Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description refers to the specific implementation, structure, characteristics and effects of a digital product generating method and system according to the invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of a digital product generation method and system provided by the invention with reference to the accompanying drawings.
Embodiment one:
referring to fig. 1, a flowchart of a digital product generating method according to an embodiment of the present invention is shown, including:
step S101, acquiring first access data of a first user in a predetermined period of time and second access data of a second user in the predetermined period of time.
The access data comprises access time, access times and access duration of each user for accessing each digital product in a preset time period.
Specifically, the predetermined period of time may be a period of time elapsed, such as N consecutive days elapsed, where N may be an integer greater than 1, and the first access data includes an access time, an access number, and an access duration for each first user to access each digital product within the predetermined period of time, and the second access data includes an access time, an access number, and an access duration for each second user to access each digital product within the predetermined period of time. The access times refer to the total access times of a certain second user for accessing a certain digital product in a preset time period, the access duration refers to the duration of a single access of the certain second user for the certain digital product in the preset time period, and the access time refers to the moment of a word access of the certain second user for the certain digital product in the preset time period.
The access data of each first user and second user in the past period of time can be further obtained from the applet management background, wherein the access data comprises the access data of each user to at least one digital product, for example, the access data of a certain user to a certain digital product in a preset period of time is thatWherein, the method comprises the steps of, wherein,Indicating that the number of accesses is 120,Which means that the 2023, 3, 13, 14 points 25 minutes and 32 seconds start the visit, the visit duration is 3 minutes,It indicates that the access is started at 10 minutes, 15 minutes and 11 seconds on day 14 of month 2023, and the access time is 2 minutes.
Further, in order to facilitate analysis of the access data of the second user, after the second access data of the second user in the predetermined period of time is obtained, coarse clustering may be performed on the second user based on a K clustering algorithm, so as to cluster the second users similar to the accessed digital products.
Specifically, a K clustering algorithm (K-means clustering algorithm) is used for coarse clustering of the second users, the second users similar to the accessed digital products are clustered, and independent analysis can be carried out for each cluster during subsequent analysis, so that accuracy of analysis results is improved. In the process of performing coarse clustering on the second users by using the K-means clustering algorithm, the distance index between the second users can be based on the digital products in the access data of the second usersThe correlation coefficient is determined, the optimal cluster number of the K-means clustering algorithm can be determined according to a contour coefficient method, wherein the contour coefficient method refers to that for each possible cluster number K, the K-means clustering algorithm is operated, the overall contour coefficient is calculated, and the K value which enables the contour coefficient to be the maximum is found out, and the K value is the optimal cluster number. After the clustering is completed, the digital products accessed by the second user in each cluster are similar, so that independent analysis can be performed by taking the clusters as units, and the accuracy of the subsequent analysis results is improved.
Step S103, determining the first access necessity of the second user for accessing each digital product according to the access time, the access times and the access time length of the second user.
Wherein the first access certainty indicates a degree of interest of the second user in accessing the digital product, the first access certainty being proportional to the degree of interest.
Specifically, the first access necessity can quantify the interestingness, the contingency and whether the user accesses the digital products or not, and the corresponding access time, the access times and the access duration of each digital product of each user are obtained in the steps, so that the access process of each digital product has time sequence characteristics according to the access sequence, the second access necessity of each digital product in the digital products accessed every day can be determined according to the time sequence characteristics of each digital product, the second access necessity can reflect the access randomness and the browsing depth of the digital product, and the larger the second access necessity is, the higher the randomness of the user accessing the digital product is and the no deep browsing is indicated. Based on the second access certainty, a first access certainty of each user's access to each digital product may be determined. The greater the first access certainty, the greater the degree of interest of the second user in the digital product, so that the problem of poor recommendation effect caused by accidental access of the user to the digital product can be eliminated.
Step S105, determining a similar user similar to the first user access habit from the second users according to the first access data and the second access data.
