Search resource anomaly identification method and device, and electronic equipment

Abstract

The present disclosure provides a search resource anomaly identification method and device and electronic equipment, which relates to the technical field of artificial intelligence, specifically the technical field of intelligent search, computer vision and the like. The specific implementation scheme is: based on the historical record of the search result page presenting the detail information of the target search object, the historical presentation times of the target search object are generated; based on the historical presentation times, the predicted presentation times of the search result page presenting the detail information of the target search object within the target time are predicted; the real presentation times of the search result page presenting the detail information of the target search object within the target time are counted; based on the real presentation times and the predicted presentation times, the special search resource of the target search object is abnormally identified to obtain the first abnormal identification result of the special search resource, and the special search resource is the resource used for presenting the detail information of the target search object on the search result page. The present disclosure can improve the abnormal identification accuracy of the special search resource.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, specifically to the fields of intelligent search, computer vision, etc., and in particular to a method, apparatus, and electronic device for identifying anomalies in search resources. Background Technology

[0002] Web search is a technology frequently used in daily life. To improve or ensure search quality, it is often necessary to monitor search resources, specifically to identify whether search resources are abnormal. Currently, the identification of abnormal search resources mainly focuses on judging whether a search resource is abnormal based on a single search behavior. Summary of the Invention

[0003] This disclosure provides a method, apparatus, and electronic device for identifying anomalies in search resources.

[0004] According to one aspect of this disclosure, a method for identifying anomalies in search resources is provided, comprising:

[0005] Based on the historical records of displaying detailed information of the target search object on the search results page, the historical display count of the target search object is generated;

[0006] Based on the historical display count, predict the predicted number of times the detailed information will be displayed on the search results page within the target time period;

[0007] The actual number of times the detailed information was displayed on the search results page within the target time period is counted.

[0008] Based on the actual number of impressions and the predicted number of impressions, anomaly identification is performed on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources, wherein the special search resources are resources used to display the details information on the search results page.

[0009] According to one aspect of this disclosure, optimization instructions for generating the details information based on the display duration information are provided, including:

[0010] Based on the display duration information and cursor trajectory information, an optimized instruction for the details information is generated;

[0011] The cursor trajectory information is used to represent the cursor movement trajectory of the terminal displaying the search results page;

[0012] If the display duration information is lower than or equal to a preset duration threshold, the optimization indicator is used to indicate that the details information needs to be optimized;

[0013] If the display duration information is higher than a preset duration threshold, and the cursor trajectory information indicates that the cursor is moving back and forth on the details information, the optimization indicator is used to indicate that the details information does not need to be optimized.

[0014] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0015] At least one processor; and

[0016] A memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the search resource anomaly identification method provided in this disclosure.

[0018] According to another aspect of this disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the search resource anomaly identification method provided in this disclosure.

[0019] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the search resource anomaly identification method provided in this disclosure.

[0020] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0021] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0022] Figure 1 This is a flowchart of a method for identifying anomalies in search resources provided in this disclosure;

[0023] Figure 2 This is a schematic diagram of a search results page provided in this public disclosure;

[0024] Figure 3 This is a schematic diagram of a method for identifying anomalies in search resources provided in this disclosure;

[0025] Figure 4 This is a schematic diagram of a method for identifying anomalies in search resources provided in this disclosure;

[0026] Figures 5a to 5c This is a structural diagram of the search resource anomaly identification device disclosed herein;

[0027] Figure 6 This is a block diagram of an electronic device used to implement embodiments of the present disclosure. Detailed Implementation

[0028] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0029] Please see Figure 1 , Figure 1 This is a flowchart of a search resource anomaly identification method provided in this disclosure, such as... Figure 1 As shown, it includes the following steps:

[0030] Step S101: Based on the historical records of displaying the details of the target search object on the search results page, generate the historical display count of the target search object.

[0031] The search results page described above is a page that displays search results to users. This page may include multiple search results, where one search result is detailed information about the searched object, and the remaining search results are brief information about the searched object.

[0032] The target search object mentioned above can be a person, a piece of music, a company, a vehicle, a film or television program, a product, a tourist destination, etc. The detailed information mentioned above is used to introduce, describe, or display the detailed information of the target search object. For example, taking the song "XXX" as the target search object, the search results page could look like this: Figure 2 As shown, Figure 2 The 201 in the image refers to the detailed information for the song "XXX". This information includes the lyrics and playback controls, meaning users can play the song directly from the search results page without needing to navigate to other pages. Additionally, in... Figure 2 The search results page also includes brief information 202 related to the song "XXX" mentioned above. This brief information does not include the lyrics of the song, nor does it include the playback controls for the song. Instead, it includes brief information about the pages associated with the song.

