to be recommended is determined according to the click occupation ratio, the heat score occupation ratio and the historical information of the current login user; and selecting proper information to be recommended as recommendation information according to the selection probability corresponding to each information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user, wherein the selection probability of the information to be recommended is calculated based on the click rate occupation ratio, the heat score occupation ratio and the historical information of the current login user, so that the recommendation information selected according to the selection probability of the information to be recommended better conforms to the interest point of the current login user.
Further, the information obtaining module 501 is further configured to determine an hour-level heat score and a minute-level heat score of each heat information according to the user click rate, the information click rate, and the information association attribute of each heat information; determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score; and determining the information to be recommended according to the heat scores of the heat information.
Further, the information obtaining module 501 is further configured to adjust a weighting of weighted summation according to the obtaining time of the minute-level heat score; and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
Further, the information obtaining module 501 is further configured to sort the heat information according to the heat scores from large to small, and select a preset number of heat information as information to be recommended according to a sorting result.
Further, the information obtaining module 501 is further configured to obtain a first user click rate, a first information click rate, and a first information association attribute of each hotness information within a first preset time range; acquiring a second user click rate, a second information click rate and second information association attributes of each hot information within a second preset time range; determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information correlation attribute; and determining the minute-level heat score of each piece of heat information according to the second user click rate, the second information click rate and the second information correlation attribute.
Further, the information obtaining module 501 is further configured to classify the first user click rate according to the first information association attribute, so as to obtain a recall user click rate, a portrait user click rate, and a no-portrait user click rate; classifying the first information click rate according to the first information correlation attribute to obtain a recall information click rate, an image information click rate and a no-image information click rate; calculating the hour-grade heat scores of the heat information according to the recall user click rate, the portrait user click rate, the no portrait user click rate, the recall information click rate, the portrait information click rate and the no portrait information click rate through a heat score formula;
the heat fraction formula is:
score＝a*(PVCNP+UVCNP)+b*(PVCWP+UVCWP)+c*(PVback+UVback)
in the formula: a. b and c are weight parameters, a + b + c is 1, UVCNPFor no image user click rate, PVCNPClick rate for no image information, UVCWPTo portray user click rate, PVCWPFor image information click-through rate, UVbackTo recall user click-through rates, PVbackFor recallingAnd (4) information click rate.
Further, the probability calculation module 502 is further configured to determine a click exposure gain according to the click occupation ratio and the heat score occupation ratio; determining user exposure gain according to the heat fraction percentage and the historical information of the current login user; and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
Further, the probability calculation module 502 is further configured to calculate a click exposure gain according to the click occupation ratio and the heat score occupation ratio by a click exposure gain formula;
the click exposure gain formula is as follows:
pv_gianitem＝(ratescore*confidence-ratePV)/(ratescore*confidence)
in the formula: pv _ gianitemFor click exposure gain, confidence is a preset confidence parameter, ratePVTo click to rate, ratescoreIs the heat fraction ratio.
Further, the probability calculation module 502 is further configured to determine a user exposure occupation ratio within a preset time range according to the history information of the current login user;
calculating user exposure gain through a user exposure gain formula according to the heat fraction occupation ratio and the user exposure occupation ratio;
the user exposure gain formula is:
pv_gianuser＝(ratescore*confidence-ratendays)/(rateScore*confidence)
in the formula: pv _ gianuserConfidence is a pre-set confidence parameter, rate, for the user exposure gainndaysExposure to the user, n is a predetermined value, ratescoreIs the heat fraction ratio.
Further, the probability calculation module 502 is further configured to calculate, according to the heat score occupation ratio, the click exposure gain, and the user exposure gain, a selection probability corresponding to each piece of information to be recommended by using a selection probability formula;
the selection probability formula is as follows:
Probshow＝ratescore*exp(pv_gianitem)*exp(pv_gianuser)
in the formula: probshowTo select the probability, ratescorePv _ gian, a percentage of heatitemFor click exposure gain, pv _ gianuserFor user exposure gain, exp (pv _ gian)item) Exp (pv _ gian) as an exponential function with click exposure gain as a parameteruser) Is an exponential function with the user exposure gain as a parameter.