Specifically, when determining similar users that are similar to the access habits of the first user, collaborative filtering algorithms may be used to calculate similar users that are similar to the access habits of the first user. Specifically, by using the method, the first access certainty of each second user to each digital product accessed by the second user is calculated to construct an evaluation matrix in a collaborative filtering algorithm, and the set of the second users is set asThe set of digital products corresponding to the second user is. The corresponding evaluation matrix is:
wherein the method comprises the steps ofRepresenting the first access certainty of the second user m to the digital product k. And then, calculating the similarity between the first user and the second user by using a collaborative filtering algorithm and utilizing the digital product and access data accessed by the first user and the digital product and access data accessed by the second user, wherein the similarity can be represented by Euclidean distance, a Pearson correlation coefficient, cosine similarity and the like, and after the similarity is calculated, sorting the similarity in a descending order, and selecting the second user with n digits before the sorting of the similarity as the similar user. It should be noted that the collaborative filtering algorithm may refer to a known technology, and the embodiments of the present application are not described herein.
In step S107, a target digital product is determined from among the digital products accessed by the similar users according to the first access certainty, and the target digital product is recommended to the first user.
Specifically, after determining the first access necessity, the product recommendation can be directly performed according to the first access necessity of the similar user for accessing the digital products, that is, the first access necessity of the similar user for accessing each digital product is sorted in descending order, the digital product with the highest first access necessity for accessing each digital product in the similar user is selected to be recommended to the first user as the target digital product, or the digital product with the first access necessity corresponding to the first access necessity of the N bits before sorting is selected to be recommended to the first user as the target digital product. Further, predictive scores of digital products accessed by similar users can be calculated, and product recommendation is performed according to the predictive scores. During subsequent use of the applet by the first user, the recommended digital product may be displayed on the hot-sell page of the applet. Wherein N is an integer greater than 0.
In one possible implementation, a predictive score for a visited digital product of a similar user is calculated from a first visit certainty based on a collaborative filtering algorithm; and sorting the predictive scores in a descending order, and determining the digital products with the top N digits and the digital products which are not accessed by the first user as target digital products. Wherein N is an integer greater than 0. Specifically, the first access certainty can reflect the interestingness of the user for accessing the digital product, the prediction scores of the digital products accessed by similar users can be calculated according to the interestingness, for example, the prediction scores are calculated according to the weights of the digital products accessed by the similar users and the first access certainty of each digital product, for example, the weights and the first access certainty are multiplied to be used as the prediction scores, the prediction scores can reflect the recommendation degree of the digital product recommended to the first user, so that the preference degree of the first user is represented, the digital products with N bits before the ranking are selected, and the digital products which are not accessed by the first user are used as target digital products, so that the user experience is improved.
Further, determining the first access certainty of the second user to access each digital product according to the access time and the access duration of the second user includes:
And taking each cluster as a user group, and determining first access certainty of the second user accessing each digital product in each cluster according to the access time and the access duration of the second user in each cluster.
Specifically, after the second users are clustered, users with similar access habits are clustered into a cluster, and then each cluster is used as a user group to analyze the second users in the cluster, so that the calculation accuracy of the first access certainty is improved.
According to the technical scheme disclosed by the embodiment of the application, the access necessity of the second user for accessing each digital product is quantified through the acquired access time, access times and access duration of the second user, the access necessity can indicate the interest level of the second user for accessing the digital product, and the greater the access necessity is, the greater the interest level of the user for accessing the digital product is. And then determining a similar user similar to the access habit of the first user according to the access data of the first user and the access data of the second user, determining a target digital product from the digital products accessed by the similar user based on the access necessity of the second user to the digital products, and recommending the target digital product to the first user.
Embodiment two:
referring to fig. 2, a flowchart of another digital product generation method according to an embodiment of the present invention is shown, including:
step S201, acquiring first access data of a first user in a predetermined period of time and second access data of a second user in a predetermined period of time.
The access data includes access time, access times, and access duration for each user to access each digital product for a predetermined period of time.
Step S2030, determining a first access time and a first access duration of the second user for accessing each digital product every day in a predetermined period according to the access time, the access times and the access duration of the second user for each digital product.
The first access duration is the access duration of each access of the second user to each digital product in each day.
Specifically, each digital product of the second user has a corresponding access time, access times and access duration, so that the access process of each digital product has a time sequence characteristic according to the access sequence, and the access necessity of each digital product is determined according to the time sequence characteristic of each digital product. If access data of each second user in s days is obtained, the first access times are all access times of the second user for accessing the a-th digital product in s days, and the first access time length is the access time length in the i-th access of the a-th digital product in s days.
Step S2031, determining a second access necessity of each digital product accessed by the second user every day according to the first access times and the first access duration.
The second access certainty indicates the contingency of the user's access to the digital product, the second access certainty being proportional to the contingency.
