Content sorting method, apparatus and electronic equipment based on user browsing history
Patent Information
- Application Number
- CN202610050152.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-14
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2046-01-14
Smart Images

Figure CN122087186B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of network application technology, and in particular to a method, apparatus, and electronic device for sorting content based on a user's browsing history. Background Technology
[0002] In real-world scenarios, many websites and applications analyze historical browsing data of user groups to infer user interests and rank the content displayed on pages. The ranking methods used in these technologies typically determine the default display order of content tags by analyzing group data such as click-through rates of popular content tags and average browsing time. However, this method often overlooks individual user interests, resulting in low ranking accuracy. Furthermore, this method requires computation based on a large amount of historical browsing data from various users to infer user interests, leading to a significant overall computational burden. Summary of the Invention
[0003] In view of this, this application provides a method, apparatus, and electronic device for sorting content based on user browsing history, so as to improve sorting accuracy and reduce overall computational load.
[0004] The technical solution provided in this application is as follows: According to an embodiment of the first aspect of this application, a method for sorting content based on a user's browsing history is provided, the method comprising: After detecting any content tag among the sequentially arranged content tags on the user-triggered page, for each content tag, if the content tag is the currently triggered content tag, then the current weight value of the content tag is adjusted; wherein, the adjusted weight value is greater than or equal to the weight value before adjustment; if the content tag is the currently untriggered content tag, then the current weight value of the content tag is adjusted; wherein, the adjusted weight value is less than or equal to the weight value before adjustment. Based on the current weight value of each content tag, the order of each content tag on the page is adjusted so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
[0005] Optionally, adjusting the current weight value of the content tag if it is the currently triggered content tag includes: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content resource is adjusted to the preset base value; If the current weight value of the content tag is greater than or equal to the preset base value, then a first target weight value is determined based on the current weight value, and the current weight value is adjusted to the first target weight value; the first target weight value is greater than or equal to the current weight value before adjustment.
[0006] Optionally, determining the first target weight value based on the current weight value includes: The historical consecutive click count corresponding to the current weight value is determined in the constructed growth curve function; the growth curve function is used to indicate the relationship between the consecutive click count of the content tag and the weight value; wherein, for any weight value within the range of the first minimum weight value and the first maximum weight value, the larger the weight value, the larger the consecutive click count; when the consecutive click count is greater than or equal to a set number, the weight value corresponding to the consecutive click count is the first maximum weight value; Based on the historical consecutive click count, determine the current consecutive click count; In the growth curve function, the weight value corresponding to the current number of consecutive clicks is determined, and the weight value corresponding to the current number of consecutive clicks is determined as the first target weight value.
[0007] Optionally, adjusting the current weight value of the content tag if it is one that has not yet been triggered includes: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value.
[0008] Optionally, determining the second target weight value based on the current weight value includes: The historical consecutive non-click count corresponding to the current weight value is determined from the constructed descent curve function; the reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count and the weight value of the content tag; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value; Based on the historical consecutive non-click counts, determine the current consecutive non-click count; In the descending curve function, the weight value corresponding to the current reference consecutive non-click count is determined, and the weight value corresponding to the current reference consecutive non-click count is determined as the second target weight value.
[0009] Optionally, determining the historical reference number of consecutive non-clicks corresponding to the current weight value in the constructed descending curve function includes: If the current weight value is less than the preset base value, then the historical reference number of consecutive non-clicks corresponding to the current weight value in the descending curve function is determined according to the inverse function of the descending curve function; If the current weight value is greater than or equal to a preset base value, then the current weight value is mapped from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the decline curve function, to obtain the reference weight value corresponding to the current weight value on the decline curve function; according to the inverse function of the decline curve function, the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined; the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined as the historical reference consecutive non-click count corresponding to the current weight value.
[0010] Optionally, each content tag is also configured with a base weight value; adjusting the arrangement order of the content tags on the page based on the current weight value of each content tag includes: For each content tag, determine the comprehensive weight value of the content tag based on its base weight value and current weight value; The order of content tags on the page is adjusted based on their overall weight values.
