CN110457580B

CN110457580B - Hotspot recommendation method and device based on search

Info

Publication number: CN110457580B
Application number: CN201910703049.6A
Authority: CN
Inventors: 陈圣灵; 罗珺; 薇娜; 王春伟; 李升起
Original assignee: Baidu com Times Technology Beijing Co Ltd
Current assignee: Baidu com Times Technology Beijing Co Ltd
Priority date: 2019-07-31
Filing date: 2019-07-31
Publication date: 2022-05-10
Anticipated expiration: 2039-07-31
Also published as: CN110457580A

Abstract

The embodiment of the invention provides a hotspot recommendation method and device based on search, wherein the method comprises the following steps: and acquiring keywords corresponding to the hot spots to be sorted and the generation time of the keywords. And acquiring a first search quantity and a second search quantity corresponding to the keywords, wherein the first search quantity is the search quantity of the keywords in the current time period, the second search quantity is the maximum search quantity determined in the search quantities of the keywords corresponding to each time period in the previous N time periods, and N is an integer greater than 1. And determining the heat value of the hot spot to be sorted according to the first search quantity, the second search quantity, the generation time and the current time. And sorting and displaying the hot spots to be sorted according to the hot value of the hot spots to be sorted. The hot degree value of the hot spot corresponding to the keyword is determined by combining the first search amount and the second search amount, so that the hot point value can be obtained by combining the current state and the hot degree trend of the hot spot based on the search, the ranked hot spot can be obtained, and the accuracy and the effectiveness of the hot spot list are improved.

Description

Hotspot recommendation method and device based on search

Technical Field

The embodiment of the invention relates to the technical field of internet, in particular to a hot spot recommendation method and device based on search.

Background

In the current internet era, information is growing explosively. The convenience of information acquisition has greatly changed people's lifestyle, and more users have become accustomed to acquiring information over the internet.

In an internet search engine, a user can search for information details through keywords. In this process, the popularity of the hotspots corresponding to the keywords is generally measured according to the search amount of the keywords queried by the user, and the hotspot lists are generated in sequence. The hot spot list can reflect the most concerned hot spot at present, and a user can conveniently and quickly know the current hot spot in time.

However, most of the hotspots are first developed on the social media platform and the information streaming media platform, the search volume of the reaction of the keyword corresponding to the hotspot in the search engine is generally lagged, and the hotspots are ranked according to the search volume, so that a timely and effective hotspot list cannot be provided for the user.

Disclosure of Invention

The embodiment of the invention provides a hotspot recommendation method and device based on search, and aims to solve the problem that a user cannot be provided with timely and effective hotspot list because hotspots are ranked according to the amount of search.

In a first aspect, an embodiment of the present invention provides a hot spot recommendation method based on search, including:

acquiring keywords corresponding to hotspots to be sorted and the generation time of the keywords;

acquiring a first search quantity and a second search quantity corresponding to the keyword, wherein the first search quantity is the search quantity of the keyword in the current time period, the second search quantity is the maximum search quantity determined in the search quantities of the keyword corresponding to each time period in the previous N time periods, and N is an integer greater than 1;

determining the heat value of the hot spot to be sorted according to the first search quantity, the second search quantity, the generation time and the current time;

and sequencing and displaying the hotspots to be sequenced according to the heat value of the hotspots to be sequenced.

In one possible design, the determining the heat value of the hotspot to be sorted according to the first search volume, the second search volume, the generation time, and the current time includes:

determining a first heat index according to the first search quantity and a third search quantity, wherein the third search quantity is the search quantity of the keyword in the previous period;

determining a second heat index according to the second search quantity and the third search quantity;

determining a third heat index according to the generation time, the current time and a fourth search quantity, wherein the fourth search quantity is the maximum search quantity determined in the search quantities corresponding to the keywords of each hotspot in the current time period;

and determining the heat value of the hot spot to be sorted according to the first heat index, the second heat index, the third heat index and the fourth search amount.

In one possible design, the determining a first heat index according to the first search volume and a third search volume of the keyword in a previous period includes:

determining a search variation according to the first search quantity and the third search quantity;

and determining the first heat index according to the search variation, the first search quantity and a smoothing function.

In one possible design, the determining a second heat index based on the second search volume and the third search volume includes:

determining peak value variation according to the second search quantity and the third search quantity;

and determining the second heat index according to the peak value variation, the first search amount and a smoothing function.

In one possible design, the determining a third heat index according to the generation time, the current time, and a fourth search amount includes:

determining a time difference value according to the generation time and the current time;

and determining the third heat index according to the time difference, the fourth search quantity and a time attenuation coefficient.

In one possible design, the determining the heat value of the hotspot to be ranked according to the first heat index, the second heat index, the third heat index and the fourth search amount includes:

acquiring a heat index and a value according to the first heat index, the second heat index and the third heat index;

and determining the heat value of the hot spot to be sorted according to the heat index sum value, the fourth search quantity and the heat adjusting coefficient.

In a possible design, after the sorting and displaying the hot spots to be sorted according to the hot value of the hot spot to be sorted, the method further includes:

acquiring the heat characteristics of keywords corresponding to target hotspots ranked before a preset rank in the hotspot list;

inputting the heat characteristic into a click rate model, and acquiring a classification value output by the click rate model, wherein the classification value is used for indicating the click rate corresponding to the heat characteristic;

and re-sequencing the target hotspots according to the classification value output by the click rate model to obtain an updated hotspot list.