[0033] In this disclosure, the detailed information of the target search object displayed on the search results page can be referred to as a search for a specific product or a special type of product, meaning that the information of the search object displayed on the search results page is different from other information displayed on the search results page. Furthermore, the aforementioned detailed information can be structured knowledge results.

[0034] The aforementioned historical records can be records of search results from one month ago, two months ago, six months ago, or even longer ago.

[0035] The historical display count of the target search object mentioned above can be the number of times the detailed information of the target search object is displayed on the search results page, based on the above historical records.

[0036] Step S102: Based on the historical display count, predict the predicted display count of the details information on the search results page within the target time period.

[0037] The predicted number of times the details information is displayed on the search results page within the target time period based on the historical display count can be obtained by calculating the average number of times the details information of the target search object is displayed within the target time period based on the historical display count, or by calculating the median number of times the details information of the target search object is displayed within the target time period based on the historical display count, to obtain the predicted number of times.

[0038] The target timeframe can be one week, two weeks, or one month.

[0039] Step S103: Count the actual number of times the detailed information is displayed on the search results page within the target time period.

[0040] The above statistics on the actual number of times the detailed information is displayed on the search results page within the target time period can be calculated based on the record information of the detailed information displayed on the search results page within the target time period.

[0041] Step S104: Based on the actual number of impressions and the predicted number of impressions, perform anomaly identification on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources, wherein the special search resources are resources used to display the details information on the search results page.

[0042] The above-mentioned anomaly identification of special search resources of the target search object based on the actual number of impressions and the predicted number of impressions can be achieved by comparing the difference between the actual number of impressions and the predicted number of impressions. If the difference is greater than a preset difference, the special search resource is determined to be abnormal; otherwise, the special search resource is determined to be normal.

[0043] In this disclosure, unusual special search resources may also be referred to as bad resources or bad products (bad cases).

[0044] The aforementioned special search resources may be pre-set search resources designed to display detailed information about the target search object on the search results page.

[0045] This disclosure identifies anomalies in special search resources of the target search object based on the actual and predicted number of times the detailed information of the target search object is displayed on the search results page. This allows for the identification of whether special search resources are abnormal from the perspective of the number of times the search results are displayed, avoiding the situation where a special search resource is determined to be abnormal based on a single search or a single search query, thereby improving the accuracy of anomaly identification for special search resources.

[0046] In this disclosure, the above method can be applied to an electronic device that performs all the steps included in the above method. The electronic device includes, but is not limited to, electronic devices such as servers, calculators, and mobile phones.

[0047] In one embodiment, the predicted number of impressions is the predicted number of impressions within a unit time period of the target time period, and the actual number of impressions includes the actual number of impressions within each unit time period of the target time period. The target time period includes multiple unit time periods. Based on the actual number of impressions and the predicted number of impressions, anomaly identification is performed on special search resources of the target search object to obtain anomaly identification results for the special search resources, including:

[0048] The actual number of impressions per unit time is compared with the predicted number of impressions to obtain the comparison results;

[0049] Based on the comparison results, anomaly identification is performed on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources.

[0050] The aforementioned unit of time can be a day, an hour, 12 hours, or other units of time.

[0051] The predicted number of impressions within a unit of time during the aforementioned target time period can be the average predicted number of impressions within a unit of time during the aforementioned target time period, for example: predicting the number of impressions per day within a week based on historical impression counts.

[0052] The comparison results above may include the difference between each actual number of impressions and the predicted number of impressions. For example, if the target time is one week, the comparison results above represent the difference between the actual number of impressions and the predicted number of impressions for each day of that week.

[0053] In this embodiment, since the anomaly identification of special search resources is based on the comparison results of comparing the actual number of times displayed in each unit of time with the predicted number of times displayed, the anomaly identification of special search resources can be comprehensively considered in multiple units of time, thereby improving the accuracy of anomaly identification of special search resources.

[0054] It should be noted that this disclosure does not limit the identification of whether a special search resource is abnormal to the actual number of times it is displayed in each unit of time. For example, in some embodiments, the total number of times it is displayed in the target time can also be used to identify whether a special search resource is abnormal.

[0055] In one embodiment, if the comparison result indicates that the number of target unit time within the target time is greater than or equal to a preset threshold, the first anomaly identification result indicates that the special search resource is abnormal.

[0056] If the comparison result indicates that the number of target units within the target time is less than a preset threshold, the first anomaly identification result indicates that the special search resource is normal.

[0057] The target unit of time is defined as the unit of time during which the actual number of impressions is less than the predicted number of impressions.