Specifically, the total number of accesses to all the digital products on the s-th day can be calculated according to the number of accesses to the a-th digital product by the second user on the s-th day. The greater the second access certainty, the higher the randomness of the user accessing the digital product and the greater the contingency of the user accessing the digital product. According to the acquired access data of the second user, the digital products accessed every day are in sequence according to the access time, so that the access time fluctuation series of all the digital products every day are firstly acquired according to the access sequence. The second access certainty of each digital product in the digital products accessed every day is determined by analyzing the fluctuation condition of the access time length of each digital product in the access time length fluctuation series in multiple accesses, assuming that the access data of the second user in s days is acquired, taking the access time length fluctuation series of the s th day as an example.
In one possible implementation, determining, based on the first number of accesses and the first access duration, a second access certainty of each digital product accessed by the second user each day includes: superposing the first access times to obtain the total access times of the second user for accessing all the digital products in each day; calculating the average value of the first access time length of all digital products in each day; calculating a second difference value between the first access time length and the average value, traversing all the first access times, and superposing the second difference value to obtain a second superposition value; calculating a first ratio of the second superimposed value to the first access times and a second ratio of the first access times to the total access times; and carrying out index operation on the first ratio value, and determining that the product of the index operation result and the second ratio value is the second access certainty of the digital product.
Specifically, assuming that access data of each second user for s days is obtained together, the digital products accessed every day are in sequence according to the access time, and therefore the access time duration fluctuation series of all the digital products every day are obtained according to the access sequence first. Taking the access duration fluctuation sequence of the s day as an example, determining the second access certainty of each digital product in the digital products accessed every day by analyzing the fluctuation condition of the access duration of each digital product in multiple accesses on the access duration fluctuation sequence, wherein the second access certainty can be calculated by adopting the following formula:
In the above-mentioned method, the step of,Representing a second access certainty of the a-th digital product in the access data of the s-th day.Representing the first number of accesses of the a-th digital product in the access data of all the second users on the s-th day.Indicating the total number of accesses to all digital products on day s.Representing an exponential function.Representing the first access duration in the ith access of the a-th digital product on the s-th day.Representing the average of all first access durations for all digital products on day s.The larger the value of (c) indicates the higher the first access number of the a-th digital product in the access data of the s-th day, and the higher the attention degree of the a-th digital product in the access data of the s-th day.The larger the value of the (a) is, the longer the first access duration in the ith access of the (a) th digital product in the access data of the(s) th day is than the normal access duration, and the higher the possibility that the ith access of the (a) th digital product in the corresponding access data of the(s) th day belongs to accidental access is. Thus (2)The larger the value of (2) is at the same timeThe larger the value of (c) indicates that the more likely the access to the a-th digital product in the access data on the s-th day is to be accidentally accessed, the greater the corresponding second access must be.
Further, since the access times of the a-th digital product in the access data of the s-th day have time sequence, as for the a-th digital product, if the access time length of the a-th digital product shows the upward trend characteristic along with the time sequence, the a-th digital product is free from accidental access, so that the calculated access certainty of the a-th digital product in the access data of the s-th day needs to be corrected according to the characteristic, and the authenticity and reliability of the second access certainty are improved. In one possible implementation, after determining the second access necessity of each digital product accessed by the second user every day according to the first access times and the first access duration, the method further includes: according to the time sequence of the access time, determining a first fluctuation curve of a first access duration of each digital product accessed by a second user every day in a preset time period and a first fitting curve corresponding to the first fluctuation curve; determining second access times, wherein the second access times are second access times which are larger than access time lengths of all digital products in first access time lengths of all the digital products accessed by a second user every day; determining a correction coefficient according to the slope value of the first fitting curve, the first access time length, the access time length during the first access and the second access times; and correcting the second access necessity by using the correction coefficient to obtain the real access necessity.