In one possible design, before the inputting the heat feature into the click rate model and obtaining the classification value output by the click rate model, the method further includes:

obtaining click rate information of a historical hotspot list, wherein the click rate information comprises a first click rate of each hotspot in a first time length before the current time and a second click rate of each hotspot in a second time length, the second time length is before the first time length, and the second time length is greater than the first time length;

obtaining a training sample according to the second click rate of each hot spot in the second time length and the heat characteristics of the keywords corresponding to each hot spot;

obtaining a test sample according to the first click rate of each hotspot in the first duration and the hot degree characteristics of the keywords corresponding to each hotspot;

and training the click rate model to be trained according to the training sample and the test sample to obtain the trained click rate model.

In a second aspect, an embodiment of the present invention provides a device for recommending hotspots based on search, including:

the acquisition module is used for acquiring keywords corresponding to the hotspots to be sorted and the generation time of the keywords;

the acquisition module is further configured to acquire a first search amount and a second search amount corresponding to the keyword, where the first search amount is a search amount of the keyword in a current time period, the second search amount is a maximum search amount determined in the search amounts of the keyword corresponding to each of the previous N time periods, and N is an integer greater than 1;

the determining module is used for determining the heat value of the hot spot to be sorted according to the first searching amount, the second searching amount, the generating time and the current time;

and the sorting module is used for sorting and displaying the hot spots to be sorted according to the heat value of the hot spots to be sorted.

In one possible design, the determining module is specifically configured to:

in one possible design, the ranking module is further to:

after the hot spots to be ranked are ranked and displayed according to the hot value of the hot spots to be ranked, acquiring the hot feature of the keyword corresponding to the target hot spot ranked before a preset rank in the hot spot list;

In one possible design, the obtaining module is further configured to:

before the hot characteristics are input into a click rate model and a classification value output by the click rate model is obtained, click rate information of a historical hotspot list is obtained, wherein the click rate information comprises a first click rate of each hotspot in a first duration before the current time and a second click rate of each hotspot in a second duration before the first duration, and the second duration is longer than the first duration;

obtaining a test sample according to the first click rate of each hotspot in the first duration and the heat characteristics of the keywords corresponding to each hotspot;

In a third aspect, an embodiment of the present invention provides a hotspot recommendation device based on search, including:

a memory for storing a program;

a processor for executing the program stored by the memory, the processor being adapted to perform the method as described above in the first aspect and any one of the various possible designs of the first aspect when the program is executed.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, including instructions, which, when executed on a computer, cause the computer to perform the method as described above in the first aspect and any one of various possible designs of the first aspect.

The embodiment of the invention provides a hotspot recommendation method and device based on search, wherein the method comprises the following steps: and acquiring keywords corresponding to the hot spots to be sorted and the generation time of the keywords. And acquiring a first search quantity and a second search quantity corresponding to the keywords, wherein the first search quantity is the search quantity of the keywords in the current time period, the second search quantity is the maximum search quantity determined in the search quantities of the keywords corresponding to each time period in the previous N time periods, and N is an integer greater than 1. And determining the heat value of the hot spot to be sorted according to the first search quantity, the second search quantity, the generation time and the current time. And sorting and displaying the hot spots to be sorted according to the hot value of the hot spots to be sorted. The heat degree value of the hot spot corresponding to the keyword is determined by combining the first search amount and the second search amount corresponding to the keyword, so that the heat degree value can be obtained by combining the current state and the heat degree trend of the hot spot based on searching, and the hot spot is sequenced according to the heat degree value of the hot spot to obtain a heat list, so that the problem that a timely and effective hot list cannot be provided for a user due to the fact that the hot list is determined only according to the real-time search amount is avoided, and the accuracy and the effectiveness of the hot list are improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

Fig. 1 is a system diagram of a hot spot recommendation method based on search according to an embodiment of the present invention;

FIG. 2 is a first flowchart of a hot spot recommendation method based on search according to an embodiment of the present invention;

FIG. 3 is a flowchart II of a hot spot recommendation method based on search according to an embodiment of the present invention;

fig. 4 is a flowchart three of the hot spot recommendation method based on search according to the embodiment of the present invention;

fig. 5 is a schematic structural diagram of a search-based hotspot recommendation device according to an embodiment of the present invention;

fig. 6 is a schematic hardware structure diagram of a hot spot recommendation device based on search according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. 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.

Fig. 1 is a schematic system diagram of a hot spot recommendation method based on search according to an embodiment of the present invention, and as shown in fig. 1, the system includes: a terminal device 101 and a server 102;

for example, the terminal device 101 may be operated with a browser, and a user may search for a hotspot or obtain related information through the browser, or may also perform the above operations through a client of a search engine installed on the terminal device 101, for example, the terminal device 101 may be a computer device, a tablet computer, or a mobile phone (or referred to as a "cellular" phone), and the terminal device 101 may also be a portable, pocket, handheld, or computer-embedded mobile device or apparatus, which is not limited herein.

The server 102 is configured to process the hotspot data according to the search information sent by the terminal device 101, sort the hotspots to obtain a hotspot list, and present the hotspot list to the user through the terminal device 101, so that the user can quickly obtain current hotspot information.

The terminal device 101 interacts with the server 101, where the interaction mode may be, for example, a wired network, the wired network may include, for example, a coaxial cable, a twisted pair, an optical fiber, and the like, and the interaction mode may also be, for example, a Wireless network, which may be a 2G network, a 3G network, a 4G network, a 5G network, a Wireless Fidelity (WIFI) network, and the like. The embodiment of the present invention does not limit the specific type or specific form of the interaction, as long as the function of the server and the terminal interaction can be realized.