[0058] The phrase "the above-mentioned special search resources are normal" can be understood as meaning that the above-mentioned special search resources have not experienced any abnormalities.

[0059] The aforementioned preset threshold can be a frequency threshold or a proportion threshold. For example, if the difference between the actual number of displays and the preset predicted number of displays is greater than 10%, then the unit time is considered abnormal, which is the aforementioned target unit time.

[0060] In this embodiment, if the actual number of times displayed within multiple unit time periods is less than the preset predicted number of times within the target time period, the special search resource is determined to be abnormal; otherwise, the special search resource is determined to be normal.

[0061] In this embodiment, since the special search resource is determined to be abnormal when the number of target units within the target time is greater than or equal to a preset threshold, this can avoid the judgment of special search resource abnormality due to sudden situations, and further improve the accuracy of special search resource identification.

[0062] In one embodiment, the number of impressions can be the number of page views (PV), such as... Figure 3 As shown, it includes the following steps:

[0063] Step S301: Obtain the historical true daily-level display PV realThe PV of this day real It could be the daily page views (PV) since the launch of a special search resource (also known as a specialized search product). real ;

[0064] Step S302: Predict the daily PV for one week. pre ;

[0065] Step S303: Calculate the actual page views (PV) over a week. real The daily real-time performance (PVreal) will be compared with the daily real-time performance (PVpre) for the next week.

[0066] Step S304: Identify the abnormal special search resources, that is, identify the abnormal special search products.

[0067] If the deviation between the predicted value and the actual value on a certain day is greater than 10%, it is considered that the PV on that day is abnormal. If there are more than 3 abnormal days in a week (Dn>3), it is considered that this particular search resource has a problem in that week.

[0068] In this embodiment, anomaly identification can be performed on all special resource searches through a traversal approach. For example, the historical daily page views (PV) of the search objects corresponding to all special resource searches up to the current date can be obtained. real (dayi), i = 1, ..., n, where n represents the number of special resource searches and the actual page views (PV) displayed in the following week. real (dayj), j = 1, 2, ..., 7, are traversed in the above manner to obtain the set of special search resources set1 that have problems displaying PV in the next week.

[0069] In one embodiment, the method further includes:

[0070] Calculate the target feature metrics when the detailed information is displayed on the search results page;

[0071] Based on the target feature index and the mean value of the feature index, anomaly identification is performed on the special search resource to obtain the second anomaly identification result of the special search resource;

[0072] Wherein, the average feature index is the average feature index of the vertical category to which the target object belongs, the average feature index of the vertical category is the average feature index of multiple search objects under the vertical category, and the feature index of each search object is the feature index when the corresponding details information is displayed on the search results page.

[0073] The aforementioned target feature indicators can be indicator parameters used to represent the detailed information of the aforementioned target search object on the search results page.

[0074] The aforementioned vertical categories can be those that divide demand based on search terms, such as tourism, games, novels, film and television, and commodities.

[0075] The above-mentioned anomaly identification of the special search resource based on the target feature indicator and the mean of the feature indicator can be performed by comparing the target feature indicator with the mean of the feature indicator. When the difference between the target feature indicator and the mean of the feature indicator is greater than a preset value, the special search resource is determined to be abnormal; otherwise, the special search resource is determined to be normal. Alternatively, when the target feature indicator includes multiple feature indicators, the mean of each of these multiple feature indicators can be compared. When the number of abnormal feature indicators is greater than a preset value, the special search resource is determined to be abnormal; otherwise, the special search resource is determined to be normal. Here, an abnormal feature indicator refers to a feature indicator whose difference from its corresponding mean of the feature indicator is greater than a preset value.

[0076] In this embodiment, it is possible to identify whether a special search resource is abnormal based on the feature index dimension, and thus it is possible to identify whether a special search resource is abnormal through two dimensions, thereby further improving the accuracy of special search resource anomaly identification.

[0077] In one embodiment, the above-mentioned target feature indicators include at least one of the following:

[0078] Click-through rate, which is equal to: the number of times the details information displayed on the search results page is clicked divided by the total number of times it is displayed, where the total number of times the details information is displayed on the search results page;

[0079] The ratio of search content changes (query change ratio) is equal to the number of times the search content is displayed divided by the total number of times it is displayed. The number of times the search content is displayed is the number of times the detailed information is displayed on the search results page by changing the search content.