Specifically, taking a fluctuation curve of the first access duration of the a-th digital product in the access data of the s-th day in the acquired predetermined time period as an example, as shown in fig. 3, the fluctuation curve represents a first fluctuation curve and a first fitting curve, after the fluctuation curve of the first access duration of the a-th digital product in the access data of the s-th day is acquired, a polynomial fitting is used to acquire a fitting curve of the first fluctuation curve, and the fitting curve can represent the overall change trend of the access duration of the a-th digital product in the access data of the s-th day. Wherein the slope value of the fitted curve can be represented by the mean of the slope values of all points on the fitted curve. However, the slope value of the fitted curve can only represent the overall trend of the access time of the a-th digital product, but cannot represent the trend of detail. As shown in fig. 4, the slope of the first fitted curve is negative at this time, which indicates that the overall trend of variation of the access duration of the digital product is downward. However, the access duration that is significantly higher than the initial access duration occurs during the fluctuation of the access duration of the first fluctuation curve, which may be caused by sudden interest or the like, so that the trend of the access duration corresponding to this case increases accordingly. Therefore, the change trend can be corrected by fitting all the access time durations which are larger than the initial access time duration on the curve to the initial access time duration. It should be noted that the wavy curve in fig. 3 and fig. 4 is the first wavy curve in the present embodiment, and the fitted curve is the first fitted curve in the present embodiment.
In one possible implementation, determining the correction factor according to the slope value of the first fitted curve, the first access duration, the access duration at the first access, and the second access number includes: calculating a first difference value of each first access time length and the access time length in the first access, traversing all second access times, and superposing the first difference values to obtain a first superposition value; normalizing the first superposition value to obtain a normalized superposition value; and calculating the sum of the normalized superimposed value and the slope value to obtain a sum, and taking the sum as a correction coefficient.
Specifically, taking a fluctuation curve of the first access duration of the a-th digital product in the access data of the s-th day in the predetermined time period as an example, the correction coefficient may be calculated by the following formula:
In the above-mentioned method, the step of,A correction factor representing the second visit certainty of the a-th digital product in the visit data on the s-th day.Representing the normalization function.The slope value of the first fitted curve representing the first visit duration of the a-th digital product in the visit data on the s-th day is expressed as the slope average of all points on the first fitted curve.And the second access times which are longer than the access time length of the 1 st access in all the first access time lengths of the a-th digital products in the access data of the s day are represented.The larger the value of (c) is, the larger the corresponding degree of correction of the trend of variation of the first access duration is. And then correspond toThe greater the correction factor representing the second visit certainty of the a-th digital product in the visit data on the s-th day. After the correction coefficient of the second access necessity of the a-th digital product in the access data of the s-th day is calculated, the original second access necessity is corrected.
The correcting the second access necessity by using the correction coefficient may be normalizing the correction coefficient to obtain a normalized correction coefficient, and calculating a product of the normalized correction coefficient and the second access necessity to obtain the corrected real access necessity. Further, the correction of the second access certainty is specifically as follows:
In the above-mentioned method, the step of,Indicating the actual access certainty of the a-th digital product in the access data of the s-th day.Representing a second access certainty of the a-th digital product in the access data of the s-th day.Representing the normalization function.A correction factor representing the second visit certainty of the a-th digital product in the visit data on the s-th day.
Step S2032, determining the interest level of the second user in each digital product using the second access certainty.
Specifically, after determining the second visit necessity of each digital product every day, determining a fluctuation curve of the second visit necessity of each digital product on different days, and determining a fitting curve corresponding to the fluctuation curve, wherein the fitting curve can be obtained through polynomial fitting, and further, the slope value of the fitting curve is determined, and the slope value of the fitting curve can be represented by the slope average value of all points on the fitting curve; under the condition that the slope value of the fitted curve is not smaller than a preset value, determining the interestingness as the slope value; and under the condition that the slope value of the fitting curve is smaller than a preset value, determining the interestingness to be the preset value, wherein the preset value can be 0.
In one possible implementation, the corrected real access certainty may also be employed to determine the interestingness, determining the interest level of the second user in each digital product using the second access certainty includes: and determining the interest degree of the second user on each digital product by using the real access certainty.
Specifically, determining the interest level of the second user in each digital product using the real access certainty includes: determining a second fluctuation curve of the real access necessity of a second user to each digital product on different days, and determining a second fitting curve corresponding to the second fluctuation curve; determining a slope value of the second fitted curve; under the condition that the slope value of the second fitting curve is not smaller than a preset value, determining the interestingness as the slope value; and under the condition that the slope value of the second fitting curve is smaller than a preset value, determining the interestingness as the preset value.
