A database interaction recommendation optimization method, device, equipment and storage medium

Abstract

The application discloses an optimization method and device for database interaction recommendation, equipment and a storage medium, and relates to the technical field of computers. The method comprises the following steps: acquiring a first operation of a user and a candidate recommendation operation set corresponding to the first operation; acquiring a second operation actually selected by the user; determining the precision of the user according to the second operation, determining the consistency of the user according to the second operation and the first operation, and determining the coverage of the user according to the candidate recommendation operation set; determining a feedback optimization coefficient of the user according to the precision, the coverage and the consistency; and optimizing the candidate recommendation operation set according to the feedback optimization coefficient and process information of the second operation selected by the user. The method can more accurately optimize the recommended content.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and in particular to a database interaction recommendation optimization method and device, equipment and a storage medium. BACKGROUND

[0002] In a database interaction scenario, when a user performs a first operation (for example, a basic query), subsequent operations often have multiple reasonable choices. For example, by adding a location condition to achieve accurate data screening, sorting by a time field to regularize the result set, directly exporting data into an analysis link, and the like, the system will select one candidate from multiple candidates to recommend to the user.

[0003] In a conventional scheme, when the system recommends candidate 1 to the user, the user selects candidate 2, thereby reducing the probability of recommending candidate 1 and increasing the probability of recommending candidate 2. This optimization scheme for recommended content has many disadvantages and cannot accurately optimize the recommended content. SUMMARY

[0004] The present application provides a database interaction recommendation optimization method, device, equipment and a storage medium, which can more accurately optimize the recommended content.

[0005] To achieve the above object, the present application adopts the following technical scheme:

[0006] In a first aspect, the present application provides a database interaction recommendation optimization method, which comprises:

[0007] Obtaining a first operation of a user and a candidate recommendation operation set corresponding to the first operation; obtaining a second operation actually selected by the user;

[0008] According to the second operation, determining the precision of the user, according to the second operation and the first operation, determining the continuity of the user, and according to the candidate recommendation operation set, determining the coverage of the user;

[0009] According to the precision, the coverage and the continuity, determining a feedback optimization coefficient of the user;

[0010] According to the feedback optimization coefficient and process information of the user selecting the second operation, optimizing the candidate recommendation operation set.

[0011] Optionally, after obtaining the second operation actually selected by the user, the method further comprises:

[0012] Obtaining a third operation selected by the user;

[0013] calculating a similarity between the third operation and all candidate operations within a preset time interval;

[0014] If there is a target candidate operation with a similarity greater than or equal to a similarity threshold among all candidate operations, a recommendation probability of the target candidate operation is optimized by increasing the target similarity between the target candidate operation and the third operation.

[0015] Optionally, the optimization of the candidate recommendation operation set according to the feedback optimization coefficient and the process information of the second operation selected by the user comprises:

[0016] updating the initial learning rate according to the feedback optimization coefficient to obtain a dynamic learning rate;

[0017] determining an initial heat index of the user for a candidate operation in the candidate recommendation operation set according to the process information of the second operation selected by the user, and updating the initial heat index according to the sensitivity coefficient to obtain a dynamic heat index;

[0018] obtaining an initial recommendation probability of a candidate operation in the candidate recommendation operation set;

[0019] updating the initial recommendation probability according to the dynamic learning rate and the dynamic heat index to obtain a dynamic recommendation probability corresponding to the candidate operation.

[0020] Optionally, the optimization of the recommendation probability of the target candidate operation by increasing the target similarity between the target candidate operation and the third operation comprises:

[0021] determining a target relevance between the target candidate operation and the third operation according to a historical recommendation probability of the target candidate operation, the target similarity between the target candidate operation and the third operation, and a target time interval between the third operation and the target candidate operation;

[0022] optimizing the recommendation probability of the target candidate operation by increasing the target relevance.

[0023] Optionally, the method further comprises:

[0024] obtaining a target role label of the user;

[0025] determining a target weight coefficient corresponding to the target role label according to a mapping relationship between a pre-configured reference role label and a reference weight coefficient, the target weight coefficient comprising a target precision rate coefficient, a target coverage rate coefficient, and a target coherence coefficient;

[0026] the determination of the feedback optimization coefficient of the user according to the precision rate, the coverage rate, and the coherence comprises:

[0027] According to the precision rate, the target precision rate coefficient, the coverage rate, the target coverage rate coefficient, the coherence, and the target coherence coefficient, a feedback optimization coefficient of the user is determined.

[0028] Optionally, the precision rate is determined by a hit of the second operation actually selected by the user in a candidate recommended operation set corresponding to the first operation; the coverage rate is obtained from reasonable candidate operations included in the candidate recommended operation set corresponding to the first operation and all reasonable candidate operations corresponding to the first operation; and the coherence is obtained from a correlation between the second operation and the first operation.