Generally, in an internet search engine, the following three ways exist for generating a hotspot list:

1) and calculating scores based on data such as hot source news information reading amount or topic mutual amount in a weighting manner, sequencing, and directly defining event popularity and sequencing rules from a hot event source.

However, this solution is suitable for generating a hot list of information topics, and is not suitable for adopting such score calculation and ranking rules in a search engine-based hot list.

2) Based on the list display and the click data sorting, because the corresponding keywords of the newly generated hot spots have no search amount, corresponding keywords are inserted into the generated list at random or fixed positions in a rotation display mode, and the displayed click data is applied to the next round of sorting.

However, the heat score of the scheme of randomly or fixedly inserting new words is unreasonably calculated, and the heat of the event and the attention of the user to the event cannot be effectively reflected.

3) According to the search amount of the keywords queried by the user, measuring the heat degree of the hotspots corresponding to the keywords and generating the hotspot list in sequence

However, most of the hotspots are exploded first in a social media platform and an information streaming media platform (such as a community platform and a news platform), the search volume of the reaction of the keyword corresponding to the hotspot in a search engine is generally lagged, if the ranking of the hotspot is performed only according to the real-time search volume of the keyword, a newly generated hotspot event obviously cannot be listed in time, an old hotspot event is likely to stay on a list for a long time, a "madai effect" is easily generated, that is, a hotspot event with a higher heat degree and a top ranking is continuously overlooked due to the fact that the hotspot event is more and more concerned by the user, and a back ranking event is continuously leaned due to the fact that the hotspot event is not concerned by the user, so that a timely and effective hotspot list cannot be provided to the user.

Based on the above problem, an embodiment of the present invention provides a method for recommending hotspots based on search, which is described in detail below with reference to specific embodiments, and first described with reference to fig. 2, where fig. 2 is a first flowchart of the method for recommending hotspots based on search provided by the embodiment of the present invention, as shown in fig. 2, the method includes:

s201, obtaining keywords corresponding to hotspots to be sorted and generation time of the keywords.

In this embodiment, each hotspot corresponds to a related keyword, wherein the keyword may be understood as an abbreviation of the hotspot, the information included in the hotspot may be simply and clearly reflected, when the terminal device provides the hotspot list to the user, the keyword corresponding to the hotspot is actually displayed on the hotspot list, and when the user clicks the keyword, the detailed information of the hotspot corresponding to the keyword may be obtained.

For example, the detailed information of the existing hot spot is that "according to incomplete statistics, until now, 21 provinces determine the basic pension up-regulation scheme of the retired staff in 2019, and the total regulation level is about 5% of the pension in 2018. "the corresponding keyword may be" 2019 pension in many countries of the country, and those skilled in the art will understand that the keyword corresponding to each hotspot may be automatically generated according to the detailed information of the hotspot, or may be manually set.

In a possible implementation manner, keywords corresponding to mined hotspots can be stored in a database, so that keywords corresponding to hotspots to be ranked can be directly acquired from the database, wherein each keyword also corresponds to generation time, and the generation time of the keyword can be the review passing time of the keyword.

S202, obtaining a first search volume and a second search volume corresponding to the keywords, wherein the first search volume is the search volume of the keywords in the current time period, the second search volume is the maximum search volume determined in the search volumes of the keywords corresponding to each time period in the previous N time periods, and N is an integer larger than 1.

Specifically, a user searches on a platform of a search engine to generate a search log, where the search log records relevant search information of the platform user, such as search content, search time, and the like, and the embodiment performs matching in the search log according to a keyword, and when the keyword appears in the search log, the search amount of the keyword is increased by 1, so as to obtain a first search amount and a second search amount corresponding to the keyword.

In this embodiment, the time may be divided into a plurality of time periods, where the time periods may be divided according to a preset time duration, for example, 5 minutes is used as one time period, or the time periods may also be divided randomly, which is not limited in this embodiment, and the search volumes of the keywords in the respective time periods may be obtained by obtaining the search logs in different time periods and matching the search logs with the keywords, where the first search volume is the search volume of the keywords in the current time period, the second search volume is the maximum search volume determined in the search volumes of the keywords corresponding to each of the previous N time periods, and N is an integer greater than 1.

Taking N as an example for explanation, the first 3 time periods may be a first time period, a second time period, and a third time period, and assuming that the search volume of the keyword corresponding to the first time period is 20, the search volume of the keyword corresponding to the second time period is 30, and the search volume of the keyword corresponding to the third time period is 25, the second search volume is the search volume 30 corresponding to the second time period.

S203, determining the heat value of the hot spot to be sorted according to the first search quantity, the second search quantity, the generation time and the current time.

If the hotspot list is determined only according to the size of the real-time search amount, a timely and effective hotspot list cannot be provided for the user, so that the second search amount is introduced in the embodiment, wherein the second search amount reflects the historical peak search amount of the keyword, and can indicate the heat trend of the hotspot, and the first search amount reflects the search amount in the current time period, and can indicate the current state of the hotspot.

Therefore, the heat value of the hot spot to be sorted is determined according to the first search amount, the second search amount, the generation time and the current time, so that the heat value can be effectively obtained according to the current state and the heat trend of the hot spot, and the heat value is determined according to the size of the real-time search amount.