Specifically, the second visit certainty of each digital product visited each day can be calculated according to the above-described embodiments. Since each digital product may be accessed on other days, such as food, video affiliates, etc. The first access certainty of each digital product for each user is thus determined by analyzing the varying relationship of the actual access certainty of each digital product on different days. For example, for the nth digital product of the mth user, the second fluctuation curve of the real access certainty of the nth digital product of the mth user on different days is acquired according to the time sequence access of the days, and the corresponding second fitting curve is acquired, wherein the second fitting curve can be obtained by adopting polynomial fitting. As the mth user becomes familiar with a certain digital product, the access time length of the mth user to the same digital product gradually decreases and becomes stable with time. The visit certainty value will then also decrease and tend to stabilize, with the slope of the corresponding fitted curve being negative. If a digital product is interested in the beginning, the corresponding access time length and the access times of the digital product are increased, the corresponding access necessity value is also increased, and the corresponding slope of the fitting curve is positive. Thus, in order to better represent the user's interest in the new digital product, the first access certainty of the nth digital product of the mth user is quantified by the real access certainty fluctuation curve.
The slope value of the second fitting curve may be represented by a slope average value of all points on the fitting curve, the predetermined value may be 0, and the interestingness may be represented by the following formula:
In the above-mentioned method, the step of,Representing the interest level of the mth user in the nth digital product, whereinSlope value of a second fitted curve corresponding to a second fluctuation curve representing the real visit certainty of an nth digital product of an mth user on different days whenIs positive, the interestingness isAnd is itself otherwise 0.The larger the value of (c) indicates the higher the interest in the access of the mth user to the nth digital product, the greater the first access certainty of the corresponding mth user to the nth digital product.
Step S2033, determining the number of days of access to each digital product by the second user according to the access time.
Specifically, when the second user accesses each digital product, the second user corresponds to each access time, and the total access days of the second user to each digital product can be determined according to the access time. For example, the mth user has access to the nth digital product for the respective times of，，The total number of days of access for the mth user to the nth digital product is determined to be 3 days.
Step S2034, determining the first access necessity based on the number of days of access and the interestingness of the second user to each digital product.
Specifically, the access duration of each digital product by the second user varies with the familiarity of the user with the digital product, and the access days of the same digital product by each second user are different, so that the difference between different access days and the interest level of each digital product can be used to determine the first access certainty.
In one possible implementation, determining the first access necessity based on the number of days and the interestingness of access to each digital product by the second user comprises: ascending order of the access days of each digital product is carried out, and an ordering sequence is obtained; calculating a variance of a third difference between adjacent access days in the ordered sequence; calculating a third ratio of the number of days of access to the digital product to the variance; overlapping the interestingness with a preset value to obtain a third overlapping value; and multiplying the third ratio by the third superposition value to obtain a product, and normalizing the product to obtain the first access certainty.
Specifically, the preset value may be 1, the number of access days for each digital product refers to the number of access days for each second user for each digital product, and the first access necessity is described by taking the nth digital product accessed by the mth user as an example, where the first access necessity may be represented by the following formula:
In the above-mentioned method, the step of,Representing a first access certainty of an nth digital product of an mth user.The normalization function is shown.Indicating the total number of days of access to the nth digital product for the mth user.The access regularity feature representing the total number of days of access of the mth user to the nth digital product viewing is expressed by the variance of the differences between adjacent days of the total number of access of all second users to the nth digital product viewing, for example, the total number of access of all second users to the nth digital product viewing is arranged in ascending order such that the number of access days is [2,3,5,7,8,9], and the differences between adjacent days are [1,2, 1].Representing the interest level of the mth user in the nth digital product,The larger the value of (c) is, the more the mth user has access to the nth digital product, whileThe smaller the value of (c) indicates that the access to the nth digital product by the mth user exhibits regularity, indicating that the mth user has a higher interest in the nth digital product.Is equivalent to a pair ofWhen (1) is modified byAfter the determination is made, the user can,The larger the value of (c) indicates the higher the interest in the access of the mth user to the nth digital product, the greater the first access certainty of the corresponding mth user to the nth digital product. Thus (2)The larger the value of (c) indicates the greater the access certainty of the nth digital product of the mth user, the greater the access certainty indicates the greater the interest in the nth digital product of the mth user.
Step S205, determining a similar user similar to the first user access habit from the second users according to the first access data and the second access data.
In step S207, a target digital product is determined from digital products accessed by similar users according to the first access certainty, and the target digital product is recommended to the first user.
It should be noted that, the step S201, the step S205 and the step S207 have the same or similar implementation manner as the step S101, the step S105 and the step S107 in the first embodiment, which can be referred to each other, and the embodiments of the present application are not repeated here.