[0011] According to an embodiment of a second aspect of this application, a content sorting device based on user browsing history is provided, the device comprising: An adjustment unit is used to, after detecting any content tag among the sequentially arranged content tags on the user-triggered page, adjust the current weight value of each content tag if it is the currently triggered content tag; wherein the adjusted weight value is greater than or equal to the original weight value; if the content tag is not currently triggered, adjust the current weight value of the content tag; wherein the adjusted weight value is less than or equal to the original weight value. The sorting unit is used to adjust the order of the content tags on the page based on the current weight value of each content tag, so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
[0012] Optionally, the adjustment unit is specifically used for: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content resource is adjusted to the preset base value; If the current weight value of the content tag is greater than or equal to the preset base value, then a first target weight value is determined based on the current weight value, and the current weight value is adjusted to the first target weight value; the first target weight value is greater than or equal to the current weight value before adjustment. And / or, the adjustment unit is specifically used for: The historical consecutive click count corresponding to the current weight value is determined in the constructed growth curve function; the growth curve function is used to indicate the relationship between the consecutive click count of the content tag and the weight value; wherein, for any weight value within the range of the first minimum weight value and the first maximum weight value, the larger the weight value, the larger the consecutive click count; when the consecutive click count is greater than or equal to a set number, the weight value corresponding to the consecutive click count is the first maximum weight value; Based on the historical consecutive click count, determine the current consecutive click count; In the growth curve function, the weight value corresponding to the current number of consecutive clicks is determined, and the weight value corresponding to the current number of consecutive clicks is determined as the first target weight value; And / or, the adjustment unit is specifically used for: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value. And / or, the adjustment unit is specifically used for: The historical consecutive non-click count corresponding to the current weight value is determined from the constructed descent curve function; the reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count and the weight value of the content tag; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value; Based on the historical consecutive non-click counts, determine the current consecutive non-click count; In the descending curve function, the weight value corresponding to the current reference number of consecutive unclicked times is determined, and the weight value corresponding to the current reference number of consecutive unclicked times is determined as the second target weight value; And / or, the adjustment unit is specifically used for: If the current weight value is less than the preset base value, then the historical reference number of consecutive non-clicks corresponding to the current weight value in the descending curve function is determined according to the inverse function of the descending curve function; If the current weight value is greater than or equal to a preset base value, then the current weight value is mapped from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the decline curve function, to obtain the reference weight value corresponding to the current weight value on the decline curve function; according to the inverse function of the decline curve function, the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined; the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined as the historical reference consecutive non-click count corresponding to the current weight value; And / or, each content tag is also configured with a basic weight value; the sorting unit is specifically used for: For each content tag, determine the comprehensive weight value of the content tag based on its base weight value and current weight value; The order of content tags on the page is adjusted based on their overall weight values.
[0013] According to an embodiment of a third aspect of this application, an electronic device is provided, comprising: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor; the processor is configured to execute the machine-executable instructions to implement the method described in the first aspect.
[0014] As can be seen from the above technical solution, after detecting any content tag among the sequentially arranged content tags on the user-triggered page, this application adjusts the current weight value of each content tag. After each user interaction, the weight of all content tags is recalculated and updated, and the tags are re-sorted to ensure that the sorting results follow the latest changes in the user's interests. This makes the sorting results more accurate for different users and achieves personalized content display for different users. At the same time, this application does not require the analysis of other users' historical data, thereby reducing the overall computational load and improving the system's operating efficiency. Attached Figure Description
[0015] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this application and, together with the specification, serve to explain the principles of this application.
[0016] Figure 1 This is a schematic diagram illustrating the content tag sorting in an online education platform provided in an embodiment of this application; Figure 2This is a schematic flowchart of a content sorting method based on user browsing history provided in an embodiment of this application; Figure 3 A schematic diagram of a growth curve function and a decline curve function provided in this application embodiment; Figure 4 A structural diagram of a content sorting device based on user browsing history provided in an embodiment of this application; Figure 5 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation
[0017] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, and to make the above-mentioned objectives, features and advantages of the embodiments of this application more apparent and understandable, the technical solutions in the embodiments of this application will be further described in detail below with reference to the accompanying drawings.
[0018] In real-world scenarios, such as on some online education platforms, in order to improve users' learning efficiency, the system usually needs to display selectable content tags on a page (such as "My Learning Homepage" or "Course Recommendation Center"). Different content tags can correspond to different courses in different subjects or different courses in the same subject, such as "Trigonometric Functions Explained", "Ancient Poetry Appreciation", "Quick Word Memorization Methods", etc. This application does not impose any restrictions on this.
[0019] Please refer to Figure 1 , Figure 1 This is a schematic diagram illustrating the content tag sorting in an online education platform provided in this application embodiment.
[0020] like Figure 1 As shown, taking the "My Learning Homepage" page opened via a mobile terminal as an example, the page displays content tags corresponding to multiple courses such as Course 1 and Course 2. Users can view the corresponding course content by clicking on the content tag of the corresponding course.
[0021] To prioritize displaying content tags that users are most interested in, the content tags on the page need to be sorted. For example, the content tags that users are more interested in should be placed higher up or in positions that are more likely to attract users' attention.
[0022] The solutions proposed in related technologies usually rely on analyzing a large amount of users' historical course browsing data, such as the click-through rate of popular courses and the average learning time, to uniformly determine the default display order of these content tags. That is, for all users, the content tags on the page are sorted according to the above default display order.
[0023] However, this method often overlooks the individual interests of students. For example, some students prefer to choose course 1, while others prefer course 2. According to the proposed solutions in related technologies, it is difficult to accurately sort the content tags (i.e., different courses) on the student's learning homepage according to the different interests of students, resulting in a poor student experience.
[0024] Meanwhile, the proposed solutions in the related technologies rely on a large amount of historical course browsing data for analysis, resulting in a large overall computational load. Furthermore, for some new courses, it is impossible to accurately sort the new courses if sufficient historical course browsing data cannot be obtained.
[0025] Based on this, this application proposes a content sorting method based on user browsing history to improve sorting accuracy and reduce overall computational load.
[0026] It should be noted that the above scenario is only an illustrative example. For other pages that require personalized sorting of content tags for users, such as content recommendation systems and resource management systems, the solution proposed in this application can also be adopted, and this application does not impose any restrictions on it.
[0027] Please refer to Figure 2 , Figure 2 This is a schematic flowchart of a content sorting method based on user browsing history provided in an embodiment of this application.
[0028] In this embodiment, the method can be applied to the server or the client locally, and this application does not impose any limitations on it.