Further, the information obtaining module 501 is further configured to, when it is detected that a user logs in, obtain history information of a currently logged-in user; and judging whether the current login user is a cold user or not according to the historical information.
Further, the information obtaining module 501 is further configured to determine whether the current login user belongs to a new user according to the history information; when the current login user belongs to a new user, judging that the current login user is a cold user; when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information; and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
It should be understood that the above is only an example, and the technical solution of the present invention is not limited in any way, and in a specific application, a person skilled in the art may set the technical solution as needed, and the present invention is not limited thereto.
It should be noted that the above-described work flows are only exemplary, and do not limit the scope of the present invention, and in practical applications, a person skilled in the art may select some or all of them to achieve the purpose of the solution of the embodiment according to actual needs, and the present invention is not limited herein.
In addition, the technical details that are not described in detail in this embodiment may refer to the information recommendation method provided in any embodiment of the present invention, and are not described herein again.
Further, it is to be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention or portions thereof that contribute to the prior art may be embodied in the form of a software product, where the computer software product is stored in a storage medium (e.g. Read Only Memory (ROM)/RAM, magnetic disk, optical disk), and includes several instructions for enabling a terminal device (e.g. a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
The invention discloses A1 and an information recommendation method, which comprises the following steps:
when the current login user is a cold user, acquiring the click percentage and the heat percentage of each piece of information to be recommended, wherein the cold user is a user without the click behavior of the recommended information within a preset time range;
determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user.
A2, the information recommendation method as in a1, where when the current logged-in user is a cold user, the click percentage and the heat score percentage of each piece of information to be recommended are obtained, and before the step of the cold user being a user who has no information recommended click behavior within a preset time range, the method further includes:
determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate and the information correlation attributes of each heat information;
determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score;
and determining the information to be recommended according to the heat scores of the heat information.
A3, the information recommendation method of A2, the step of determining the heat scores of each heat information according to the hour-scale heat scores and the minute-scale heat scores, comprising:
adjusting the weight of weighted summation according to the acquisition time of the minute-level heat fraction;
and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
A4, the information recommendation method as in a2, wherein the step of determining the information to be recommended according to the heat scores of the respective heat information comprises:
and sorting the heat information from large to small according to the heat scores, and selecting a preset number of heat information as information to be recommended according to a sorting result.
The information recommendation method of a5, as in a2, the step of determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate and the information association attribute of each heat information includes:
acquiring a first user click rate, a first information click rate and first information association attributes of each hot information within a first preset time range;
acquiring a second user click rate, a second information click rate and second information association attributes of each hot information within a second preset time range;
determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information correlation attribute;
and determining the minute-level heat score of each piece of heat information according to the second user click rate, the second information click rate and the second information correlation attribute.
A6, the information recommendation method of A5, wherein the step of determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information association attribute comprises:
classifying the first user click rate according to the first information correlation attribute to obtain a recall user click rate, an image user click rate and an image-free user click rate;
classifying the first information click rate according to the first information correlation attribute to obtain a recall information click rate, an image information click rate and a no-image information click rate;
calculating the hour-grade heat scores of the heat information according to the recall user click rate, the portrait user click rate, the no portrait user click rate, the recall information click rate, the portrait information click rate and the no portrait information click rate through a heat score formula;
the heat fraction formula is:
score＝a*(PVCNP+UVCNP)+b*(PVCWP+UVCWP)+c*(PVback+UVback)
in the formula: a. b and c are weight parameters, a + b + c is 1, UVCNPFor no image user click rate, PVCNPClick rate for no image information, UVCWPTo portray user click rate, PVCWPFor image information click-through rate, UVbackTo recall user click-through rates, PVbackClick-through rates for recall information.