[0080] Bounce rate, which is equal to the number of bounces divided by the total number of times the page is displayed, wherein the following process constitutes a bounce: after entering the link page of the details information displayed on the search results page, the user returns to the link page of other brief information on the search results page;

[0081] Page turn rate, which is equal to the number of pages turned divided by the total number of times the page is displayed, wherein the following process constitutes one page turn: moving from the search results page displaying the details information to the next search results page, wherein the next search results page does not include the details information of the search results;

[0082] Long click rate, which is equal to the number of long clicks divided by the total number of times the page is displayed, wherein the following process constitutes a long click: clicking on the details information on the search results page, and then clicking on another piece of information displayed on the search results page after clicking on the details information, and the time interval between clicking on the details information and clicking on the other piece of information exceeds a preset time interval.

[0083] The other piece of information mentioned above can be any information on the search results page other than the details mentioned above.

[0084] The total number of times the detailed information is displayed on the search results page can be the total number of times the detailed information of the target search object is displayed on the search results page within a certain period of time. The number of clicks, the number of times the search content is changed, the bounce rate, the number of page turns, and the number of long clicks also refer to the number of clicks, the number of times the search content is changed, the bounce rate, the number of page turns, and the number of long clicks within that period of time.

[0085] In addition, the click-through rate, search content change ratio, bounce rate, page turn rate, and long click rate mentioned above can be click-through rate, search content change ratio, bounce rate, page turn rate, and long click rate under search PV. Search PV can refer to a user initiating a search and displaying the details of the search object on the search results page, which is considered a search PV.

[0086] The average value of the aforementioned feature indicators may include the average value of the feature indicators corresponding to the click-through rate, the proportion of search content changed, the bounce rate, the page turn rate, and the long click rate. When comparing, the click-through rate, the proportion of search content changed, the bounce rate, the page turn rate, and the long click rate are compared with the average value of their respective feature indicators. If the number of abnormal feature indicators is greater than the preset value, the aforementioned special search resource is determined to be abnormal; otherwise, the aforementioned special search resource is determined to be normal.

[0087] In this embodiment, the click-through rate, the proportion of search content changed, the bounce rate, the page-turning rate, and the long click rate can be used to more accurately identify whether special search resources are abnormal.

[0088] In one embodiment, the above calculation of the target feature metrics when displaying the details information on the search results page includes:

[0089] If the first anomaly identification result indicates that the special search resource is normal, calculate the target feature index when displaying the details information on the search results page.

[0090] In this embodiment, the identification methods corresponding to the first and second anomaly identification results can be complemented to further improve the accuracy of anomaly identification of special search resources and identify more abnormal special search resources.

[0091] It should be noted that in some embodiments, the above-mentioned target feature indicators are not limited to being calculated only when the first anomaly identification result indicates that the special search resource is normal. For example, in some embodiments, the anomaly of the special search resource can be identified based on both the first and second anomaly identification results. For example, if both the second and first anomaly identification results indicate that the special search resource is abnormal, the special search resource is ultimately determined to be abnormal.

[0092] In one embodiment, such as Figure 4 As shown, it includes the following steps:

[0093] Step S401: Obtain feature indicator data of search PV for multiple special search resources corresponding to search objects over a week, such as click-through rate, query change ratio, page turn rate, long click rate, bounce rate, etc.

[0094] Step S402: Calculate the daily average characteristic index (METRICS(srccid)) for each specific search resource. i ), and calculate the mean of feature metrics under vertical category requirements (METRICS_TYPE(type j ));

[0095] Step S403: Compare the average daily feature index and the average feature index under the vertical category for each special search resource;

[0096] Step S404: Identify the abnormal special search resources, that is, identify the abnormal special search products.

[0097] If two-thirds of the daily feature indicators of a special search resource are worse than the average feature indicators of the vertical category, then the special search resource is determined to be abnormal.

[0098] In this embodiment, anomaly identification can be performed on all special resource searches through a traversal approach. For example, the set of average characteristic indicators (METRICS(srcid)) of special search resources over a week can be obtained. i Given i = 1, ..., n, the output is: set2, a special set of unusual search resources.

[0099] In some embodiments, the abnormal special search resources corresponding to the first and second anomaly identification results can be obtained. For example, each time, a set SET of resource numbers for 20-50 abnormal special resources can be generated using the two methods described above (such as the union of set1 and set2). To reduce product development manpower, the number of resources can be further reduced, and the 10 special search resources with more serious anomalies can be selected, such as the 10 special search resources with the lowest characteristic indicators. After selecting the special search resources, an alarm can be issued, which can send the relevant information of the abnormal special search resources to the corresponding maintenance personnel, thus introducing a manual evaluation mechanism. For example, the resource data under 100 random search terms for each resource in the special search resource set SET can be manually evaluated to find out if there are information relevance issues, information accuracy issues, etc. In addition, the conclusions of the manual evaluation can be fed back to the product development personnel, who can then process the abnormal special search resources.