[0029] like Figure 2 As shown, the method may include the following steps: Step 201: After detecting any content tag among the sequentially arranged content tags on the user-triggered page, for each content tag, if the content tag is the currently triggered content tag, then adjust the current weight value of the content tag; wherein the adjusted weight value is greater than or equal to the weight value before adjustment; if the content tag is the currently untriggered content tag, then adjust the current weight value of the content tag; wherein the adjusted weight value is less than or equal to the weight value before adjustment.
[0030] In this embodiment, a user can trigger any content tag by clicking the content tag, touching the content tag with a gesture, or other methods to view the content corresponding to the content tag. This application does not impose any restrictions on this.
[0031] In this embodiment, after detecting any content tag among the content tags arranged in sequence on the user-triggered page, the current weight value of each recorded content tag can be obtained. The current weight value of each content tag refers to the adjusted weight value recorded after the last adjustment of the weight value of the content tag. The weight value can be recorded in memory or non-volatile storage device, and this application does not impose any restrictions on this.
[0032] The weight value of the content tag is used to represent the user's interest in the content tag. The higher the weight value, the more interested the user is in the content tag, and it will be ranked higher or more likely to attract the user's attention during the sorting process.
[0033] In this embodiment, when a user triggers a content tag, it indicates that the user is interested in the content corresponding to that tag. At this point, the current weight value of the content tag can be increased so that the content tag will rank higher in the next sorting. It should be noted that since the weight value cannot increase indefinitely, it usually has a maximum value (e.g., 1). Once the weight value of a content tag reaches this maximum value, even if the user continues to trigger the content tag, the weight value of the content tag can only remain at the maximum value and cannot continue to increase.
[0034] Therefore, for the currently triggered content tag, the current weight value of the content tag can be adjusted, and the adjusted weight value is greater than or equal to the weight value before adjustment.
[0035] In this embodiment, if a user does not trigger a content tag, it indicates that the user is not interested in the content corresponding to that tag. In this case, the current weight value of the content tag can be reduced so that the content tag will be ranked lower in the next sorting. It should be noted that since the weight value cannot be reduced indefinitely, it usually has a minimum value (such as 0). After the weight value of a content tag is reduced to this minimum value, even if the user continues not to trigger the content tag, the weight value of the content tag can only remain at the minimum value and cannot be reduced further.
[0036] Therefore, for content tags that are not currently being triggered, the current weight value of the content tag can be adjusted, and the adjusted weight value will be less than or equal to the original weight value.
[0037] The specific methods for adjusting the current weight value of the currently triggered content tag and the current weight value of the currently untriggered content tag will be described in detail below, and will not be repeated here.
[0038] This concludes the description of step 201. We will now proceed to step 202.
[0039] Step 202: Based on the current weight value of each content tag, adjust the arrangement order of each content tag on the page so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
[0040] In this embodiment, after obtaining the current weight value of each content tag through step 201, the arrangement order of each content tag on the page can be adjusted based on the current weight value of each content tag.
[0041] After adjusting the order of the content tags, the adjusted order can be recorded so that when the user browses the page again based on the page identifier, that is, when the user opens the page corresponding to the same page identifier again (e.g., reopens the page), the order can be recorded. Figure 1 When viewing the "My Learning Homepage" page, the adjusted content tags are displayed on the page according to their order, thus enabling personalized content tag sorting for different users.
[0042] The specific method for adjusting the order of content tags on the page based on their current weight values will be detailed below and will not be repeated here.
[0043] This concludes the description of step 202.
[0044] As can be seen, with the above-mentioned solution proposed in this application, even if user A and user B open the same page, such as the "Course Recommendation Center" page, the sorting of content tags in the "Course Recommendation Center" page for user A will be different from that for user B because the content tags triggered by user A and user B are different. This achieves personalized content tag sorting for different users and can significantly improve the accuracy of content tag sorting based on different user interests.
[0045] This concludes the discussion on... Figure 2 The description.
[0046] This application adjusts the current weight value of each content tag after detecting any content tag among the sequentially arranged content tags on the user-triggered page. After each user interaction, the weight of all content tags is recalculated and updated, and the tags are re-sorted to ensure that the sorting results follow the latest changes in the user's interests. This makes the sorting results more accurate for different users and enables personalized content display. At the same time, this application does not require the analysis of other users' historical data, thereby reducing the overall computational load and improving the system's operating efficiency.
[0047] The following describes the specific methods for adjusting the current weight value of the currently triggered content tag and the current weight value of the currently untriggered content tag.
[0048] In this embodiment, the method for adjusting the current weight value of the currently triggered content tag may include: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content resource is adjusted to the preset base value; If the current weight value of the content tag is greater than or equal to the preset base value, then a first target weight value is determined based on the current weight value, and the current weight value is adjusted to the first target weight value; the first target weight value is greater than or equal to the current weight value before adjustment.
[0049] Specifically, a base value can be preset, denoted as the preset base value. This preset base value can be a threshold benchmark, representing the minimum weight value required to transition from a state of no interest to a state of interest. It also represents the minimum weight value when a content tag is triggered. In other words, as long as a content tag is triggered, the adjusted current weight value is at least the preset base value.
[0050] When a content tag is triggered, if the current weight value of the content tag is found to be less than the preset base value, it means that the content tag may be a new content tag that has not been clicked, or it may be a content tag that the user was not interested in before. In this case, the current weight value of the content tag can be directly adjusted to the preset base value, which is equivalent to setting an initial level of interest for the content tag.