[0029] Optionally, the process information of the second operation selected by the user includes a residence time of a cursor in each candidate operation and a position of each candidate operation during the process in which the user selects the second operation.

[0030] In a second aspect, the present application provides an optimization device for database interaction recommendation, and the device comprises:

[0031] An acquisition module is configured to acquire a first operation of a user and a candidate recommended operation set corresponding to the first operation, and acquire a second operation actually selected by the user.

[0032] A determination module is configured to determine a precision rate of the user according to the second operation, determine a coherence of the user according to the second operation and the first operation, and determine a coverage rate of the user according to the candidate recommended operation set; and determine a feedback optimization coefficient of the user according to the precision rate, the coverage rate, and the coherence.

[0033] An optimization module is configured to optimize the candidate recommended operation set according to the feedback optimization coefficient and process information of the second operation selected by the user.

[0034] In a third aspect, the present application provides a computing device comprising a memory and a processor.

[0035] One or more computer programs are stored in the memory, and the one or more computer programs comprise instructions; when the instructions are executed by the processor, the computing device performs the method according to any one of the first aspect.

[0036] In a fourth aspect, the present application provides a computer readable storage medium for storing a computer program, and the computer program is used to perform the method according to any one of the first aspect.

[0037] From the above technical solution, the present application has at least the following beneficial effects:

[0038] The present application provides an optimization method for database interaction recommendation. In the optimization process, a first operation of a user and a candidate recommendation operation set corresponding to the first operation are obtained, and then a second operation actually selected by the user is obtained after the user makes an actual selection. Then, based on the second operation, the precision of the user is determined, based on the second operation and the first operation, the consistency of the user is determined, based on the candidate recommendation operation set, the coverage of the user is determined, based on the precision, the coverage and the consistency, the feedback optimization coefficient of the user is determined; and finally, based on the feedback optimization coefficient and the process information of the user selecting the second operation, the candidate recommendation operation set is optimized. In this method, not only whether the user selects the operation in the candidate recommendation set is considered for feedback optimization, but also the process information of the user actually selecting the second operation, the feedback optimization coefficient determined based on multiple angles, and the sorting of the candidate operations in the candidate recommendation operation set are considered, and the candidate recommendation operation set is optimized comprehensively. In this way, the method can more accurately optimize the recommended content.

[0039] It should be understood that the description of technical features, technical solutions, beneficial effects or similar language in the present application does not imply that all features and advantages can be achieved in any single embodiment. On the contrary, it can be understood that the description of a feature or a beneficial effect means that the specific technical feature, technical solution or beneficial effect is included in at least one embodiment. Therefore, the description of technical features, technical solutions or beneficial effects in the specification does not necessarily refer to the same embodiment. Further, the technical features, technical solutions and beneficial effects described in the embodiments can be combined in any appropriate manner. Those skilled in the art will understand that the embodiments can be implemented without one or more specific technical features, technical solutions or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects can be identified in specific embodiments that do not embody all embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 A flowchart of an optimization method for database interaction recommendation provided by the embodiments of the present application is provided;

[0041] Figure 2 A schematic diagram of an optimization device for database interaction recommendation provided by the embodiments of the present application is provided;

[0042] Figure 3 A schematic diagram of a computing device provided by the embodiments of the present application is provided. DETAILED DESCRIPTION

[0043] The terms "first", "second", and "third" and the like in the specification and the drawings of the present application are used to distinguish different objects, rather than to define a particular order.

[0044] In the embodiments of the present application, the words "exemplary" or "for example" are used to mean serving as an example, instance, or illustration, and not to imply any preference or superiority. In fact, the use of the words "exemplary" or "for example" is intended to present concepts in a particular manner.

[0045] In order to make the technical solutions of the present application clearer, the technical solutions of the present application will be introduced below in combination with application scenarios.

[0046] Application scenario: a user who is a data analyst uses a database interface to query stored order information. After the user performs an operation of selecting (SELECT * FROM orders) from the stored order information, the system predicts that the next possible operation of the user includes: 1, adding a location condition to filter orders of a specific location (for example, filtering orders in Asia, WHERE region = 'Asia'); 2, querying a time trend according to time sorting (for example, order_date); 3, counting sales of various products (for example, GROUP BY); and 4, exporting data for making a report.

[0047] These predictions all conform to business logic, but a traditional precision rate index can only determine whether the prediction "hits" the actual operation of the user (for example, the user finally performs operation 2), and ignores the rationality of operations 1, 3, and 4. If only the precision rate is used for evaluation, the model may be misjudged as predicting that the effects of operations 1, 2, and 4 are poor because operations 1, 3, and 4 are not hit, but actually other predictions are also valuable, only a single index cannot support the system to consider other values.