[0100] In one embodiment, the method further includes:

[0101] If the first anomaly identification result indicates that the special search resource is normal, obtain the display duration information of the search results page when the details information is displayed on the search results page;

[0102] Based on the display duration information, an optimization instruction for the details information is generated, which indicates whether the details information should be optimized.

[0103] The aforementioned display duration information may be provided by the terminal displaying the aforementioned search results page. For example, the terminal records the aforementioned display duration information when displaying the aforementioned search results page and sends the display duration information to the execution device of this method, such as sending it to the server.

[0104] The optimization instruction for generating the details information based on the display duration information can be as follows: if the display duration information indicates that the display duration of the details information of the target search object is less than a preset duration, an optimization instruction indicating that the details information of the target search object needs to be prioritized is generated; otherwise, an optimization instruction indicating that the details information of the target search object does not need to be prioritized is generated.

[0105] In this embodiment, the display duration information is used to determine whether the details of the target search object need to be optimized, which can improve the monitoring effect of the details of the target search object.

[0106] In practical applications, since the search results page displays detailed information about the target search object, if the detailed information is reasonable, users can directly browse it on the search results page to obtain the relevant information. If the user's browsing time is too short, it means that the detailed information does not contain the information the user needs, and thus optimization is required. Conversely, if the user browses the detailed information for a relatively long time, it can be determined that the detailed information includes the information the user needs.

[0107] In one embodiment, generating the optimization indication for the details information based on the display duration information includes:

[0108] Based on the display duration information and cursor trajectory information, an optimized instruction for the details information is generated;

[0109] The cursor trajectory information is used to represent the cursor movement trajectory of the terminal displaying the search results page;

[0110] If the display duration information is lower than or equal to a preset duration threshold, the optimization indicator is used to indicate that the details information needs to be optimized;

[0111] If the display duration information is higher than a preset duration threshold, and the cursor trajectory information indicates that the cursor is moving back and forth on the details information, the optimization indicator is used to indicate that the details information does not need to be optimized.

[0112] The cursor trajectory information mentioned above may be provided by the terminal displaying the search results page. For example, the terminal records the cursor trajectory information when displaying the search results page and sends the cursor trajectory information to the execution device of this method, such as sending it to the server.

[0113] The aforementioned movement of the cursor back and forth over the details of the target search object can be achieved by moving the cursor among multiple sub-information items included in the details. For example, if these multiple sub-information items are displayed side by side, then if the user controls the cursor to move over these sub-information items, it indicates that the user is carefully reading these sub-information items. Therefore, it is determined that the aforementioned details include the information that the user needs.

[0114] The optimization indicator is used to indicate that the details of the target search object need to be optimized. In addition to considering the display duration information mentioned above, cursor trajectory information can also be considered. One implementation method is as follows:

[0115] If the display duration is less than a preset duration threshold, and the cursor trajectory indicates a direct movement from the front to the back of the details information, the optimization indicator is used to suggest that the details information needs optimization. This is because if the user controls the cursor to move directly from the front to the back of the details information, it often indicates that the user is not interested in the details, thus confirming that the details need optimization.

[0116] In this embodiment, the display duration information and cursor trajectory information described above can further improve the monitoring effect of the details information and more accurately determine the details information that needs to be optimized.

[0117] This disclosure identifies anomalies in special search resources of the target search object based on the actual and predicted number of times the detailed information of the target search object is displayed on the search results page. This allows for the identification of whether special search resources are abnormal from the perspective of the number of times the search results are displayed, avoiding the situation where a special search resource is determined to be abnormal based on a single search or a single search query, thereby improving the accuracy of anomaly identification for special search resources.

[0118] Please see Figure 5a , Figure 5a This disclosure provides a search resource anomaly identification device, such as... Figure 5a As shown, the search resource anomaly identification device 500 includes:

[0119] The first generation module 501 is used to generate the historical display count of the target search object based on the historical record of displaying the detailed information of the target search object on the search results page.

[0120] The prediction module 502 is used to predict the predicted number of times the detailed information will be displayed on the search results page within a target time period based on the historical display count.

[0121] The statistics module 503 is used to count the actual number of times the details information is displayed on the search results page within the target time period;

[0122] The first identification module 504 is used to identify anomalies in special search resources of the target search object based on the actual number of times displayed and the predicted number of times displayed, and to obtain a first anomaly identification result of the special search resources, wherein the special search resources are resources used to display the details information on the search results page.