[0051] If the current weight value of the content tag is found to be greater than or equal to the preset base value, it means that the content tag has usually been triggered at least once, indicating that the content tag has been triggered multiple times by the user. At this time, the first target weight value can be determined based on the current weight value. The first target weight value is greater than or equal to the current weight value before adjustment, and the current weight value is adjusted to the first target weight value to further improve the current weight value of the content tag.
[0052] Specifically, the method for determining the first target weight value based on the current weight value may include: The historical consecutive click count corresponding to the current weight value is determined in the constructed growth curve function; the growth curve function is used to indicate the relationship between the consecutive click count of the content tag and the weight value; wherein, for any weight value within the range of the first minimum weight value and the first maximum weight value, the larger the weight value, the larger the consecutive click count; when the consecutive click count is greater than or equal to a set number, the weight value corresponding to the consecutive click count is the first maximum weight value; Based on the historical consecutive click count, determine the current consecutive click count; In the growth curve function, the weight value corresponding to the current number of consecutive clicks is determined, and the weight value corresponding to the current number of consecutive clicks is determined as the first target weight value.
[0053] In this embodiment, a growth curve function can be pre-constructed to indicate the relationship between the number of consecutive clicks on a content tag and its weight value. The growth curve function can use the number of consecutive clicks on the content tag as the horizontal axis and the weight value as the vertical axis, and the range of weight value variation can be pre-set. Here, the minimum weight value of the growth curve function is recorded as the first minimum weight value, and the maximum weight value of the growth curve function is recorded as the first maximum weight value.
[0054] For any weight value within the range of the first minimum weight value and the first maximum weight value, the larger the weight value, the greater the number of consecutive clicks; when the number of consecutive clicks is greater than or equal to a set number, the weight value corresponding to the number of consecutive clicks is the first maximum weight value.
[0055] Please refer to Figure 3 , Figure 3 This is a schematic diagram of a growth curve function and a decline curve function provided in an embodiment of this application.
[0056] like Figure 3 As shown in the figure, the blue curve is the growth curve function, which can be implemented using a logarithmic function.
[0057] For example, the growth curve function can be: Where fLog represents the y value corresponding to the given value x on the ln function within the specified numerical range; y represents the weight value of the content tag, x represents the number of consecutive clicks on the content tag (here, the unit can be selected as hours, days, etc., to represent consecutive clicks for x hours, consecutive clicks for x days, etc., according to actual needs); a represents the maximum value of the ln function value range, that is, the maximum value of y; c represents the zero-point shift of the ln function; d represents the minimum number of consecutive clicks required for the weight value to grow from the first minimum weight value of the growth curve function to the first maximum weight value (usually 1 in the ln function); base is the first minimum weight value, which in this embodiment can be the above-mentioned preset base value, and the specified numerical range is [base, 1]; minN is the minimum value of the ln function under the above parameters; maxN is the maximum value of the ln function under the above parameters.
[0058] vRange is a linear mapping function that maps a specified value from one range of values to the corresponding position in another range of values.
[0059] Specifically, the linear mapping function can be: Where, value is the specified value to be mapped, maxV is the maximum value of the current value range of value, minV is the minimum value of the current value range of value, maxR is the maximum value of the target value range, and minR is the minimum value of the target value range.
[0060] exist Figure 3 In the growth curve function shown, a=1, c=1, d=7, base=0.6.
[0061] After setting the growth curve function, it can be used to characterize the non-linear relationship between the weight value and the number of consecutive clicks. If the content tag is the currently triggered content tag, and the current weight value of the content tag is greater than or equal to the preset base value, the historical number of consecutive clicks corresponding to the current weight value can be determined using the aforementioned growth curve function.
[0062] Specifically, by using the inverse function of the growth curve function, the current weight value can be used as the independent variable to obtain the historical consecutive click count corresponding to the current weight value.
[0063] In this embodiment, the inverse function of the growth curve function can be the rfLog function, which is used to calculate the x value of the y value in the corresponding fLog given the y value.
[0064] Specifically, the inverse function of this growth curve function can be: Here, rate refers to the y-value of the current fLog function, i.e., the current weight value. Other parameters have been detailed above and will not be repeated here.
[0065] It should be noted that since the physical meaning of the determined x value is the number of consecutive clicks in history, the x value is an integer value. Here, the determined x value can be rounded up, down, or to the nearest integer, etc. This application does not impose any restrictions on this.
[0066] Preferably, the determined x value can be rounded up to make the final updated current weight value larger. The rounding error can be used as a reward for the user to continuously trigger the content tag.
[0067] After obtaining the historical consecutive click count corresponding to the current weight value, the current consecutive click count can be determined based on the historical consecutive click count. One way to determine the current consecutive click count is to increase the historical consecutive click count by a preset value, such as 1.
[0068] Furthermore, after obtaining the current number of consecutive clicks, the weight value corresponding to the current number of consecutive clicks can be determined in the growth curve function, and the weight value corresponding to the current number of consecutive clicks can be determined as the first target weight value.
[0069] This concludes the description of the method for determining the current weight value of the currently triggered content tag.
[0070] In this embodiment, the method for adjusting the current weight value of a content tag that has not yet been triggered may include: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value.
[0071] The second minimum weight value here can be a pre-set minimum threshold of all weight values, such as 0. The second minimum weight value here must be at least less than the first minimum weight value mentioned above.
[0072] If the current weight value of a content tag is found to be the second minimum weight value, it means that the weight value of the content tag has reached the minimum weight value and cannot be reduced further. At this time, the current weight value of the content tag can be kept at the second minimum weight value.