[0048] Therefore, the embodiments of the present application provide an optimization method for database interaction recommendation, which can be executed by a processing device. The processing device can be a terminal or a server. The terminal includes, but is not limited to, a smart phone, a tablet computer, a notebook computer, a personal digital assistant, or a smart wearable device, and the like. The server can be a cloud server, for example, a central server in a central cloud computing cluster or an edge server in an edge cloud computing cluster. Of course, the server can also be a server in a local data center. The local data center refers to a data center directly controlled by a user.

[0049] In the method, the multi-dimensional evaluation indexes of precision, coverage and continuity are obtained by adding the coverage and continuity evaluation indexes, the precision is used to directly measure the matching degree of the recommended content and the real demand of the user, the coverage is used to avoid that the recommended content is limited to the high-frequency operation of the user, and the continuity is used to ensure that the recommended content meets the requirements of the business scene. Based on the information, the feedback optimization coefficient of the user is determined. Then, the feedback optimization coefficient is combined with the process information of the second operation selected by the user and the first ranking of the candidate operations in the candidate recommended operation set to optimize the candidate recommended operation set. In this way, although the second operation of the user is not the candidate operation ranked at the rear in the candidate recommended operation set, if the process information indicates that the user pays attention to the candidate operation ranked at the rear, the recommendation probability of the candidate operation ranked at the rear can be improved. Therefore, the scheme is more tolerant to the operation that the user does not perform but pays attention to, and instead of directly reducing the recommendation probability, the recommendation probability is improved. It can be seen that the application can more accurately optimize the recommended content.

[0050] In order to make the technical solutions of the application clearer and easier to understand, the technical solutions of the application will be introduced below in combination with the drawings.

[0051] As shown in Figure 1 The figure is a flowchart of an optimization method of database interaction recommendation provided by an embodiment of the application, and the method comprises the following steps.

[0052] In S201, the processing device acquires the first operation of the user and the candidate recommended operation set corresponding to the first operation, and acquires the second operation actually selected by the user.

[0053] In some embodiments, the processing device can provide a human-computer interaction interface for the user, through which the user can control the database, for example, query the data stored in the database. The user can input "SELECT * FROM orders" in the input area of the human-computer interaction interface to perform order query. The operation of the user inputting "SELECT * FROM orders" in the input area is the first operation. After receiving the first operation input by the user, the processing device can determine the candidate recommended operation set corresponding to the first operation based on the first operation. The second operation refers to the operation actually selected by the user after the processing device displays the candidate recommended operation set corresponding to the first operation to the user.

[0054] In some examples, the set of candidate recommended operations corresponding to the first operation includes: A, adding a WHERE condition (filtering orders of a specific time); B, using GROUP BY to count the sales by commodity category; C, joining the products table to obtain commodity details; and D, exporting data to make a report. The second operation can be an operation in the set of candidate recommended operations or other operations.

[0055] S202, the processing device determines the precision of the user according to the second operation, determines the consistency of the user according to the second operation and the first operation, and determines the coverage of the user according to the set of candidate recommended operations.

[0056] The precision is used to represent the hit of the operation actually selected by the user in the set of candidate recommended operations; the consistency is used to represent the business relevance between the operation actually selected by the user and the previous operation; and the coverage is used to represent the rationality of the candidate operation in the set of candidate recommended operations.

[0057] The precision is determined by the hit of the second operation actually selected by the user in the set of candidate recommended operations corresponding to the first operation; the coverage is obtained from the reasonable candidate operations included in the set of candidate recommended operations corresponding to the first operation and all reasonable candidate operations corresponding to the first operation; and the consistency is obtained from the relevance of the second operation and the first operation.

[0058] In some examples, the processing device determines the precision of the user according to the second operation, which can be achieved by the following formula:

[0059]

[0060] In the formula, precision represents the precision of the user, n represents the number of candidate operations in the set of candidate recommended operations, and hit represents the hit of the second operation in the set of candidate recommended operations. When hit, hit represents 1; and when not hit, hit represents 0. In some examples, the processing device determines the consistency of the user according to the second operation and the first operation, which can be achieved by the following formula:

[0061] In the formula, consistency represents the consistency of the user, op1 represents the first operation, op2 represents the second operation, and sim represents the similarity between op1 and op2.

[0062]

[0063] In the formula, consistency represents the consistency of the user, op1 represents the first operation, op2 represents the second operation, and sim represents the similarity between op1 and op2. ​​​​​​​​​

[0064] In some examples, the processing device determines the coverage of the user according to the set of candidate recommended operations, which can be achieved by the following formula:

[0065]

[0066] wherein, represents the coverage, represents the number of reasonable candidate operations included in the set of candidate recommended operations, represents the number of all reasonable candidate operations corresponding to the first operation.

[0067] S203, the processing device determines the feedback optimization coefficient of the user according to the precision, the coverage and the consistency.

[0068] After the processing device obtains the precision, the coverage and the consistency, the processing device can determine the feedback optimization coefficient of the user based on the precision, the coverage and the consistency.