[0123] In one embodiment, the predicted number of impressions is the predicted number of impressions within a unit time period of the target time period, and the actual number of impressions includes the actual number of impressions within each unit time period of the target time period, wherein the target time period includes multiple unit time periods, and the first identification module is used for:

[0124] The actual number of impressions per unit time is compared with the predicted number of impressions to obtain the comparison results;

[0125] Based on the comparison results, anomaly identification is performed on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources.

[0126] In one embodiment, if the comparison result indicates that the number of target unit time within the target time is greater than or equal to a preset threshold, the first anomaly identification result indicates that the special search resource is abnormal.

[0127] If the comparison result indicates that the number of target units within the target time is less than a preset threshold, the first anomaly identification result indicates that the special search resource is normal.

[0128] The target unit of time is defined as the unit of time during which the actual number of impressions is less than the predicted number of impressions.

[0129] In one embodiment, such as Figure 5b As shown, the device further includes:

[0130] Calculation module 505 is used to calculate the target feature index when the details information is displayed on the search results page;

[0131] The second identification module 506 is used to identify anomalies in the special search resource based on the target feature index and the mean value of the feature index, and to obtain a second anomaly identification result of the special search resource.

[0132] Wherein, the average feature index is the average feature index of the vertical category to which the target object belongs, the average feature index of the vertical category is the average feature index of multiple search objects under the vertical category, and the feature index of each search object is the feature index when the corresponding details information is displayed on the search results page.

[0133] In one embodiment, the target feature index includes at least one of the following:

[0134] Click-through rate, which is equal to: the number of times the details information displayed on the search results page is clicked divided by the total number of times it is displayed, where the total number of times the details information is displayed on the search results page;

[0135] The ratio of search content to be changed is equal to the number of times the search content to be changed is divided by the total number of times the search content to be changed. The number of times the search content to be changed is the number of times the detailed information is displayed on the search results page by changing the search content.

[0136] Bounce rate, which is equal to the number of bounces divided by the total number of times the page is displayed, wherein the following process constitutes a bounce: after entering the link page of the details information displayed on the search results page, the user returns to the link page of other brief information on the search results page;

[0137] Page turn rate, which is equal to the number of pages turned divided by the total number of times the page is displayed, wherein the following process constitutes one page turn: moving from the search results page displaying the details information to the next search results page, wherein the next search results page does not include the details information of the search results;

[0138] Long click rate, which is equal to the number of long clicks divided by the total number of times the page is displayed, wherein the following process constitutes a long click: clicking on the details information on the search results page, and then clicking on another piece of information displayed on the search results page after clicking on the details information, and the time interval between clicking on the details information and clicking on the other piece of information exceeds a preset time interval.

[0139] In one embodiment, the computing module 505 is used for:

[0140] If the first anomaly identification result indicates that the special search resource is normal, calculate the target feature index when displaying the details information on the search results page.

[0141] In one embodiment, such as Figure 5c The device further includes:

[0142] The acquisition module 507 is used to acquire the display duration information of the search results page when the details information is displayed on the search results page, provided that the first anomaly identification result indicates that the special search resource is normal.

[0143] The second generation module 508 is used to generate an optimization instruction for the details information based on the display duration information, wherein the optimization instruction is used to indicate whether the details information should be optimized.

[0144] In one embodiment, the second generation module 508 is used for:

[0145] Based on the display duration information and cursor trajectory information, an optimized instruction for the details information is generated;

[0146] The cursor trajectory information is used to represent the cursor movement trajectory of the terminal displaying the search results page;

[0147] If the display duration information is lower than or equal to a preset duration threshold, the optimization indicator is used to indicate that the details information needs to be optimized;

[0148] If the display duration information is higher than a preset duration threshold, and the cursor trajectory information indicates that the cursor is moving back and forth on the details information, the optimization indicator is used to indicate that the details information does not need to be optimized.

[0149] The search resource anomaly identification device provided in this disclosure can implement all the processes implemented by the search resource anomaly identification method provided in this disclosure and achieve the same technical effect. To avoid duplication, it will not be described in detail here.

[0150] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0151] According to embodiments of this disclosure, this disclosure also provides an electronic device, an autonomous vehicle, a readable storage medium, and a computer program product.

[0152] The aforementioned electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the search resource anomaly identification method provided in this disclosure.

[0153] The aforementioned readable storage medium stores computer instructions, wherein the computer instructions are used to cause the computer to execute the search resource anomaly identification method provided in this disclosure.

[0154] The aforementioned computer program product includes a computer program that, when executed by a processor, implements the search resource anomaly identification method provided in this disclosure.

[0155] Figure 6 A schematic block diagram of an example electronic device 600 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0156] like Figure 6As shown, device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 602 or a computer program loaded from storage unit 608 into random access memory (RAM) 603. RAM 603 may also store various programs and data required for the operation of device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.