[0073] If the current weight value of a content tag is found to be greater than the second minimum weight value, it indicates that there is still room for the weight value of the content tag to decrease. Since the content tag has not been triggered, the second target weight value can be determined based on the current weight value of the content tag. The second target weight value is less than the current weight value before adjustment, and the current weight value is adjusted to the second target weight value.
[0074] Specifically, the method for determining the second target weight value based on the current weight value may include: The historical consecutive non-click count corresponding to the current weight value is determined from the constructed descent curve function; the reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count and the weight value of the content tag; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value; Based on the historical consecutive unclick counts, determine the current consecutive unclick count; In the descending curve function, the weight value corresponding to the current reference consecutive non-click count is determined, and the weight value corresponding to the current reference consecutive non-click count is determined as the second target weight value.
[0075] In this embodiment, a descent curve function can also be pre-constructed to indicate the relationship between the reference number of consecutive unclicked times for a content tag and its weight value. It should be noted that the reference number of consecutive unclicked times refers to how many more consecutive unclicked times the current weight value needs to drop to the second minimum weight value. The following will explain based on... Figure 3 The specific values in the data will be described in detail here, and will not be repeated here.
[0076] The descending curve function uses the reference number of consecutive unclicked content tags as the x-axis and the weight value as the y-axis, with a pre-defined range for the weight value. Here, the minimum weight value of the descending curve function is denoted as the second minimum weight value, and the maximum weight value is denoted as the second maximum weight value. The second maximum weight value can be the same as the first maximum weight value mentioned above.
[0077] For any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value.
[0078] Please continue to refer to this. Figure 3 ,like Figure 3 As shown in the figure, the yellow curve is the descending curve function, which can also be expressed in the same way as the growing function curve.
[0079] For example, a decreasing curve function can have the same expression as a growing function, but with different parameters, such as... Figure 3 The yellow descending curve function has a value of a=1, c=1, d=7, and base=0.
[0080] by Figure 3 Taking the downward curve function as an example, its horizontal axis represents the number of consecutive unclicked times. For example, if the horizontal axis is 2, it means that the weight value corresponding to the horizontal axis needs to be unclicked for two more consecutive times before it drops to 0. (Similarly, the unit can be selected as hours, days, etc., according to actual needs, indicating that the weight value corresponding to the horizontal axis needs to be unclicked for x consecutive hours or x consecutive days before it drops to 0.)
[0081] After setting the descent curve function, it can be used to characterize the non-linear relationship between the weight value and the reference consecutive non-click count. If the content tag is one that has not yet been triggered, and the current weight value of that content tag is determined to be greater than the second minimum weight value, the historical reference consecutive non-click count corresponding to the current weight value can be determined using the aforementioned descent curve function.
[0082] Specifically, methods for determining the number of consecutive unclicked historical references may include: If the current weight value is less than the preset base value, then the historical reference number of consecutive non-clicks corresponding to the current weight value in the descending curve function is determined according to the inverse function of the descending curve function; If the current weight value is greater than or equal to a preset base value, then the current weight value is mapped from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the decline curve function, to obtain the reference weight value corresponding to the current weight value on the decline curve function; according to the inverse function of the decline curve function, the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined; the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined as the historical reference consecutive non-click count corresponding to the current weight value.
[0083] In this embodiment, if the current weight value is found to be less than the preset base value, it indicates that the current weight value was previously determined based on the descending curve function. In this case, the historical reference number of consecutive non-clicks corresponding to the current weight value in the descending curve function can be determined directly based on the inverse function of the descending curve function.
[0084] It should be noted that the inverse function of the descending curve function has the same form as the inverse function of the growing curve function. It can be obtained by simply adjusting the relevant parameter base, which will not be elaborated here.
[0085] Since the physical meaning of the x value determined by the inverse function of the descending curve function is the number of consecutive unclicked times, the x value is also an integer value. Here, the determined x value can be rounded up, down, or to the nearest integer, etc. This application does not impose any restrictions on this.
[0086] Preferably, the determined x value can be rounded down to make the final updated current weight value smaller. The error of rounding down can be used as a penalty for the user not triggering the content tag for a period of time.
[0087] In this embodiment, if the current weight value is found to be greater than or equal to the above-mentioned preset base value, it indicates that the current weight value was usually determined based on the growth curve function, that is, the current content tag was triggered when it was triggered last time, but the current content tag was not triggered this time.
[0088] At this point, the obtained current weight value is the value on the growth curve function. Therefore, it is necessary to first use the vRange function to map the current weight value from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the decline curve function, so as to obtain the reference weight value.
[0089] The reference weight value here is a value on a descending curve function. We can then use the inverse function of the descending curve function to determine the reference consecutive non-click count corresponding to this reference weight value on the descending curve function, and use this reference consecutive non-click count as the historical reference consecutive non-click count for the current weight value. Similarly, we can round down the determined x value to obtain the reference consecutive non-click count.
[0090] After obtaining the historical consecutive non-click count corresponding to the current weight value, the current consecutive non-click count can be determined based on the historical consecutive non-click count.
[0091] It should be noted that the methods for determining the current reference consecutive non-click count based on the historical reference consecutive non-click count are different for the two different situations: the current weight value is less than the preset base value and the current weight value is greater than or equal to the preset base value.
[0092] If the current weight value is less than the preset base value, it means that the content tag has not been triggered for several consecutive times. In this case, the historical reference number of consecutive unclicked times can be reduced by the preset value, for example, by 1, to get the current reference number of consecutive unclicked times.