[0069] In some embodiments, different weighting methods can be set based on the role of the user, for example, for the user whose role label is data analyst, the user pays more attention to the comprehensiveness of data, so the weight of the coverage will be higher, which can prompt the user to obtain more comprehensive data for subsequent processing, for example, the weights of the precision, the coverage and the consistency are 0.2, 0.5 and 0.3 respectively; for the user whose role label is operation and maintenance engineer, the user pays more attention to the accuracy of data, so the weight of the precision will be higher, which can prompt the user to obtain more accurate data for subsequent processing, for example, the weights of the precision, the coverage and the consistency are 0.5, 0.2 and 0.3 respectively.

[0070] The processing device can also obtain the target role label of the user, and then determine the target weight coefficient corresponding to the target role label according to the pre-configured mapping relationship between the reference role label and the reference weight coefficient, wherein the target weight coefficient includes a target precision coefficient, a target coverage coefficient and a target consistency coefficient. Then, based on the precision, the target precision coefficient, the coverage, the target coverage coefficient, the consistency and the target consistency coefficient, the feedback optimization coefficient of the user is determined.

[0071] Specifically, the processing device can determine the feedback optimization coefficient by the following formula:

[0072]

[0073] wherein, S represents the feedback optimization coefficient, represents the target precision coefficient, represents the target coverage coefficient, represents the target consistency coefficient.

[0074] S204, the processing device optimizes the candidate recommended operation set according to the feedback optimization coefficient and the process information of the second operation selected by the user.

[0075] The process information of the second operation selected by the user includes the dwell time of the control cursor in each candidate operation during the process of the second operation selected by the user, and the position of each candidate operation. The position of each candidate operation can be represented by the row, for example, the first row, the tenth row, etc. The smaller the row number of the position, the more forward the candidate operation, that is, the more in the conspicuous position, and vice versa, that is, the more in the inconspicuous position.

[0076] The first ranking of the candidate operation in the candidate recommended operation set refers to the arrangement order of the candidate operation in the candidate recommended operation set. The higher the recommendation probability of the candidate operation, the more forward the ranking of the candidate operation in the candidate recommended operation set; the lower the recommendation probability of the candidate operation, the more backward the ranking of the candidate operation in the candidate recommended operation set.

[0077] In some embodiments, the processing device can update the initial learning rate according to the feedback optimization coefficient to obtain a dynamic learning rate. Specifically, the processing device can achieve the following formula:

[0078]

[0079] Wherein, The dynamic learning rate is represented by The initial learning rate is represented by S, and the feedback optimization coefficient is represented by S. As can be seen from the formula, in the case of a low feedback optimization coefficient, that is, in the case of poor recommendation quality, the learning rate is increased to speed up the update, and in the case of a large feedback optimization coefficient, that is, in the case of good recommendation quality, the learning rate is reduced to slow down the update. The embodiment of the present application can dynamically adjust the optimization process by means of the dynamic learning rate.

[0080] In some embodiments, the processing device can determine the initial heat index of the user for the candidate operation in the candidate operation set according to the process information of the second operation selected by the user, and then update the initial heat index based on the sensitivity coefficient to obtain a dynamic heat index. Specifically, the processing device can achieve the following formula:

[0081]

[0082] Wherein, The initial heat index of the kth candidate operation is represented by The hovering time of the user for the kth candidate operation in the candidate recommended operation set is represented by The position of the kth candidate operation in the candidate recommended operation set is represented by represents the average hovering time of the user for each candidate operation in the candidate recommended operation set, represents the average position of each candidate operation in the candidate recommended operation set.

[0083] After the processing device obtains the initial thermal index, the processing device can update the initial thermal index according to the sensitivity coefficient, and further obtain a dynamic thermal index. Specifically, the processing device can achieve this by the following formula:

[0084]

[0085] wherein, represents the dynamic thermal index, represents the initial thermal index, represents the sensitivity coefficient, represents the coefficient of the coverage of the user, for example, when the role label of the user is a data analyst, represents the coverage coefficient of the user, that is, .

[0086] In some embodiments, the processing device obtains an initial recommendation probability of the candidate operation in the candidate operation set.

[0087] Then, the processing device can update the initial recommendation probability based on the dynamic learning rate and the dynamic thermal index, and obtain a dynamic recommendation probability corresponding to the candidate operation. The processing device can achieve this by the following formula:

[0088]

[0089] wherein, represents the dynamic recommendation probability, represents the initial recommendation probability, represents the dynamic learning rate, represents the dynamic thermal index.

[0090] For example, it is assumed that , , , , , , Therefore, , , (the probability is increased by 2.3%).

[0091] In the technical solution, when there is a candidate operation that is of interest to the user but not selected in the candidate recommended operation set, the recommendation probability of the candidate operation can be improved in the subsequent recommendation process, so as to avoid the situation that the user does not select the candidate operation this time and the next time, and then the user needs to find the candidate operation in the next operation process, thereby affecting the business processing efficiency.