[0157] Multiple components in device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of monitors, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and communication unit 609, such as network card, modem, wireless transceiver, etc. Communication unit 609 allows device 600 to exchange information / data with other devices through computer grids such as the Internet and / or various telecommunications grids.

[0158] The computing unit 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as the search resource anomaly identification method. For example, in some embodiments, the search resource anomaly identification method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the search resource anomaly identification method described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform the search resource anomaly identification method by any other suitable means (e.g., by means of firmware).

[0159] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0160] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0161] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0162] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0163] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or grid browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication mesh). Examples of communication meshes include local area networks (LANs), wide area networks (WANs), and the Internet.

[0164] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact through a communication mesh. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0165] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0166] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for identifying anomalies in search resources, comprising: Based on the historical records of displaying detailed information of the target search object on the search results page, the historical display count of the target search object is generated; Based on the historical display count, predict the predicted number of times the detailed information will be displayed on the search results page within the target time period; The actual number of times the detailed information was displayed on the search results page within the target time period is counted. Based on the actual number of impressions and the predicted number of impressions, anomaly identification is performed on the special search resources of the target search object to obtain a first anomaly identification result of the special search resources, wherein the special search resources are resources used to display the details information on the search results page; Wherein, the predicted display count is the predicted display count within a unit time period of the target time period, the actual display count includes the actual display count for each unit time period of the target time period, the target time period includes multiple unit time periods, and based on the actual display count and the predicted display count, anomaly identification is performed on the special search resources of the target search object to obtain a first anomaly identification result for the special search resources, including: The actual number of impressions per unit time is compared with the predicted number of impressions to obtain the comparison results; Based on the comparison results, anomaly identification is performed on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources.

2. The method according to claim 1, wherein, If the comparison result indicates that the number of target unit time within the target time is greater than or equal to a preset threshold, the first anomaly identification result indicates that the special search resource is abnormal. If the comparison result indicates that the number of target units within the target time is less than a preset threshold, the first anomaly identification result indicates that the special search resource is normal. The target unit of time is defined as the unit of time during which the actual number of impressions is less than the predicted number of impressions.

3. The method according to any one of claims 1 to 2, further comprising: Calculate the target feature metrics when the detailed information is displayed on the search results page; Based on the target feature index and the mean value of the feature index, anomaly identification is performed on the special search resource to obtain the second anomaly identification result of the special search resource; Wherein, the average feature index is the average feature index of the vertical category to which the target search object belongs, the average feature index of the vertical category is the average feature index of multiple search objects under the vertical category, and the feature index of each search object is the feature index when the corresponding details information is displayed on the search results page.

4. The method according to claim 3, wherein, The target feature indicators include at least one of the following: Click-through rate, which is equal to: the number of times the details information displayed on the search results page is clicked divided by the total number of times it is displayed, where the total number of times the details information is displayed on the search results page; The ratio of search content to be changed is equal to the number of times the search content to be changed is divided by the total number of times the search content to be changed. The number of times the search content to be changed is the number of times the detailed information is displayed on the search results page by changing the search content. Bounce rate, which is equal to the number of bounces divided by the total number of impressions, wherein the following process constitutes a bounce: after entering the link page of the details information displayed on the search results page, the user returns to the link page of other brief information on the search results page; Page turn rate, which is equal to the number of pages turned divided by the total number of impressions, wherein the following process constitutes one page turn: moving from the search results page displaying the details to the next search results page, wherein the next search results page does not include the details of the search results; Long click rate, which is equal to the number of long clicks divided by the total number of impressions, wherein the following process constitutes a long click: clicking the details information on the search results page, and then clicking another piece of information displayed on the search results page after clicking the details information, and the time interval between clicking the details information and clicking the other piece of information exceeds a preset time interval.

5. The method according to claim 3, wherein, The calculation of the target feature metrics when displaying the details information on the search results page includes: If the first anomaly identification result indicates that the special search resource is normal, calculate the target feature index when displaying the details information on the search results page.

6. The method according to any one of claims 1 to 2, further comprising: If the first anomaly identification result indicates that the special search resource is normal, obtain the display duration information of the search results page when the details information is displayed on the search results page; Based on the display duration information, an optimization instruction for the details information is generated, which indicates whether the details information should be optimized.

