Third-party prediction method for activeness of news clientTechnical Field
The invention relates to the field of statistics, in particular to a third-party news client liveness prediction method.
Background
News information is one of the most concerned industry applications in the internet industry, news clients are more endless, and in order to evaluate the influence of news media, the evaluation of the news client is more important under the current internet. The one-way declaration mode of the number of active users of each news client lacks a public and uniform measurement scale for upstream and downstream users.
Disclosure of Invention
The invention aims to provide a third-party news client liveness prediction method, so that the problems in the prior art are solved.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a third-party prediction method for activeness of news client comprises the following steps,
a third-party prediction method for activeness of news client comprises the following steps,
s1, acquiring news contents of each news client by using a crawler, setting sampling constants of each news client according to the difference of postings numbers of different news clients, wherein all the news clients follow the same sampling period;
s2, defining draft sending quantity, total reading number, APP reading number balance factor, draft sending number curvature in unit time, reading number curvature in unit time, manuscript comment number and comment number in unit time according to the obtained news content;
and S3, predicting the liveness of each news client by adopting a liveness numerical formula according to the parameters defined in the step S2.
Preferably, the contribution amount is the sum of the number of contributions issued by a certain news client in a sampling period; the manuscript refers to an article visible in the news client list and is defined as Ps; the reading total is the sum of all manuscripts of the news client in a sampling period and is defined as Vs; the APP reading number balance factor is a balance factor for fitting the reading number of the news client to a uniform reference and is defined as Avgs; the curvature of the number of releases in unit time is the curvature of the number of releases in unit time of the news client in a sampling period, and is defined as Dpr, and the value is taken through the following formula,
the curvature of the reading number in unit time is the curvature of the reading number of the news client in unit time in a sampling period, and is defined as Vpr, and the value is taken through the following formula,
the number of the comments of the manuscript is the sum of the number of the comments of all original manuscripts of the news client; the number of reviews in unit time is the number of reviews of the news client in unit time in a sampling period, and is defined as Cpr, and the Cpr is taken as the value through the following formula,
preferably, the numerical formula of the activity is as follows,
wherein, Dau is the liveness of the news client; rri is a penalty coefficient; i is a calculation period; maxi is the maximum number of activities in a calculation period; mini is the minimum number of activities in a calculation period.
Preferably, the penalty factor takes the following values,
where x represents the review browsing activity ratio.
Preferably, x is calculated by the following formula,
wherein y represents the ratio of the manuscript browsing liveness; cr represents review liveness; dr represents the activity of manuscript sending; vr represents browsing liveness.
Preferably, the comment liveness, the draft liveness and the browsing liveness are respectively obtained by the following formulas,
wherein,c represents the number of comments of a single article; d represents the number of articles; v represents a single article reading number; j represents adoptSample period; os represents the original manuscript quantity.
The invention has the beneficial effects that: 1. by the method, universal prediction can be performed on any news client. 2. The method can smoothly and accurately adjust the self-adaptive parameters according to different clients, and avoids the problem that the transverse contrast of the statistical data of a single client cannot be measured. 3. By adopting the machine learning-based method, the vitality prediction of the news client is realized, and news workers, advertisement delivery workers and public opinion workers can further utilize the prediction result to predict the working effect in advance.
Drawings
FIG. 1 is a flow chart illustrating a prediction method according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
As shown in fig. 1, the present invention provides a third party news client liveness prediction method, which includes the following steps,
s1, acquiring news contents of each news client by using a crawler, setting sampling constants of each news client according to the difference of postings numbers of different news clients, wherein all the news clients follow the same sampling period;
s2, defining draft sending quantity, total reading number, APP reading number balance factor, draft sending number curvature in unit time, reading number curvature in unit time, manuscript comment number and comment number in unit time according to the obtained news content;
and S3, predicting the liveness of each news client by adopting a liveness numerical formula according to the parameters defined in the step S2.
In this embodiment, the contribution amount is the total number of contributions issued by a certain news client in a sampling period; the manuscript refers to an article visible in the news client list and is defined as Ps; the reading total is the sum of all manuscripts of the news client in a sampling period and is defined as Vs; the APP reading number balance factor is a balance factor for fitting the reading number of the news client to a uniform reference and is defined as Avgs; the curvature of the number of releases in unit time is the curvature of the number of releases in unit time of the news client in a sampling period, and is defined as Dpr, and the value is taken through the following formula,
the curvature of the reading number in unit time is the curvature of the reading number of the news client in unit time in a sampling period, and is defined as Vpr, and the value is taken through the following formula,
the number of the comments of the manuscript is the sum of the number of the comments of all original manuscripts of the news client; the number of comments in unit time is the number of comments in unit time of the news client in a sampling period, and is defined as Cpr, the value of the Cpr is obtained by taking down a formula,
in this embodiment, the activity numerical formula is as follows,
wherein, Dau is the liveness of the news client; rri is a penalty coefficient; i is a calculation period; maxi is the maximum number of activities in a calculation period; mini is the minimum number of activities in a calculation period.
In this embodiment, the penalty factor takes the following values,
where x represents the review browsing activity ratio.
In this embodiment, x is calculated by the following formula,
wherein y represents the ratio of the manuscript browsing liveness; cr represents review liveness; dr represents the activity of manuscript sending; vr represents browsing liveness.
In this embodiment, the review liveness, the submission liveness and the browsing liveness are respectively obtained by the following formulas,
wherein,c represents the number of comments of a single article; d represents the number of articles; v represents a single article reading number; j represents the sampling period; os represents the original manuscript quantity.
By adopting the technical scheme disclosed by the invention, the following beneficial effects are obtained:
the invention provides a third-party news client liveness prediction method, which can be used for universally predicting any news client; the method can smoothly and accurately adjust the self-adaptive parameters according to different clients, and avoids the problem that the transverse contrast of the statistical data of a single client cannot be measured; meanwhile, the machine learning-based method is adopted, so that the vitality prediction of the news client is realized, and news workers, advertisement delivery workers and public opinion workers can further utilize the prediction result to predict the working effect in advance.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and improvements can be made without departing from the principle of the present invention, and such modifications and improvements should also be considered within the scope of the present invention.