In a possible implementation manner, the change rate of the search volume of the keyword in a period of time and the change rate of the peak search volume of the search volume in a period of time may be determined according to the first search volume, the second search volume, the generation time and the current time, so that the change rate is used as a measure of the heat degree of the hot spot to obtain the heat degree value of the hot spot to be ranked.

And S204, sorting and displaying the hot spots to be sorted according to the hot degree value of the hot spots to be sorted.

In an alternative implementation manner, for example, the hot spots to be sorted may be sorted in the order of the hot value from large to small; or, the hotspots can be ranked in a descending order (or other possible order) according to the user requirement, and displayed.

Or, a preset number of hotspots ranked in the top may be selected according to the magnitude of the heat value, so as to obtain the ranked hotspots, and the specific implementation manner of obtaining the hotspot list is not limited in this embodiment, as long as the hotspot list can be obtained according to the heat value of the hotspot to be ranked, and the heat condition of the hotspot can be effectively reflected.

The hot spot recommendation method based on search provided by the embodiment of the invention comprises the following steps: and acquiring keywords corresponding to the hot spots to be sorted and the generation time of the keywords. And acquiring a first search quantity and a second search quantity corresponding to the keywords, wherein the first search quantity is the search quantity of the keywords in the current time period, the second search quantity is the maximum search quantity determined in the search quantities of the keywords corresponding to each time period in the previous N time periods, and N is an integer greater than 1. And determining the heat value of the hot spot to be sorted according to the first search quantity, the second search quantity, the generation time and the current time. And sorting and displaying the hot spots to be sorted according to the hot value of the hot spots to be sorted. The heat degree value of the hot spot corresponding to the keyword is determined by combining the first search amount and the second search amount corresponding to the keyword, so that the heat degree value can be obtained by combining the current state and the heat degree trend of the hot spot based on search, and the hot spot is sequenced according to the heat degree value of the hot spot to obtain a heat list, so that the condition that the timely and effective hot list cannot be displayed to a user due to the fact that the hot list is determined only according to the real-time search amount is avoided, and the accuracy and the effectiveness of the hot list are improved.

On the basis of the foregoing embodiment, the following describes in further detail a search-based hotspot recommendation method provided by an embodiment of the present invention with reference to fig. 3, where fig. 3 is a flowchart of a search-based hotspot recommendation method provided by an embodiment of the present invention, and as shown in fig. 3, the method includes:

s301, obtaining keywords corresponding to the hotspots to be sorted and the generation time of the keywords.

S302, a first search amount and a second search amount corresponding to the keywords are obtained, the first search amount is the search amount of the keywords in the current time period, the second search amount is the maximum search amount determined in the search amounts of the keywords corresponding to each time period in the previous N time periods, and N is an integer larger than 1.

The implementation manners of S301 and S302 are similar to those of S201 and S202, and are not described herein again

And S303, determining a search variation according to the first search amount and a third search amount, wherein the third search amount is the search amount of the keyword in the previous period.

The third search volume refers to the search volume of the keyword in the previous time period relative to the current time period, and the change rate of the keyword can be determined according to the first search volume of the keyword in the current time period and the third search volume in the previous time period, where the implementation manner of determining the search change rate can be seen in the following formula one:

wherein main _ pv is a first search amount, pre _ pv is a third search amount, and main _ pv _ change is a search variation.

The first expression of the above formula means that when the third search amount of the keyword in the previous time interval is 0 (which may be a new hotspot), the search variation of the keyword is the first search amount of the keyword in the current time interval, and when the third search amount of the keyword in the previous time interval is not 0, the search variation of the keyword is the search amount variation rate between the two time intervals.

S304, determining a first heat index according to the search variation, the first search quantity and the smoothing function.

An implementation manner of determining the first heat index may be seen in the following formula two:

index1 ═ main _ pv ═ ln (1+ (1+ main _ pv _ change)) formula two

Wherein main _ pv is a first search amount, main _ pv _ change is a search variation amount, index1 is a first heat index, and ln (…) is a smoothing function.

In the second formula, the search change rate is processed by setting a smoothing function, so that the situation that the popularity is high continuously due to the fact that the overall popularity value of a certain keyword is too high when the search change rate of the keyword is too high can be avoided.

The first popularity index in this embodiment is determined according to the first search volume of the keyword in the current time period and the third search volume in the previous time period, and is determined by combining the change rate of the first search volume in the current time period and the third search volume in the previous time period and the smoothing function, so that the first popularity index can reflect the current state of the hotspot corresponding to the keyword, and the situation that the hotspot existing in the hotspot list exists for a long time is avoided.

And S305, determining the peak value variation according to the second search quantity and the third search quantity.

The peak variation may be used to indicate the rate of change of the maximum search amount, and the implementation manner may be shown in the following formula three:

wherein, peak _ pv is a second search amount, pre _ pv is a third search amount, and peak _ pv _ change is a peak variation.

The meaning of formula three is similar to that of formula one, except that the peak variation calculated by formula three reflects the variation of the maximum search volume determined in the search volumes of the keywords corresponding to each time interval in the previous N time intervals, and the variation trend of the heat of the keyword can be determined by determining the variation of the maximum search volume.

And S306, determining a second heat index according to the peak value variation, the first search amount and the smoothing function.

The second heat index may be implemented according to the following formula four:

index2 ═ main _ pv ═ ln (1+ (1+ peak _ pv _ change)) formula four

Here, main _ pv is a first search amount, peak _ pv _ change is a peak variation amount, and ln (…) is a smoothing function.