[0093] If the current weight value is greater than or equal to the preset base value, it indicates that this is the first time the content tag has not been triggered. In this case, the historical consecutive non-click count can be used as the current consecutive non-click count, and the difference between the conversion between the growth curve function and the decline curve function, as well as the error caused by rounding down, can be used as the penalty for the current weight value.
[0094] Furthermore, after obtaining the current reference consecutive non-click count, the weight value corresponding to the current reference consecutive click count can be determined in the growth curve function, and the weight value corresponding to the current reference consecutive click count can be determined as the second target weight value.
[0095] This concludes the description of the method for determining the current weight value of content tags that have not yet been triggered.
[0096] In this embodiment, a basic weight value can also be pre-configured for each content tag. This basic weight value can be obtained by manual setting or by statistical analysis of user data of the same profile. It serves as the basic weight value of the content tag itself, which cannot be adjusted or dynamically changed.
[0097] After determining the current weight value of each content tag after adjustment, the comprehensive weight value of each content tag can be determined based on its base weight value and current weight value; and the order of the content tags on the page can be adjusted based on the comprehensive weight values of each content tag.
[0098] Specifically, for each content tag, the adjusted current weight value of the content tag can be combined with the corresponding basic weight value through a specified operation, such as addition or multiplication, to obtain the comprehensive weight value of the content tag, and finally the content tags are sorted based on the comprehensive weight value.
[0099] For example, content tags can be sorted in descending order of their overall weight value; this application does not impose any restrictions on this.
[0100] As can be seen, the proposed solution differs from statistical or deep learning-based methods in related technologies. It uses function estimation to capture nonlinear relationships in data more flexibly, and is also applicable to personalized content ranking scenarios for single users. It requires less computation and can better adapt to sparse data and cold start situations.
[0101] The following is a general description of the content sorting device based on user browsing history proposed in this application through a specific embodiment.
[0102] In this embodiment, a basic ranking value α can be pre-set for each content tag, and the current weight value β of each content tag can be determined based on the user's trigger operation. Finally, α+β of each content tag is used as the final ranking value to sort the resources that the user is interested in.
[0103] The calculation method for β is as follows: Where, β i This represents the new β value after content tag i has been clicked once; baseR represents the preset base value; cur represents the current trigger value of content tag i; β i-1 This indicates the value of β before this trigger. Other parameters have been detailed above and will not be repeated here.
[0104] The following is still in the format of Figure 3 The growth and decline function curves shown briefly describe the five cases mentioned above: like Figure 3 As shown, the two curves are the blue growth function curve and the yellow decline function curve, both of which are functions of fLog. In the two curves, a=1, c=1, d=7, the base of the growth function curve is 0.6 (baseR), and the base of the decline function curve is 0.
[0105] Taking a learning scenario as an example, d can be defined as consecutive dates, that is, the number of days the user has continuously triggered the content tag. Taking content tag A as an example, its base ranking value is α=0.2. Assuming the user triggers content tag A for the first time, then the triggering situation is cur, and the β of content tag A is... i-1 If β is 0, which is less than baseR, then β i The value will be set to the baseR value, which is 0.6, corresponding to case ① above.
[0106] If the user triggers content tag A again the next day, then for the triggered condition cur, the β value will continue to rise on the growth curve function. Therefore, rfLog is used to obtain the current β. i-1 The corresponding x value is 1. Since the user continues to trigger content tag A, then the new β... i The value should be taken as the next function value on the growth curve, that is, the value at x+1, which is x=2, β i =min(1,max(0,ln(2)))=0.6934, corresponding to case ② above.
[0107] If the user does not trigger content tag A on the third day, then it is considered not triggered, i.e., "not cur". In this case, due to β... i-1 Still greater than baseR, the beta of content tag A iThe value will be transformed into a descending curve function to obtain its new value. First, the rfLog function is used to calculate the current β. i-1 The x-value corresponding to the descending curve is r=vRange(0.6934,0,1,0,1)≈0.6934, x=max(0,min(7,e0.6934-1))≈1 (rounded down). Further, the weight value on the descending curve corresponding to x=1 is determined, i.e., fLog(1,1,1,7,0), to obtain the current β. i The value is 0.3333, corresponding to case ③ above.
[0108] If the user does not trigger content tag A on the fourth day, then it is considered not triggered, i.e., "not cur". In this case, due to β... i-1 Since it is already less than baseR, β can be determined directly. i-1 The x-value corresponding to the descending curve function is determined, and the weight value corresponding to x-1 on the descending curve is determined as β. i This corresponds to situation ④ above.
[0109] For example, β determined on the fourth day i If the value is 0, and the user still hasn't triggered content tag A by the fifth day, then it falls under other cases (⑤), in which case β for the fifth day is directly determined. i It is also 0.
[0110] The corresponding current weight value β is determined each day. i Then, the sum of the current weight value and the basic weight value can be used to obtain the comprehensive weight value α+β for each content tag. Furthermore, all content tags can be sorted in descending order based on the comprehensive weight value so that when the user opens the interface again, the content tags will be displayed according to the sorted order.
[0111] This concludes the description of content sorting based on user browsing history.