[0092] In the embodiments of the present application, the processing device dynamically optimizes the initial recommendation probability by using dynamic parameters (such as dynamic learning rate and dynamic heat index) in the process of optimizing the recommended content, thereby obtaining a dynamic recommendation probability. This optimization method can dynamically optimize the recommendation probability from multiple dimensions instead of static optimization, thereby further improving the accuracy of the recommended content and making the recommended content more matched to the user.

[0093] In some embodiments, the processing device can further obtain a third operation selected by the user after the second operation actually selected by the user, calculate the similarity between the third operation and all candidate operations within a preset time interval, and if there is a target candidate operation with a similarity greater than or equal to a similarity threshold value among all candidate operations, increase the recommendation probability of the target candidate operation according to the target similarity between the target candidate operation and the first operation.

[0094] For example, the third operation refers to the operation performed by the user after the second operation, the preset time interval can be 30 minutes, the time when the candidate recommended operation set corresponding to the first operation is recommended to the user is 9:00, the preset time interval is 30 minutes, and the time of the third operation is 10:00. The candidate operations within the preset time interval include the candidate operations recommended in the 9:30-10:00 time period. The similarity threshold value can be 70%, and the processing device needs to calculate the similarity between the third operation and the candidate operations recommended in the 9:30-10:00 time period, find the target candidate operation with a similarity greater than the similarity threshold value, and then increase the recommendation probability of the target candidate operation.

[0095] In the technical solution, the association between the operation with the delay and the historical recommended operation is identified, the recommendation failure caused by the time interval is avoided, and the subsequent recommendation probability of the target candidate operation is increased by 15%.

[0096] In some embodiments, the processing device determines the target relevance between the target candidate operation and the third operation according to the historical recommendation probability of the target candidate operation, the target similarity between the target candidate operation and the third operation, and the target time interval between the third operation and the target candidate operation, and then increases the recommendation probability of the target candidate operation based on the target relevance. Specifically, the processing device can achieve this by the following formula:

[0097]

[0098]

[0099] wherein, represents the target relevance of the target candidate operation to the third operation, represents the target similarity of the target candidate operation to the third operation, represents the historical recommendation probability of the target candidate operation, represents the target time interval, for example, the actual time interval, in hours; represents the increased recommendation probability of the target candidate operation, represents the dynamic learning rate.

[0100] In the above technical solution, through semantic understanding, the user's seemingly unrelated operations are connected together, so that the real demand is more accurately predicted. Through the time decay mechanism, the system is prevented from over-relying on outdated data (such as 30-minute-old recommendations), ensuring that the recommendation priority is always close to the user's current demand. This delay feedback correlation capability cannot be achieved by traditional linear recommendation models (such as only looking at the last operation).

[0101] Based on the above content description, the technical solution of the present application has at least the following beneficial effects:

[0102] Precise matching of business targets: The hierarchical evaluation framework dynamically adjusts the evaluation weight according to the user role (such as analyst or developer) and the task stage, so that the recommendation is tailored to the actual demand. In the analyst scenario, the high coverage rate weight prevents low accuracy operations from being misjudged, and in the developer scenario, the high accuracy weight prioritizes recommending syntax-accurate operations. The comprehensive score calculation is more scientific, solving the problems of multi-solution evaluation ambiguity and target mislocation. Intelligent identification of user interest: The heat calibration mechanism captures implicit feedback such as mouse hovering, eliminating the interference of the recommended position. For example, the bottom recommendation item "Sort by date", even if not clicked, the high hovering time increases the probability after heat index calibration, reducing the misjudgment rate by 28%, avoiding the system from misjudging the user's interest due to position factors. Cross-stage operation correlation optimization: Through semantic similarity and time decay calculation, the relevance of operations after a certain time and historical recommendations is associated, the success rate is increased from 12% to 53%, avoiding misjudgment of recommendation failure, and the operation subsequent recommendation probability is increased by 15%, optimizing the model's understanding of long-term operation logic.

[0103] In some embodiments, the method can directly measure the matching degree of the recommended result and the user's real demand by calculating the hit ratio (precision rate) of the actual selection operation (second operation) of the user in the candidate recommendation set. High precision rate means that the recommended result is more in line with the user's intention, reduces invalid information interference, and improves operation efficiency (such as the user finding the target function or content faster); analyzing the relevance (coherence) of the first operation and the second operation can identify the rationality of the user's operation path. For example, if the user first searches for "running shoes" and then clicks "purchase", high coherence indicates that the recommended process meets the user's expectations; if the user turns to a non-recommended operation (such as exiting the page), it indicates that there is a gap in the recommendation logic, and the operation guide sequence or associated recommended content needs to be adjusted; the coverage rate of the candidate recommendation set reflects its coverage ability for the user's potential operation (such as covering similar functions, associated content or alternative options). High coverage rate can avoid "information cocoon", ensuring that the user has rich choices in different scenarios, especially in scenarios that require exploratory operations (such as e-commerce recommendation, function navigation); combining precision rate, coverage rate and coherence can avoid the limitations of a single indicator: precision rate ensures "recommended accuracy", coverage rate ensures "recommended fullness", and coherence ensures "flow order", and the three-dimensional optimization of the three can achieve a balance between accuracy, diversity and logic of the recommended result.