The smoothing function acts on the equation two identically, and is not repeated here, the second heat index in this embodiment is determined according to the second search quantity that is the largest among the search quantities of the keyword in the previous N periods and the third search quantity in the previous period, and the second heat index is determined by combining the change rates of the second search quantity and the third search quantity and the smoothing function, so that the second heat index can reflect the state of the peak value of the hotspot corresponding to the keyword, so as to indicate the heat trend of the hotspot.

And S307, determining a time difference value according to the generation time and the current time.

And S308, determining a third heat index according to the time difference, a fourth search quantity and a time attenuation coefficient, wherein the fourth search quantity is the maximum search quantity determined in the search quantities corresponding to the keywords of each hot spot in the current time period.

For a new keyword, because the hotspot corresponding to the keyword just appears, the hotspot will not be reflected in the search engine, and the search volume of the new keyword is absent or very low, and it is also very difficult for the new keyword to appear in the hotspot list, so that the embodiment obtains the maximum search volume in the search volumes corresponding to the keywords of each hotspot in the current time period as the fourth search volume, and obtains the third heat index by combining the time difference, the fourth search volume, and the time decay coefficient, so that the new keyword can be given a relatively large heat, and the new keyword can enter the hotspot list quickly.

Specifically, an implementation manner of determining the third heat index may be seen in the following formula five:

index3＝max_pv*α^Δtformula five

Where Δ t is a time difference value, α … is a time decay coefficient, max _ pv is a fourth search amount, and index3 is a third heat index.

The formula four-way process combines the time decay coefficient to process the fourth search quantity of the keyword with the maximum search quantity in the current time period, so that a relatively large heat degree is given to the new keyword, the effect of preheating the new keyword in a search engine is achieved, and the problem that the heat degree of the new keyword cannot be quickly improved is avoided.

And after a period of time, the heat degree of the new keyword may be relatively high, due to the processing of the time decay coefficient, the new keyword can be gradually exited from the list with the passage of time, so as to effectively avoid that the keywords corresponding to some hotspots exist in the hotspot list for a long time, and enable the new keyword to be effectively listed, wherein the time decay coefficient can be used for adjusting the heat degree value and reflecting the heat degree aging state of the hotspot, and in a possible implementation manner, the range of the time decay coefficient is 0< α < 1.

S309, acquiring a heat index and a value according to the first heat index, the second heat index and the third heat index.

S310, determining the heat value of the hot spot to be sorted according to the heat index sum value, the fourth search quantity and the heat adjusting coefficient.

In this embodiment, the first heat index, the second heat index and the third heat index respectively reflect the heat indexes of the keywords from different dimensions, wherein the first heat index can reflect the current state of the hot spot corresponding to the keyword, the second heat index can reflect the change trend of the hot spot corresponding to the keyword, and the third heat index can enable a new keyword to obtain a larger heat value.

Specifically, the first heat index, the second heat index and the third heat index are added to obtain a heat index sum, and then the heat value of the hot spot to be sorted is determined according to the heat index sum, the fourth search volume and the heat adjustment coefficient, which can be implemented according to the following formula six:

where index1 is the first heat index, index2 is the second heat index, and index3 is the third heat index, (index1+ index2+ index3) is the sum of the heat indexes, α and β are coefficients for adjusting the magnitude range of the magnitude of the heat value, max _ pv is the fourth search amount, and final _ index is the heat value.

The heat value of the hot spot to be sorted is determined according to the heat index sum value, the fourth searching amount and the heat adjusting coefficient, meanwhile, the range of the magnitude of the heat value is adjusted through alpha and beta, the heat value of the hot spot to be sorted can be made to accord with the searching amount in a period of time (because the heat value is determined according to the searching amount and is adjusted through alpha and beta), and the magnitude of the heat value is made to be proper and not to be too high or too low.

It will be understood by those skilled in the art that the above formulas are not limited to the implementation of the parameters, and the formulas may be replaced according to actual requirements, for example, the smoothing function and the time attenuation coefficient may be selected according to actual requirements, but are not limited to the above manners.

By way of example, it is assumed that the first search volume of the keyword a in the current time period is 30, the second search volume is 80, the second search volume is the maximum search volume determined in the search volumes of the keywords corresponding to each of the previous N time periods, the third search volume in the previous time period is 10, and the fourth search volume is 40, the fourth search volume is the maximum search volume determined in the search volumes corresponding to the keywords of each hotspot in the current time period, and it can be seen from the search volumes of the keyword a that the hotspot corresponding to the keyword a is an already existing hotspot, and the heat of the hotspot is slowly reduced.

The search variation of the keyword a can be determined to be 2 according to formula one, and the first heat index of the keyword a can be determined to be 41.58 according to formula two; according to a third formula, the peak value variation of the keyword A can be determined to be 7, and meanwhile, according to a fourth formula, the second heat index of the keyword A can be determined to be 65.92; moreover, for convenience of explanation, assuming that the time attenuation coefficient and α and β are both 1, the third heat index of the keyword a may be determined to be 40 according to the fifth formula, and finally the heat value of the hot spot corresponding to the keyword a may be determined to be 23.32 according to the sixth formula.