A7, the information recommendation method according to A1, wherein the step of determining the selection probability corresponding to each piece of information to be recommended according to the click percentage, the popularity score percentage and the history information of the current login user includes:
determining click exposure gain according to the click occupation ratio and the heat fraction occupation ratio;
determining user exposure gain according to the heat fraction percentage and the historical information of the current login user;
and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
A8, the information recommendation method of A7, wherein the step of determining the click exposure gain according to the click percentage and the heat score percentage comprises:
calculating click exposure gain through a click exposure gain formula according to the click occupation ratio and the heat fraction occupation ratio;
the click exposure gain formula is as follows:
pv_gianitem＝(ratescore*confidence-ratePV)/(ratescore*confidence)
in the formula: pv _ gianitemFor click exposure gain, confidence is a preset confidence parameter, ratePVTo click to rate, ratescoreIs the heat fraction ratio.
A9, the information recommendation method of A7, wherein the step of determining the user exposure gain according to the heat score percentage and the history information of the currently logged-in user comprises:
determining the exposure occupation ratio of the user within a preset time range according to the historical information of the current login user;
calculating user exposure gain through a user exposure gain formula according to the heat fraction occupation ratio and the user exposure occupation ratio;
the user exposure gain formula is:
pv_gianuser＝(ratescore*confidence-ratendays)/(ratescore*confidence)
in the formula: pv _ gianuserConfidence is a pre-set confidence parameter, rate, for the user exposure gainndaysExposure to the user, n is a predetermined value, ratescoreIs the heat fraction ratio.
A10, the information recommendation method according to A7, wherein the step of determining the selection probability corresponding to each piece of information to be recommended according to the heat score percentage, the click exposure gain and the user exposure gain comprises:
calculating the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain through a selection probability formula;
the selection probability formula is as follows:
Probshow＝ratescore*exp(pv_gianitem)*exp(pv_gianuser)
in the formula: probshowTo select the probability, ratescorePv _ gian, a percentage of heatitemFor click exposure gain, pv _ gianuserFor user exposure gain, exp (pv _ gian)item) Exp (pv _ gian) as an exponential function with click exposure gain as a parameteruser) Is an exponential function with the user exposure gain as a parameter.
A11, the information recommendation method as in a1, where when the current logged-in user is a cold user, the click percentage and the heat score percentage of each piece of information to be recommended are obtained, and before the step of the cold user being a user who has no information recommended click behavior within a preset time range, the method further includes:
when the user login is detected, acquiring the history information of the current login user;
and judging whether the current login user is a cold user or not according to the historical information.
A12, the information recommendation method of A11, wherein the step of determining whether the current logged-on user is a cold user according to the history information comprises:
judging whether the current login user belongs to a new user or not according to the historical information;
when the current login user belongs to a new user, judging that the current login user is a cold user;
when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information;
and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
The invention discloses B13 and an information recommendation device, which comprises the following modules:
the information acquisition module is used for acquiring the click percentage and the heat percentage of each piece of information to be recommended when the current login user is a cold user, wherein the cold user is a user without the click behavior of the recommendation information within a preset time range;
the probability calculation module is used for determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and the information recommendation module is used for selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended and pushing the recommendation information to the terminal equipment corresponding to the current login user.
B14, the information recommendation device as described in B13, the information obtaining module is further configured to obtain the history information of the currently logged-in user when the user login is detected; and judging whether the current login user is a cold user or not according to the historical information.
B15, the information recommendation device of B14, the information acquisition module is further configured to determine whether the current logged-in user belongs to a new user according to the history information; when the current login user belongs to a new user, judging that the current login user is a cold user; when the current login user does not belong to a new user, judging whether a recommended information clicking behavior exists in a preset time range of the current login user according to the historical information; and when the current login user has no recommended information clicking behavior within a preset time range, judging that the current login user is a cold user.