[0104] The above Figure 1 The optimization method of the database interaction recommendation provided by the embodiments of the present application is described in detail, and the device and equipment provided by the embodiments of the present application will be introduced below with reference to the drawings.

[0105] As Figure 2 shown, the figure is a schematic diagram of an optimization device for database interaction recommendation provided by an embodiment of the present application, and the device comprises:

[0106] The acquisition module 201 is configured to acquire a first operation of a user and a candidate recommendation operation set corresponding to the first operation, and acquire a second operation actually selected by the user.

[0107] The determination module 202 is configured to determine a precision rate of the user according to the second operation, determine a coherence of the user according to the second operation and the first operation, and determine a coverage rate of the user according to the candidate recommendation operation set; and determine a feedback optimization coefficient of the user according to the precision rate, the coverage rate and the coherence.

[0108] The optimization module 203 is configured to optimize the candidate recommendation operation set according to the feedback optimization coefficient and process information of the second operation selected by the user.

[0109] Optionally, the acquisition module 201 is further configured to acquire a third operation selected by the user.

[0110] The optimization module 203 is specifically configured to calculate a similarity between the third operation and all candidate operations in a preset time interval, and increase the recommendation probability of a target candidate operation in all candidate operations if the similarity is greater than or equal to a similarity threshold.

[0111] Optionally, the optimization module 203 is specifically configured to update the initial learning rate according to the feedback optimization coefficient to obtain a dynamic learning rate, determine an initial heat index of the user for a candidate operation in the candidate recommendation operation set according to process information of the user selecting the second operation, update the initial heat index according to the sensitivity coefficient to obtain a dynamic heat index, obtain an initial recommendation probability of the candidate operation in the candidate recommendation operation set, and update the initial recommendation probability according to the dynamic learning rate and the dynamic heat index to obtain a dynamic recommendation probability corresponding to the candidate operation.

[0112] Optionally, the optimization module 203 is specifically configured to determine a target relevance between the target candidate operation and the third operation according to a historical recommendation probability of the target candidate operation, a target similarity between the target candidate operation and the third operation, and a target time interval between the third operation and the target candidate operation, and increase the recommendation probability of the target candidate operation according to the target relevance.

[0113] Optionally, the acquisition module 201 is further configured to acquire a target role label of the user, and the determination module 202 is configured to determine a target weight coefficient corresponding to the target role label according to a mapping relationship between a preconfigured reference role label and a reference weight coefficient, the target weight coefficient including a target precision rate coefficient, a target coverage rate coefficient and a target coherence coefficient, and determine a feedback optimization coefficient of the user according to the precision rate, the target precision rate coefficient, the coverage rate, the target coverage rate coefficient, the coherence and the target coherence coefficient.

[0114] Optionally, the precision rate is determined by a hit condition of the second operation actually selected by the user in a candidate recommendation operation set corresponding to the first operation, the coverage rate is obtained from reasonable candidate operations included in the candidate recommendation operation set corresponding to the first operation and all reasonable candidate operations corresponding to the first operation, and the coherence is obtained from a relevance between the second operation and the first operation.

[0115] Optionally, the process information of the user selecting the second operation includes a residence time of a cursor in each candidate operation and a position of each candidate operation during the user selecting the second operation.

[0116] The database interaction recommendation optimization apparatus according to the embodiments of the present application can correspond to performing the methods described in the embodiments of the present application, and the above-mentioned other operations and / or functions of each module / unit of the database interaction recommendation optimization apparatus are respectively implemented in order to achieve Figure 1 The corresponding flow of each method in the illustrated embodiments is not described here again for brevity.

[0117] The embodiments of the present application also provide a computing device. As shown in the figure, the figure is a schematic diagram of a computing device provided by the embodiments of the present application, and the computing device 400 includes a bus 401, a processor 402, a communication interface 403 and a memory 404. The processor 402, the memory 404 and the communication interface 403 communicate through the bus 401. Figure 3 The figure is a schematic diagram of a computing device provided by the embodiments of the present application, and the computing device 400 includes a bus 401, a processor 402, a communication interface 403 and a memory 404. The processor 402, the memory 404 and the communication interface 403 communicate through the bus 401.

[0118] The bus 401 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 In the figure, only one thick line is used to represent, but it does not mean that there is only one bus or only one type of bus.

[0119] The processor 402 can be any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP) processor, etc.