7. The method according to claim 6, wherein, The step of generating optimization instructions for the details information based on the display duration information includes: Based on the display duration information and cursor trajectory information, an optimized instruction for the details information is generated; The cursor trajectory information is used to represent the cursor movement trajectory of the terminal displaying the search results page; If the display duration information is lower than or equal to a preset duration threshold, the optimization indicator is used to indicate that the details information needs to be optimized; If the display duration information is higher than a preset duration threshold, and the cursor trajectory information indicates that the cursor is moving back and forth on the details information, the optimization indicator is used to indicate that the details information does not need to be optimized.

8. A search resource anomaly identification device, comprising: The first generation module is used to generate the historical display count of the target search object based on the historical records of displaying detailed information of the target search object on the search results page. The prediction module is used to predict the number of times the detailed information will be displayed on the search results page within a target time period, based on the historical display count. The statistics module is used to count the actual number of times the detailed information is displayed on the search results page within the target time period; The first identification module is used to identify anomalies in special search resources of the target search object based on the actual number of times displayed and the predicted number of times displayed, and to obtain a first anomaly identification result of the special search resources, wherein the special search resources are resources used to display the details information on the search results page; Wherein, the predicted display count is the predicted display count within a unit time period of the target time period, the actual display count includes the actual display count for each unit time period of the target time period, and the target time period includes multiple unit time periods. The first identification module is used for: The actual number of impressions per unit time is compared with the predicted number of impressions to obtain the comparison results; Based on the comparison results, anomaly identification is performed on the special search resources of the target search object to obtain the first anomaly identification result of the special search resources.

9. The apparatus according to claim 8, wherein, If the comparison result indicates that the number of target unit time within the target time is greater than or equal to a preset threshold, the first anomaly identification result indicates that the special search resource is abnormal. If the comparison result indicates that the number of target units within the target time is less than a preset threshold, the first anomaly identification result indicates that the special search resource is normal. The target unit of time is defined as the unit of time during which the actual number of impressions is less than the predicted number of impressions.

10. The apparatus according to any one of claims 8 to 9, wherein the apparatus further comprises: The calculation module is used to calculate the target feature indicators when the detailed information is displayed on the search results page; The second identification module is used to identify anomalies in the special search resource based on the target feature index and the mean value of the feature index, and to obtain a second anomaly identification result for the special search resource. Wherein, the average feature index is the average feature index of the vertical category to which the target search object belongs, the average feature index of the vertical category is the average feature index of multiple search objects under the vertical category, and the feature index of each search object is the feature index when the corresponding details information is displayed on the search results page.

11. The apparatus according to claim 10, wherein, The target feature indicators include at least one of the following: Click-through rate, which is equal to: the number of times the details information displayed on the search results page is clicked divided by the total number of times it is displayed, where the total number of times the details information is displayed on the search results page; The ratio of search content to be changed is equal to the number of times the search content to be changed is divided by the total number of times the search content to be changed. The number of times the search content to be changed is the number of times the detailed information is displayed on the search results page by changing the search content. Bounce rate, which is equal to the number of bounces divided by the total number of impressions, wherein the following process constitutes a bounce: after entering the link page of the details information displayed on the search results page, the user returns to the link page of other brief information on the search results page; Page turn rate, which is equal to the number of pages turned divided by the total number of impressions, wherein the following process constitutes one page turn: moving from the search results page displaying the details to the next search results page, wherein the next search results page does not include the details of the search results; Long click rate, which is equal to the number of long clicks divided by the total number of impressions, wherein the following process constitutes a long click: clicking the details information on the search results page, and then clicking another piece of information displayed on the search results page after clicking the details information, and the time interval between clicking the details information and clicking the other piece of information exceeds a preset time interval.

12. The apparatus according to claim 10, wherein, The calculation module is used for: If the first anomaly identification result indicates that the special search resource is normal, calculate the target feature index when displaying the details information on the search results page.

13. The apparatus according to any one of claims 8 to 9, wherein the apparatus further comprises: The acquisition module is used to acquire the display duration information of the search results page when the details information is displayed on the search results page, provided that the first anomaly identification result indicates that the special search resource is normal. The second generation module is used to generate an optimization instruction for the details information based on the display duration information. The optimization instruction is used to indicate whether the details information should be optimized.

14. The apparatus according to claim 13, wherein, The second generation module is used for: Based on the display duration information and cursor trajectory information, an optimized instruction for the details information is generated; The cursor trajectory information is used to represent the cursor movement trajectory of the terminal displaying the search results page; If the display duration information is lower than or equal to a preset duration threshold, the optimization indicator is used to indicate that the details information needs to be optimized; If the display duration information is higher than a preset duration threshold, and the cursor trajectory information indicates that the cursor is moving back and forth on the details information, the optimization indicator is used to indicate that the details information does not need to be optimized.

15. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

17. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-7.

Images