B16, the information recommendation device of B13, the information obtaining module further configured to determine an hour-level heat score and a minute-level heat score of each heat information according to the user click rate, the information click rate and the information association attribute of each heat information; determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score; and determining the information to be recommended according to the heat scores of the heat information.
B17, the information recommendation device as described in B16, the information acquisition module further configured to adjust the weight of the weighted sum according to the acquisition time of the minute-scale heat score; and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
B18, the information recommendation device as described in B13, the probability calculation module is further used for determining click exposure gain according to the click percentage and the heat score percentage; determining user exposure gain according to the heat fraction percentage and the historical information of the current login user; and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
The invention also discloses C19 and an information recommendation device, wherein the information recommendation device comprises: the information recommendation system comprises a memory, a processor and an information recommendation program stored on the memory and capable of running on the processor, wherein the information recommendation program realizes the steps of the information recommendation method when being executed by the processor.
The invention also discloses D20 and a computer readable storage medium, wherein the computer readable storage medium is stored with an information recommendation program, and when the information recommendation program is executed, the steps of the information recommendation method are realized.

Claims (10)

1. An information recommendation method, characterized in that the information recommendation method comprises the steps of:
when the current login user is a cold user, acquiring the click percentage and the heat percentage of each piece of information to be recommended, wherein the cold user is a user without the click behavior of the recommended information within a preset time range;
determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended, and pushing the recommendation information to the terminal equipment corresponding to the current login user.
2. The information recommendation method according to claim 1, wherein the step of obtaining the click percentage and the heat score percentage of each piece of information to be recommended when the current logged-in user is a cold user, and before the step of the cold user being a user who has no click behavior of recommendation information within a preset time range, further comprises:
determining the hour-level heat score and the minute-level heat score of each heat information according to the user click rate, the information click rate and the information correlation attributes of each heat information;
determining the heat score of each piece of heat information according to the hour-level heat score and the minute-level heat score;
and determining the information to be recommended according to the heat scores of the heat information.
3. The information recommendation method of claim 2, wherein the step of determining the heat score of each piece of heat information according to the hour-scale heat score and the minute-scale heat score comprises:
adjusting the weight of weighted summation according to the acquisition time of the minute-level heat fraction;
and carrying out weighted summation on the hour-level heat fraction and the minute-level heat fraction according to the weight to obtain the heat fraction of each heat information.
4. The information recommendation method according to claim 2, wherein the step of determining the information to be recommended according to the popularity scores of the respective popularity information comprises:
and sorting the heat information from large to small according to the heat scores, and selecting a preset number of heat information as information to be recommended according to a sorting result.
5. The information recommendation method according to claim 2, wherein the step of determining the hour-level heat score and the minute-level heat score of each of the heat information according to the user click rate, the information click rate, and the information association attribute of each of the heat information comprises:
acquiring a first user click rate, a first information click rate and first information association attributes of each hot information within a first preset time range;
acquiring a second user click rate, a second information click rate and second information association attributes of each hot information within a second preset time range;
determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information correlation attribute;
and determining the minute-level heat score of each piece of heat information according to the second user click rate, the second information click rate and the second information correlation attribute.
6. The information recommendation method of claim 5, wherein the step of determining the hour-level heat score of each heat information according to the first user click rate, the first information click rate and the first information association attribute comprises:
classifying the first user click rate according to the first information correlation attribute to obtain a recall user click rate, an image user click rate and an image-free user click rate;
classifying the first information click rate according to the first information correlation attribute to obtain a recall information click rate, an image information click rate and a no-image information click rate;
calculating the hour-grade heat scores of the heat information according to the recall user click rate, the portrait user click rate, the no portrait user click rate, the recall information click rate, the portrait information click rate and the no portrait information click rate through a heat score formula;
the heat fraction formula is:
score＝a*(PVCNP+UVCNP)+b*(PVCWP+UVCWP)+c*(PVback+UVback)
in the formula: a. b and c are weight parameters, a + b + c is 1, UVCNPFor no image user click rate, PVCNPClick rate for no image information, UVCWPTo portray user click rate, PVCWPFor image information click-through rate, UVbackTo recall user click-through rates, PVbackClick-through rates for recall information.