[0120] The communication interface 403 is used for external communication.

[0121] The memory 404 can include a volatile memory (volatile memory), such as a random access memory (RAM). The memory 404 can also include a non-volatile memory (non-volatile memory), such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD) or a solid state drive (SSD).

[0122] The executable code stored in the memory 404 is executed by the processor 402 to perform the aforementioned optimization method of database interaction recommendation.

[0123] Specifically, in the case of the embodiment shown, and Figure 2 Figure 2 In the case of the optimization apparatus of database interaction recommendation described in the embodiment being implemented by software, the software or program code required for the functions of each module / unit in the embodiment can be stored in the memory 404 in part or in whole. The processor 402 executes the program code corresponding to each unit stored in the memory 404 to perform the aforementioned optimization method of database interaction recommendation. Figure 2

[0124] The embodiment of the present application also provides a computer readable storage medium. The computer readable storage medium can be any available medium or data storage device that can be accessed by a computing device, such as a data center containing one or more available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk), etc. The computer readable storage medium includes instructions that instruct a computing device to perform the aforementioned optimization method of database interaction recommendation.

[0125] The embodiment of the present application also provides a computer program product including one or more computer instructions. When the computer instructions are loaded and executed on a computing device, the flow or function described in the embodiment of the present application is generated in whole or in part.

[0126] The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer or data center to another website, computer or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) mode.

[0127] The computer program product is executed by a computer, and the computer executes any of the aforementioned optimization methods of database interaction recommendation. The computer program product can be a software installation package, and in the case of needing to use any of the aforementioned optimization methods of database interaction recommendation, the computer program product can be downloaded and executed on the computer.

[0128] The description of the flow or structure corresponding to each of the above figures has its own emphasis, and the parts not described in detail in a certain flow or structure can be referred to the related description of other flows or structures.

[0129] ​​The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this, any change or replacement within the technical scope disclosed by the present application should be covered in the protection scope of the present application.

Claims

1. An optimization method for database interactive recommendation, characterized in that, The method comprises: obtaining a first operation of a user and a set of candidate recommended operations corresponding to the first operation; obtaining a second operation actually selected by the user; determining a precision of the user according to the second operation, determining a consistency of the user according to the second operation and the first operation, and determining a coverage of the user according to the set of candidate recommended operations; wherein the precision is used to represent a hit condition of the operation actually selected by the user in the set of candidate recommended operations; the consistency is used to represent a service correlation between the operation actually selected by the user and a previous operation; the coverage is used to represent a rationality of a candidate operation in the set of candidate recommended operations, the precision is determined by a hit condition of the second operation actually selected by the user in the set of candidate recommended operations corresponding to the first operation; the coverage is obtained from rational candidate operations included in the set of candidate recommended operations corresponding to the first operation and all rational candidate operations corresponding to the first operation; and the consistency is obtained from a correlation between the second operation and the first operation; determining a feedback optimization coefficient of the user according to the precision, the coverage and the consistency; updating an initial learning rate according to the feedback optimization coefficient to obtain a dynamic learning rate; the dynamic learning rate is realized by the following formula: wherein, denotes a dynamic learning rate, denotes an initial learning rate, S denotes a feedback optimization coefficient; determining an initial heat index of the user for a candidate operation in the set of candidate recommended operations according to process information of the user selecting the second operation; the initial heat index is realized by the following formula: wherein, represents an initial thermodynamic index of the kth candidate operation, represents a hovering time of the user for the kth candidate operation in the set of candidate recommended operations, represents a position of the kth candidate operation in the set of candidate recommended operations, represents an average hovering time of the user for each candidate operation in the set of candidate recommended operations, represents an average position of each candidate operation in the set of candidate recommended operations; updating the initial heat index according to a sensitivity coefficient to obtain a dynamic heat index; the dynamic heat index is realized by the following formula: wherein, represents a dynamic thermal index, represents an initial thermal index, represents a sensitivity coefficient, represents a coefficient of the coverage corresponding to the user; obtaining an initial recommendation probability of the candidate operation in the set of candidate recommended operations; updating the initial recommendation probability according to the dynamic learning rate and the dynamic heat index to obtain a dynamic recommendation probability corresponding to the candidate operation; the dynamic recommendation probability is realized by the following formula: wherein, represents a dynamic recommendation probability, represents an initial recommendation probability, represents a dynamic learning rate, represents a dynamic thermal index; after obtaining the second operation actually selected by the user, the method further comprises: obtaining a third operation selected by the user; calculating a similarity between the third operation and all candidate operations within a preset time interval; if there is a target candidate operation with a similarity greater than or equal to a similarity threshold in all candidate operations, increasing and optimizing a recommendation probability of the target candidate operation according to a target similarity between the target candidate operation and the third operation: wherein, represents a target relevance of the target candidate operation to the third operation, represents a target similarity of the target candidate operation to the third operation, represents a historical recommendation probability of the target candidate operation, represents a target time interval, represents an increased recommendation probability of the target candidate operation, represents a dynamic learning rate.