The new keyword B appearing at the next time period of the keyword a is introduced by comparison, and assuming that the first search volume of the current time period of the keyword B is 80, the second search volume is 10, the third search volume of the previous time period is 10, and the fourth search volume is 80, it can be seen from each search volume of the keyword B that the search volume just begins to appear at the previous time period, and the search volume from the previous time period to the current time period is rapidly increased, it can be determined that the hotspot corresponding to the keyword B is a newly appearing hotspot, and the heat degree is continuously increased.

The search variation of the keyword B may be determined to be 7 according to formula one, and the first heat index of the keyword B may be determined to be 175.77 according to formula two; according to the formula three, the peak value variation of the keyword B can be determined to be 7, and according to the formula four, the second heat index of the keyword B can be determined to be 55.45; also for convenience of explanation, assuming that the time decay coefficient and α and β are both 1, the third heat index of the keyword B may be determined to be 80 according to the fifth formula, and finally the heat value of the hot spot corresponding to the keyword B may be determined to be 34.8 according to the sixth formula.

It can be seen that the keyword a which is already a hotspot and the heat degree of which is decreasing is decreased in the heat degree value, and the newly appearing keyword B which is increasing in the heat degree value is relatively high in the heat degree value, so that the embodiment can effectively reflect the heat degree condition of the hotspot according to the search amount through the above steps.

It should be noted that, in the above example, for convenience of calculation, the time attenuation coefficient and α and β are both set to 1 for explanation, and in the actual implementation process, the time attenuation coefficient can ensure that the new keyword is assigned with a larger heat value when the new keyword starts to appear, but as time goes by, the hot point of the new keyword will be gradually attenuated, so that a situation that the heat value of the old hot point is higher for a long period of time is avoided, and at the same time, α and β can effectively change the value range of the heat value, so that the range of the heat value can be limited in a required range, which can be set according to actual requirements, and details are not repeated here.

S311, according to the hot degree value of the hot spots to be sorted, sorting and displaying the hot spots to be sorted.

The implementation of S311 is similar to S204, and is not described herein again.

The hot spot recommendation method based on search provided by the embodiment of the invention comprises the following steps: and acquiring keywords corresponding to the hot spots to be sorted and the generation time of the keywords. And acquiring a first search quantity and a second search quantity corresponding to the keywords, wherein the first search quantity is the search quantity of the keywords in the current time period, the second search quantity is the maximum search quantity determined in the search quantities of the keywords corresponding to each time period in the previous N time periods, and N is an integer greater than 1. And determining a first heat index according to the search variation, the first search quantity and the smoothing function. And determining the peak value variation according to the second search quantity and the third search quantity. And determining a second heat index according to the peak variation, the first search amount and the smoothing function. And determining a time difference value according to the generation time and the current time. And determining a third heat index according to the time difference, a fourth search quantity and a time attenuation coefficient, wherein the fourth search quantity is the maximum search quantity determined in the search quantities respectively corresponding to the keywords of each hot spot in the current time period. And acquiring a heat index sum value according to the first heat index, the second heat index and the third heat index. And determining the heat value of the hot spot to be sorted according to the heat index sum value, the fourth search quantity and the heat adjusting coefficient. And sorting and displaying the hot spots to be sorted according to the hot value of the hot spots to be sorted. The method comprises the steps of determining that a first heat index reflects the current state of a keyword, determining that a second heat index reflects the heat trend of the keyword, determining a third heat index enables the heat corresponding to a new keyword to be distributed to a larger heat value, and determining the heat value of a hot spot to be ranked by combining the first heat index, the second heat index, the third heat index, a fourth search amount and a heat regulation coefficient, so that the heat value of the hot spot can reflect the search amount and can replace the new hot spot and the old hot spot in time, and the effectiveness of a hot spot list is guaranteed.

On the basis of the foregoing embodiment, after the hotspot list is acquired, the hotspot list can be updated according to the click rate model, so that the updated hotspot list can more accurately conform to the click condition of the user, which is described below with reference to fig. 4, where fig. 4 is a flowchart three of the hotspot recommendation method based on search provided by the embodiment of the present invention.

As shown in fig. 4, the method further includes:

s401, obtaining the hot degree characteristics of the keywords corresponding to the target hotspots in the hotspot list, wherein the ranks of the keywords are before the preset ranks.

In this embodiment, the heat characteristics of the keywords may be the parameters described in the above embodiments, such as the first heat index, the second heat index, the third heat index, the first search amount, the second search amount, and the like described in the above embodiments, and details are not repeated here.

Because the number of keywords that can be displayed in the page is limited, only the popularity value of the hotspot with the top ranking can be updated, so that the processing efficiency is improved, the current hotspot list is ranked by the hotspot to be ranked, and therefore the target hotspot with the top ranking before the preset ranking can be directly obtained according to the sequence of the hotspot list, for example, the hotspot list with the top ranking of 50 can be selected for adjustment.

S402, click rate information of the historical hotspot list is obtained, wherein the click rate information comprises a first click rate of each hotspot in a first time length before the current time and a second click rate of each hotspot in a second time length, the second time length is before the first time length, and the second time length is longer than the first time length.

The historical hotspot list is a hotspot list before the current time, for example, a user click log of the historical hotspot list can be obtained, so that click rate information of the historical hotspot list is obtained, wherein the click rate information includes a first click rate of each hotspot in a first duration before the current time and a second click rate of each hotspot in a second duration.

For example, 2 to 49 hours before the current time may be used as a first duration, and 1 hour before the current time may be used as a second duration, so as to obtain the click rate of each hotspot in the first duration and the second duration, respectively, where specific implementation manners of the first duration and the second duration may be selected according to actual requirements, as long as the second duration is before the first duration, and the second duration is greater than the first duration.