7. The information recommendation method according to claim 1, wherein the step of determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio, and the history information of the current login user comprises:
determining click exposure gain according to the click occupation ratio and the heat fraction occupation ratio;
determining user exposure gain according to the heat fraction percentage and the historical information of the current login user;
and determining the selection probability corresponding to each piece of information to be recommended according to the heat fraction ratio, the click exposure gain and the user exposure gain.
8. An information recommendation device, characterized in that the information recommendation device comprises the following modules:
the information acquisition module is used for acquiring the click percentage and the heat percentage of each piece of information to be recommended when the current login user is a cold user, wherein the cold user is a user without the click behavior of the recommendation information within a preset time range;
the probability calculation module is used for determining the selection probability corresponding to each piece of information to be recommended according to the click occupation ratio, the popularity score occupation ratio and the historical information of the current login user;
and the information recommendation module is used for selecting recommendation information according to the selection probability corresponding to each piece of information to be recommended and pushing the recommendation information to the terminal equipment corresponding to the current login user.
9. An information recommendation apparatus characterized by comprising: memory, processor and information recommendation program stored on the memory and executable on the processor, the information recommendation program when executed by the processor implementing the steps of the information recommendation method according to any of claims 1-7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon an information recommendation program which, when executed, implements the steps of the information recommendation method according to any one of claims 1-7.
CN202011142991.9A 2020-10-22 2020-10-22 Information recommendation method, device, equipment and storage medium Pending CN114385904A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011142991.9A CN114385904A (en) 2020-10-22 2020-10-22 Information recommendation method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011142991.9A CN114385904A (en) 2020-10-22 2020-10-22 Information recommendation method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN114385904A true CN114385904A (en) 2022-04-22

Family

ID=81194451

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011142991.9A Pending CN114385904A (en) 2020-10-22 2020-10-22 Information recommendation method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN114385904A (en)

Similar Documents

Publication Publication Date Title
CN109684118B (en) Abnormal data detection method, device, equipment and computer readable storage medium
CN111125574B (en) Method and device for generating information
CN105490823B (en) data processing method and device
CN109685537B (en) User behavior analysis method, device, medium and electronic equipment
CN110163647A (en) A kind of data processing method and device
CN112163963B (en) Service recommendation method, device, computer equipment and storage medium
CN110866767A (en) Method, device, equipment and medium for predicting satisfaction degree of telecommunication user
CN110782291A (en) Advertisement delivery user determination method and device, storage medium and electronic device
WO2023000491A1 (en) Application recommendation method, apparatus and device, and computer-readable storage medium
CN113076416A (en) Information heat evaluation method and device and electronic equipment
CN110910201A (en) Information recommendation control method and device, computer equipment and storage medium
CN110599240A (en) Application preference value determination method, device and equipment and storage medium
CN113656699B (en) User feature vector determining method, related equipment and medium
CN113505154B (en) Digital reading statistical analysis method and system based on big data
CN113723995A (en) Method and device for determining marketing task
CN113344660A (en) House source information processing method and device, electronic equipment and storage medium
Yucel et al. Sequential hierarchical regression imputation
CN113434762A (en) Association pushing method, device and equipment based on user information and storage medium
JP7015927B2 (en) Learning model application system, learning model application method, and program
CN111881469A (en) Data processing method and device
CN113836388A (en) Information recommendation method and device, server and storage medium
CN114385904A (en) Information recommendation method, device, equipment and storage medium
CN115914363A (en) Message pushing method and device, computer equipment and storage medium
CN115659005A (en) Product pushing method and device, computer equipment and storage medium
CN111784091B (en) Method and device for processing information

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