2. The method of claim 1, wherein, the increasing and optimizing of the recommendation probability of the target candidate operation according to the target similarity between the target candidate operation and the third operation comprises: determining a target correlation between the target candidate operation and the third operation according to a historical recommendation probability of the target candidate operation, the target similarity between the target candidate operation and the third operation, and a target time interval between the third operation and the target candidate operation; increasing and optimizing the recommendation probability of the target candidate operation according to the target correlation.

3. The method of claim 1, wherein, The method further comprises: obtaining a target role label of the user; According to a pre-configured mapping relationship between a reference role label and a reference weight coefficient, a target weight coefficient corresponding to the target role label is determined, and the target weight coefficient includes a target precision rate coefficient, a target coverage rate coefficient, and a target coherence coefficient; The determining of the feedback optimization coefficient of the user according to the precision rate, the coverage rate, and the coherence includes: The feedback optimization coefficient of the user is determined according to the precision rate, the target precision rate coefficient, the coverage rate, the target coverage rate coefficient, the coherence, and the target coherence coefficient.

4. The method according to any one of claims 1 to 3, characterized in that, The process information of the second operation selected by the user includes: the residence time of the control cursor in each candidate operation and the position of each candidate operation during the selection of the second operation by the user.

5. An apparatus for optimizing database interaction recommendation, characterized by, The device includes: The acquisition module is configured to acquire a first operation of a user and a candidate recommended operation set corresponding to the first operation, and acquire a second operation actually selected by the user; The determination module is configured to determine a precision rate of the user according to the second operation, determine a coherence of the user according to the second operation and the first operation, and determine a coverage rate of the user according to the candidate recommended operation set; and determine a feedback optimization coefficient of the user according to the precision rate, the coverage rate, and the coherence; wherein the precision rate is used to represent the hit condition of the operation actually selected by the user in the candidate recommended operation set; the coherence is used to represent the business correlation between the operation actually selected by the user and the previous operation; and the coverage rate is used to represent the rationality of the candidate operation in the candidate recommended operation set, the precision rate is determined by the hit condition of the second operation actually selected by the user in the candidate recommended operation set corresponding to the first operation; the coverage rate is obtained from the reasonable candidate operations included in the candidate recommended operation set corresponding to the first operation and all reasonable candidate operations corresponding to the first operation; and the coherence is obtained from the correlation between the second operation and the first operation; The optimization module is configured to: update an initial learning rate according to the feedback optimization coefficient to obtain a dynamic learning rate; the dynamic learning rate is realized by the following formula: wherein, denotes a dynamic learning rate, denotes an initial learning rate, S denotes a feedback optimization coefficient; determine an initial heat index of the user for the candidate operation in the candidate recommended operation set according to the process information of the second operation selected by the user; the initial heat index is realized by the following formula: wherein, represents an initial thermal index of the kth candidate operation, represents a hovering time of the user for the kth candidate operation in the set of candidate recommended operations, represents a position of the kth candidate operation in the set of candidate recommended operations, represents an average hovering time of the user for each candidate operation in the set of candidate recommended operations, represents an average position of each candidate operation in the set of candidate recommended operations; update the initial heat index according to a sensitivity coefficient to obtain a dynamic heat index; the dynamic heat index is realized by the following formula: wherein, represents a dynamic thermal index, represents an initial thermal index, represents a sensitivity coefficient, represents a coefficient of the coverage corresponding to the user; acquire an initial recommendation probability of the candidate operation in the candidate recommended operation set; update the initial recommendation probability according to the dynamic learning rate and the dynamic heat index to obtain a dynamic recommendation probability corresponding to the candidate operation; the dynamic recommendation probability is realized by the following formula: wherein, represents a dynamic recommendation probability, represents an initial recommendation probability, represents a dynamic learning rate, represents a dynamic thermal index; After acquiring the second operation actually selected by the user, the method further includes: acquire a third operation selected by the user; calculate the similarity between the third operation and all candidate operations within a preset time interval; If there are target candidate operations with similarity greater than or equal to the similarity threshold in all candidate operations, the recommendation probability of the target candidate operation is increased according to the target similarity between the target candidate operation and the third operation: wherein, represents a target relevance of the target candidate operation to the third operation, represents a target similarity of the target candidate operation to the third operation, represents a historical recommendation probability of the target candidate operation, represents a target time interval, represents an increased recommendation probability of the target candidate operation, represents a dynamic learning rate.

6. A computing device, comprising: comprise a memory and a processor; wherein the memory has stored therein one or more computer programs comprising instructions which, when executed by the processor, cause the computing device to perform the method of any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer readable storage medium is used to store a computer program for executing the method of any one of claims 1 to 4.

Images