[0112] Please refer to Figure 4 , Figure 4 This is a structural diagram of a content sorting device based on user browsing history proposed in an embodiment of this application. Figure 4 As shown, the device may include an adjustment unit 401 and a sorting unit 402. Specifically, the device includes: The adjustment unit 401 is configured to, after detecting any content tag among the sequentially arranged content tags in the user-triggered page, adjust the current weight value of each content tag if it is the currently triggered content tag, wherein the adjusted weight value is greater than or equal to the original weight value; and if it is a currently untriggered content tag, adjust the current weight value of that content tag, wherein the adjusted weight value is less than or equal to the original weight value. The sorting unit 402 is used to adjust the order of the content tags on the page based on the current weight value of each content tag, so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
[0113] Optionally, the adjustment unit 401 is specifically used for: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content resource is adjusted to the preset base value; If the current weight value of the content tag is greater than or equal to the preset base value, then a first target weight value is determined based on the current weight value, and the current weight value is adjusted to the first target weight value; the first target weight value is greater than or equal to the current weight value before adjustment. Optionally, the adjustment unit 401 is specifically used for: The historical consecutive click count corresponding to the current weight value is determined in the constructed growth curve function; the growth curve function is used to indicate the relationship between the consecutive click count of the content tag and the weight value; wherein, for any weight value within the range of the first minimum weight value and the first maximum weight value, the larger the weight value, the larger the consecutive click count; when the consecutive click count is greater than or equal to a set number, the weight value corresponding to the consecutive click count is the first maximum weight value; Based on the historical consecutive click count, determine the current consecutive click count; In the growth curve function, the weight value corresponding to the current number of consecutive clicks is determined, and the weight value corresponding to the current number of consecutive clicks is determined as the first target weight value; Optionally, the adjustment unit 401 is specifically used for: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value. Optionally, the adjustment unit 401 is specifically used for: The historical consecutive non-click count corresponding to the current weight value is determined from the constructed descent curve function; the reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count and the weight value of the content tag; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value; Based on the historical consecutive unclick counts, determine the current consecutive unclick count; In the descending curve function, the weight value corresponding to the current reference number of consecutive unclicked times is determined, and the weight value corresponding to the current reference number of consecutive unclicked times is determined as the second target weight value; Optionally, the adjustment unit 401 is specifically used for: If the current weight value is less than the preset base value, then the historical reference number of consecutive non-clicks corresponding to the current weight value in the descending curve function is determined according to the inverse function of the descending curve function; If the current weight value is greater than or equal to a preset base value, then the current weight value is mapped from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the decline curve function, to obtain the reference weight value corresponding to the current weight value on the decline curve function; according to the inverse function of the decline curve function, the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined; the reference consecutive non-click count corresponding to the reference weight value in the decline curve function is determined as the historical reference consecutive non-click count corresponding to the current weight value; Optionally, each content tag is also configured with a basic weight value; the sorting unit 402 is specifically used for: For each content tag, determine the comprehensive weight value of the content tag based on its base weight value and current weight value; The order of content tags on the page is adjusted based on their overall weight values.
[0114] This concludes the process. Figure 4 Description of a content sorting device based on user browsing history.
[0115] This application also provides embodiments that... Figure 4 Hardware structure description of the illustrated device. This hardware structure is... Figure 5The structure in the illustrated electronic device. Please refer to [link / reference]. Figure 5 , Figure 5 This is a structural diagram of an electronic device provided in an embodiment of this application. Figure 5 As shown, the hardware structure may include: a processor and a machine-readable storage medium, the machine-readable storage medium storing machine-executable instructions that can be executed by the processor; the processor is used to execute the machine-executable instructions to implement the method disclosed in the above example of this application.
[0116] Based on the same concept as the above method, this application also provides a machine-readable storage medium storing a plurality of computer instructions, which, when executed by a processor, can implement the method disclosed in the above examples of this application.
[0117] For example, the aforementioned machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, etc. For instance, machine-readable storage media can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or combinations thereof.
[0118] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A content sorting method based on user browsing history, characterized in that, The method includes: After detecting any content tag among the sequentially arranged content tags on the user-triggered page, for each content tag, if the content tag is the currently triggered content tag, then the current weight value of the content tag is adjusted; wherein, the adjusted weight value is greater than or equal to the weight value before adjustment; if the content tag is the currently untriggered content tag, then the current weight value of the content tag is adjusted; wherein, the adjusted weight value is less than or equal to the weight value before adjustment. Specifically, for the currently triggered content tag, if the current weight value of the content tag is greater than or equal to the preset base value, the corresponding historical consecutive click count is determined in the constructed growth curve function based on the current weight value of the content tag, the current consecutive click count is determined based on the historical consecutive click count, and the weight value corresponding to the current consecutive click count is determined in the growth curve function as the adjusted weight value. For content tags that are not currently triggered, if the current weight value of the content tag is less than a preset base value, then the historical reference number of consecutive unclicked occurrences corresponding to the current weight value in the descending curve function is determined according to the inverse function of the constructed descending curve function; if the current weight value of the content tag is greater than or equal to the preset base value, then the current weight value is mapped from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the descending curve function, to obtain the reference weight value corresponding to the current weight value on the descending curve function; the reference number of consecutive unclicked occurrences corresponding to the reference weight value in the descending curve function is determined according to the inverse function of the descending curve function; the reference number of consecutive unclicked occurrences corresponding to the reference weight value in the descending curve function is determined as the historical reference number of consecutive unclicked occurrences corresponding to the current weight value; the current reference number of consecutive unclicked occurrences is determined based on the historical reference number of consecutive unclicked occurrences, and the weight value corresponding to the current reference number of consecutive unclicked occurrences is determined in the descending curve function as the adjusted weight value; Based on the current weight value of each content tag, the order of each content tag on the page is adjusted so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
2. The method according to claim 1, characterized in that, For the currently triggered content tag, the method also includes: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content tag will be adjusted to the preset base value.