And S403, obtaining a training sample according to the second click rate of each hotspot in the second time length and the heat characteristics of the keywords corresponding to each hotspot.

S404, obtaining a test sample according to the first click rate of each hot spot in the first time length and the heat characteristics of the keywords corresponding to each hot spot.

Specifically, the second click rate of each hotspot in the second duration can be used as a training set, and a training sample is obtained according to the second click rate and the heat characteristics of the keywords corresponding to each hotspot; and taking the first click rate of each hotspot in the first duration as a test set, and obtaining a test sample according to the first click rate of each hotspot in the first duration and the heat characteristics of the keyword corresponding to each hotspot.

It should be noted that, during model training, the heat features of the keywords corresponding to each hotspot may include a fifth search amount and a total variation, where the fifth search amount is a total search amount of the search engine in a current time period, and the total variation is determined according to the fifth search amount and the third search amount, and the implementation manner of the determination may refer to the following formula seven:

wherein all _ pv is a fifth search amount, pre _ pv is a second search amount, and all _ pv _ change is a total change rate.

And the fifth variation and all the variation rates are used as the heat characteristics to train the model, so that the accuracy and comprehensiveness of model prediction can be improved.

S405, training the click rate model to be trained according to the training sample and the test sample to obtain the trained click rate model.

In a possible implementation manner, according to click rate information of a historical hotspot list recorded in a user click log, a keyword corresponding to a hotspot where the user clicks may exist may be taken as a positive sample, and a keyword corresponding to a hotspot where the user clicks does not exist may be taken as a negative sample, for example, an identifier corresponding to the positive sample may be recorded as 1, and an identifier corresponding to the negative sample may be recorded as 0.

The method comprises the steps of firstly training a model by taking a training sample as input of the model, wherein the training sample comprises a second click rate of each hotspot in a second duration and heat characteristics of keywords corresponding to the hotspots, the model learns by taking the second click rate and the heat characteristics of the keywords as input, and accordingly training of the model is achieved, and in the process of training the model according to the training sample, the model can clearly determine whether the training sample is a positive sample or a negative sample.

Secondly, after the model training is completed according to the training samples, the model can be trained according to the test samples, wherein the click rate model is processed according to the heat characteristics of the keywords corresponding to the hotspots, so that a prediction result (for example, a numerical value between 0 and 1) of whether the test sample is a positive sample or a negative sample is output, wherein the first click rate can indicate whether the test sample is the positive sample (1) or the negative sample (0), and the click rate model is adjusted and learned according to the indication of the first click rate and the prediction result so as to continuously train the model, thereby improving the accuracy of the output result.

In an optional implementation manner, for example, eXtreme Gradient Boosting (XGBoost) may be used as a classifier, and the click rate model is trained by a method of pair, so as to obtain a trained click rate model, where a specific training method of the model may refer to any possible implementation manner in the prior art, and is not described here again.

S406, inputting the heat characteristic into the click rate model, and obtaining a classification value output by the click rate model, wherein the classification value is used for indicating the click rate corresponding to the heat characteristic.

After the click rate model training is completed, the obtained heat characteristic of the keyword corresponding to the target hotspot is used as an output of the click rate model, and a classification value is output after the click rate model training is processed, where the classification value may be used to indicate a heat state of the target hotspot, and the classification value may be, for example, a value between 0 and 1, which specifically depends on the setting of the click rate model, which is not limited herein.

S407, reordering the target hotspots according to the classification value output by the click rate model to obtain an updated hotspot list.

After the click rate model processes each target hotspot, each target hotspot corresponds to a respective classification value, which may be, for example, a numerical value of 0.9, 0.87, and the like, and then the target hotspots are reordered according to a sequence of the classification values from small to large, so as to obtain an updated hotspot list.

In this embodiment, the updated hotspot list is the hotspot list considering the initial ranking, and meanwhile, the list is ranked according to the click rate of the user, so that the ranking of hotspots with a higher click rate in the hotspot list is higher, and the status of hotspots can be reflected more timely.

The hot spot recommendation method based on search provided by the embodiment of the invention comprises the following steps: and acquiring the heat characteristics of the keywords corresponding to the target hotspots with the ranks before the preset ranks in the hotspot list. The click rate information of the historical hotspot list is obtained, the click rate information comprises a first click rate of each hotspot in a first time length before the current time and a second click rate of each hotspot in a second time length, the second time length is before the first time length, and the second time length is longer than the first time length. And obtaining a training sample according to the second click rate of each hotspot in the second duration and the hot degree characteristics of the keywords corresponding to each hotspot. And obtaining a test sample according to the first click rate of each hotspot in the first time length and the heat characteristics of the keywords corresponding to each hotspot. And training the click rate model to be trained according to the training sample and the test sample to obtain the trained click rate model. And inputting the heat characteristics into the click rate model, and acquiring the classification value output by the click rate model. And re-sequencing the target hotspots according to the classification value output by the click rate model to obtain the updated hotspot list. The click rate model is trained according to the click rate information, so that the click rate model can effectively and accurately output classification results corresponding to the click rate, then the hot spot list is processed according to the click rate model to output a classification value, and the ranking of the hot spot list is updated according to the classification value, so that the hot spot list can be ranked in combination with the click rate information, and the hot degree condition of the hot spot is reflected more accurately.