According to the technical scheme disclosed by the embodiment of the application, the access necessity of the second user for accessing each digital product is quantified through the acquired access time, access times and access duration of the second user, the access necessity can indicate the interest level of the second user for accessing the digital product, and the greater the access necessity is, the greater the interest level of the user for accessing the digital product is. And then determining a similar user similar to the access habit of the first user according to the access data of the first user and the access data of the second user, determining a target digital product from the digital products accessed by the similar user based on the access necessity of the second user to the digital products, and recommending the target digital product to the first user, so that the influence of the digital products accidentally accessed by the second user on the subsequent recommendation results is eliminated, the reliability and accuracy of the recommendation results for recommending the digital products to the target user are improved, the recommendation effect is better, and therefore, the suitable digital products are recommended for the target user in a targeted manner, and the user experience is improved. Further, the authenticity and the reliability of the second access necessity are improved by correcting the second access necessity through the correction coefficient, the reliability and the authenticity of the first access necessity are further improved, the influence of the digital product which is accidentally accessed by the second user on the follow-up recommendation result can be eliminated to a greater extent, the reliability and the accuracy of the recommendation result for recommending the digital product for the first user are further improved, the recommendation effect is improved, and therefore the proper digital product is recommended for the first user in a targeted mode, and the user experience is improved.
Embodiment III:
According to the digital product generating method provided by the above embodiment, based on the same technical concept, the embodiment of the present application further provides a digital product generating system, and fig. 5 is a schematic diagram of module composition of the digital product generating system provided by the embodiment of the present application, where the digital product generating system is used to execute the digital product generating method described in fig. 1, and as shown in fig. 5, the digital product generating system 500 includes: an obtaining module 501, configured to obtain first access data of a first user in a predetermined period of time and second access data of a second user in a predetermined period of time, where the access data includes access time, access times, and access duration for each user to access each digital product in the predetermined period of time; the determining module 502 is configured to determine, according to the access time, the access number and the access duration of the second user, a first access necessity of the second user for accessing each digital product, where the first access necessity indicates an interestingness of the second user for accessing the digital product, and the first access necessity is proportional to the interestingness; the determining module 502 is further configured to determine, from the second users, similar users having access habits similar to those of the first users according to the first access data and the second access data; the determining module 502 is further configured to determine a target digital product from digital products accessed by similar users according to the first access certainty, and recommend the target digital product to the first user.
According to the technical scheme disclosed by the embodiment of the application, the access necessity of the second user for accessing each digital product is quantified through the acquired access time, access times and access duration of the second user, the access necessity can indicate the interest level of the second user for accessing the digital product, and the greater the access necessity is, the greater the interest level of the user for accessing the digital product is. And then determining a similar user similar to the access habit of the first user according to the access data of the first user and the access data of the second user, determining a target digital product from the digital products accessed by the similar user based on the access necessity of the second user to the digital products, and recommending the target digital product to the first user.
In a possible implementation manner, the determining module 502 is further configured to determine, according to the access time, the number of times of access, and the access duration of the second user to each digital product, a first access time and a first access duration of each day of access to each digital product by the second user in a predetermined period of time, where the first access duration is an access duration of each day of access to each digital product by the second user; determining a second access necessity of each digital product accessed by a second user every day according to the first access times and the first access time length, wherein the second access necessity indicates the accidental nature of the digital product accessed by the user, and the second access necessity is in direct proportion to the accidental nature; determining the interest degree of the second user on each digital product by using the second access certainty; determining the number of days of access of the second user to each digital product according to the access time; the first access necessity is determined based on the number of days of access and the interestingness of the second user to each digital product.
In a possible implementation manner, the determining module 502 is further configured to determine, according to a time sequence of the access time, a first fluctuation curve of a first access duration of each digital product accessed by the second user every day in a predetermined period of time, and a first fitting curve corresponding to the first fluctuation curve; determining second access times, wherein the second access times are second access times which are larger than access time lengths of all digital products in first access time lengths of all the digital products accessed by a second user every day; determining a correction coefficient according to the slope value of the first fitting curve, the first access time length, the access time length during the first access and the second access times; correcting the second access necessity by using the correction coefficient to obtain the real access necessity; and determining the interest degree of the second user on each digital product by using the real access certainty.