3. The method according to claim 1, characterized in that, The growth curve function is used to indicate the relationship between the number of consecutive clicks on a content tag and its weight value; wherein, for any weight value within the range of a first minimum weight value and a first maximum weight value, the larger the weight value, the larger the number of consecutive clicks; when the number of consecutive clicks is greater than or equal to a set number, the weight value corresponding to the number of consecutive clicks is the first maximum weight value.
4. The method according to claim 1, characterized in that, If the content tag is one that has not yet been triggered, then the current weight value of the content tag will be adjusted, including: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value.
5. The method according to claim 4, characterized in that, The reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count and the weight value of the content tag; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value.
6. The method according to claim 1, characterized in that, Each content tag is also configured with a base weight value; adjusting the order of the content tags on the page based on their current weight values includes: For each content tag, determine the comprehensive weight value of the content tag based on its base weight value and current weight value; The order of the content tags on the page is adjusted based on their overall weight values.
7. A content sorting device based on user browsing history, characterized in that, The device includes: An adjustment unit is configured to, upon detecting any content tag among the sequentially arranged content tags on a user-triggered page, adjust the current weight value of each content tag as follows: if the content tag is currently triggered, the adjusted weight value is greater than or equal to the original weight value; if the content tag is not currently triggered, the current weight value is adjusted as follows, the adjusted weight value is less than or equal to the original weight value. Specifically, for a currently triggered content tag, if its current weight value is greater than or equal to a preset base value, the unit determines the corresponding historical consecutive click count in a constructed growth curve function based on the current weight value, determines the current consecutive click count based on the historical consecutive click count, and determines the weight value corresponding to the current consecutive click count in the growth curve function as the adjusted weight value. For a currently untriggered content tag, if its current weight value is less than the preset base value, the unit adjusts the current weight value. Based on the inverse function of the constructed descending curve function, determine the historical reference number of consecutive unclicked occurrences corresponding to the current weight value in the descending curve function; if the current weight value of the content tag is greater than or equal to a preset base value, map the current weight value from the range of the first minimum weight value and the first maximum weight value corresponding to the constructed growth curve function to the range of the second minimum weight value and the second maximum weight value corresponding to the descending curve function, to obtain the reference weight value corresponding to the current weight value on the descending curve function; based on the inverse function of the descending curve function, determine the reference number of consecutive unclicked occurrences corresponding to the reference weight value in the descending curve function; determine the reference number of consecutive unclicked occurrences corresponding to the reference weight value in the descending curve function as the historical reference number of consecutive unclicked occurrences corresponding to the current weight value; determine the current reference number of consecutive unclicked occurrences based on the historical reference number of consecutive unclicked occurrences, and determine the weight value corresponding to the current reference number of consecutive unclicked occurrences in the descending curve function as the adjusted weight value; The sorting unit is used to adjust the order of the content tags on the page based on the current weight value of each content tag, so that when the user browses the page again based on the page identifier, the page will display the content tags in the adjusted order.
8. The apparatus according to claim 7, characterized in that, The adjustment unit is also used for: If the current weight value of the content tag is less than the preset base value, then the current weight value of the content tag is adjusted to the preset base value; And / or, the growth curve function is used to indicate the relationship between the number of consecutive clicks on a content tag and its weight value; wherein, for any weight value within the range of a first minimum weight value and a first maximum weight value, the larger the weight value, the larger the number of consecutive clicks; when the number of consecutive clicks is greater than or equal to a set number, the weight value corresponding to the number of consecutive clicks is the first maximum weight value; And / or, the adjustment unit is specifically used for: If the current weight value of the content tag is greater than the second minimum weight value, then the second target weight value is determined based on the current weight value, and the current weight value is adjusted to the second target weight value; the second target weight value is less than the current weight value before adjustment. If the current weight value of the content tag is the second minimum weight value, then the current weight value of the content tag will remain at the second minimum weight value. And / or, the reference consecutive non-click count refers to the number of consecutive non-clicks required for the current weight value to drop to the second minimum weight value; the descent curve function is used to indicate the relationship between the reference consecutive non-click count of the content tag and the weight value; wherein, for any weight value within the range of the second minimum weight value and the second maximum weight value, the smaller the weight value, the smaller the reference consecutive non-click count; when the reference consecutive non-click count is greater than or equal to a set number, the weight value corresponding to the reference consecutive non-click count is the second maximum weight value; And / or, each content tag is also configured with a basic weight value; the sorting unit is specifically used for: For each content tag, determine the comprehensive weight value of the content tag based on its base weight value and current weight value; The order of the content tags on the page is adjusted based on their overall weight values.
9. An electronic device, characterized in that, include: A processor and a machine-readable storage medium, the machine-readable storage medium storing machine-executable instructions that can be executed by the processor; The processor is configured to execute machine-executable instructions to implement the method according to any one of claims 1 to 6.
Citation Information
Patent Citations
Website directory display method and system
CN102542055A
Click feedback type individual recommendation system
CN102685565A
Search result display method and device
CN104123332A
Column display sequence adaptive adjustment method and device, equipment, medium and product
CN120832451A