Fig. 5 is a schematic structural diagram of a hot spot recommendation device based on search according to an embodiment of the present invention. As shown in fig. 5, the apparatus 50 includes: an acquisition module 501, a determination module 502, and a sorting module 503.

An obtaining module 501, configured to obtain a keyword corresponding to a hotspot to be sorted and generation time of the keyword;

the obtaining module 501 is further configured to obtain a first search amount and a second search amount corresponding to the keyword, where the first search amount is a search amount of the keyword in a current time period, the second search amount is a maximum search amount determined in the search amounts of the keyword corresponding to each of previous N time periods, and N is an integer greater than 1;

a determining module 502, configured to determine a heat value of the hot spot to be ranked according to the first search volume, the second search volume, the generation time, and the current time;

the sorting module 503 is configured to sort and display the hotspots to be sorted according to the heat value of the hotspot to be sorted.

In one possible design, the determining module 502 is specifically configured to:

determining a first heat index according to the first search quantity and a third search quantity, wherein the third search quantity is the search quantity of the keyword in the previous period; (ii) a

in one possible design, the sorting module 503 is further configured to:

In one possible design, the obtaining module 501 is further configured to:

The apparatus provided in this embodiment may be used to implement the technical solutions of the above method embodiments, and the implementation principles and technical effects are similar, which are not described herein again.

Fig. 6 is a schematic diagram of a hardware structure of a search-based hotspot recommendation device according to an embodiment of the present invention, and as shown in fig. 6, a search-based hotspot recommendation device 60 according to this embodiment includes: a processor 601 and a memory 602; wherein

A memory 602 for storing computer-executable instructions;

the processor 601 is configured to execute computer-executable instructions stored in the memory to implement the steps performed by the search-based hotspot recommendation method in the foregoing embodiments. Reference may be made in particular to the description relating to the method embodiments described above.

Alternatively, the memory 602 may be separate or integrated with the processor 601.

When the memory 602 is separately provided, the search-based hotspot recommendation device further comprises a bus 603 for connecting the memory 602 and the processor 601.

An embodiment of the present invention further provides a computer-readable storage medium, where a computer execution instruction is stored in the computer-readable storage medium, and when a processor executes the computer execution instruction, the method for recommending a hotspot based on a search, performed by the above hotspot recommendation device based on a search, is implemented.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described device embodiments are merely illustrative, and for example, the division of the modules is only one logical division, and other divisions may be realized in practice, for example, a plurality of modules may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or modules, and may be in an electrical, mechanical or other form.

The integrated module implemented in the form of a software functional module may be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present application.

It should be understood that the Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.

The memory may comprise a high-speed RAM memory, and may further comprise a non-volatile storage NVM, such as at least one disk memory, and may also be a usb disk, a removable hard disk, a read-only memory, a magnetic or optical disk, etc.

The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.

The storage medium may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.

Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.

Finally, it should be noted that: the above embodiments are only used for illustrating the technical solutions of the embodiments of the present invention, and are not limited thereto; although embodiments of the present invention have been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the embodiments of the present invention.

Claims

1. A hot spot recommendation method based on search is characterized by comprising the following steps:

determining a third heat index according to the generation time, the current time and a fourth search quantity, wherein the fourth search quantity is the maximum search quantity determined in the search quantities corresponding to the keywords of each hot spot in the current time period;

determining the heat value of the hot spot to be sorted according to the first heat index, the second heat index, the third heat index and the fourth search quantity;

and sequencing and displaying the hot spots to be sequenced according to the hot degree value of the hot spots to be sequenced.

2. The method of claim 1, wherein determining a first heat index according to the first search volume and a third search volume of the keyword in a previous time period comprises:

3. The method of claim 1, wherein determining a second heat index based on the second search volume and the third search volume comprises:

4. The method of claim 1, wherein determining a third heat index based on the generation time, the current time, and a fourth search amount comprises:

5. The method of claim 1, wherein determining the heat value of the hotspot to be ranked according to the first heat index, the second heat index, the third heat index and the fourth search amount comprises:

6. The method according to any one of claims 1 to 5, wherein after the sorting and displaying the hot spots to be sorted according to the hot value of the hot spot to be sorted, the method further comprises:

7. The method of claim 6, wherein before inputting the heat characteristic into the click-through rate model and obtaining the classification value output by the click-through rate model, the method further comprises:

8. A search-based hotspot recommendation device, comprising:

the obtaining module is further configured to obtain a first search amount and a second search amount corresponding to the keyword, where the first search amount is a search amount of the keyword in a current time period, the second search amount is a maximum search amount determined in the search amounts of the keyword corresponding to each of the previous N time periods, and N is an integer greater than 1;

a determination module to:

9. The apparatus of claim 8, wherein the determining module is specifically configured to:

10. The apparatus of claim 8, wherein the determining module is specifically configured to:

11. The apparatus of claim 8, wherein the determining module is specifically configured to:

12. The apparatus of claim 8, wherein the determining module is specifically configured to:

13. The apparatus of any of claims 8 to 12, wherein the sorting module is further configured to:

14. The apparatus of claim 13, wherein the obtaining module is further configured to:

15. A search-based hotspot recommendation device, comprising:

a memory for storing a program;

a processor for executing the program stored by the memory, the processor being configured to perform the method of any of claims 1 to 7 when the program is executed.

16. A computer-readable storage medium comprising instructions which, when executed on a computer, cause the computer to perform the method of any one of claims 1 to 7.