In a possible implementation manner, the determining module 502 is further configured to calculate a first difference value between each first access duration and an access duration when the first access is performed, traverse all second access times, and stack the first difference values to obtain a first stack value; normalizing the first superposition value to obtain a normalized superposition value; and calculating the sum of the normalized superimposed value and the slope value to obtain a sum, and taking the sum as a correction coefficient.
In a possible implementation manner, the determining module 502 is further configured to superimpose the first access times to obtain a total access times of the second user accessing all the digital products every day; calculating the average value of the first access time length of all digital products in each day; calculating a second difference value between the first access time length and the average value, traversing all the first access times, and superposing the second difference value to obtain a second superposition value; calculating a first ratio of the second superimposed value to the first access times and a second ratio of the first access times to the total access times; and carrying out index operation on the first ratio value, and determining that the product of the index operation result and the second ratio value is the second access certainty of the digital product.
In a possible implementation manner, the determining module 502 is further configured to determine a second fluctuation curve of the real access necessity of the second user to each digital product on different days, and determine a second fitting curve corresponding to the second fluctuation curve; determining a slope value of the second fitted curve; under the condition that the slope value of the second fitting curve is not smaller than a preset value, determining the interestingness as the slope value; and under the condition that the slope value of the second fitting curve is smaller than a preset value, determining the interestingness as the preset value.
In a possible implementation manner, the determining module 502 is further configured to sort the access days of each digital product in ascending order, to obtain a sorted sequence; calculating a variance of a third difference between adjacent access days in the ordered sequence; calculating a third ratio of the number of days of access to the digital product to the variance; overlapping the interestingness with a preset value to obtain a third overlapping value; and multiplying the third ratio by the third superposition value to obtain a product, and normalizing the product to obtain the first access certainty.
In one possible implementation, the determining module 502 is further configured to calculate a predictive score of the accessed digital product of the similar user according to the first access certainty based on the collaborative filtering algorithm; and sorting the predictive scores in a descending order, determining that the digital products with the N bits before sorting and the digital products which are not accessed by the first user are target digital products, wherein N is an integer greater than 0.
In one possible implementation, the method further includes: the clustering module is used for carrying out coarse clustering on the second users based on a K clustering algorithm, and clustering the second users similar to the accessed digital products into a cluster; determining the first access certainty of the second user to access each digital product according to the access time and the access duration of the second user comprises: and taking each cluster as a user group, and determining the first access certainty of the second user in each cluster to access each digital product according to the access time and the access duration of the second user in each cluster.
It should be noted that, the digital product generating system provided by the embodiment of the present application and the digital product generating method provided by the embodiment of the present application are based on the same application concept, so that the implementation of the embodiment can refer to the implementation of the foregoing digital product generating method, and have the same or similar beneficial effects, and the repetition is omitted.
Embodiment four:
according to the digital product generating method provided by the above embodiment, based on the same technical concept, the embodiment of the present application further provides a digital product generating system, where the digital product generating system is used to execute the digital product generating method, and fig. 6 is a schematic structural diagram of a digital product generating system implementing the embodiments of the present application, and shown in fig. 6. The digital product generation system may vary widely in configuration or performance and may include one or more processors 601 and memory 602, the memory 602 being configured to store computer programs executable on the processors, the processors being configured to execute the programs stored on the memory to implement the steps of the method embodiments of fig. 1 or 2 above. Wherein the memory 602 may be transient storage or persistent storage. The application programs stored in memory 602 may include one or more modules (not shown) each of which may include a series of computer-executable instructions in the digital product generation system.
Still further, the processor 601 may be arranged to communicate with the memory 602 and execute a series of computer executable instructions in the memory 602 on the digital product generation system. The digital product generation system may also include one or more power supplies 603, one or more wired or wireless network interfaces 604, one or more input/output interfaces 605, and one or more keyboards 606.
In this embodiment, the digital product generating system includes a processor, a communication interface, a memory, and a communication bus; the processor, the communication interface and the memory complete communication with each other through a bus; a memory for storing a computer program; the processor is configured to execute the program stored in the memory, implement each step in the method embodiment in fig. 1 or fig. 2, and have the beneficial effects of the method embodiment, so that the embodiments of the present application are not repeated herein.
It should be noted that, the digital product generating system provided by the embodiment of the present application and the digital product generating method provided by the embodiment of the present application are based on the same application concept, so that the implementation of the embodiment can refer to the implementation of the foregoing digital product generating method, and have the same or similar beneficial effects, and the repetition is omitted.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.