CN109242109B - Management method of depth model and server - Google Patents
Management method of depth model and server Download PDFInfo
- Publication number
- CN109242109B CN109242109B CN201810739543.3A CN201810739543A CN109242109B CN 109242109 B CN109242109 B CN 109242109B CN 201810739543 A CN201810739543 A CN 201810739543A CN 109242109 B CN109242109 B CN 109242109B
- Authority
- CN
- China
- Prior art keywords
- depth model
- server
- standby
- result
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The embodiment of the invention relates to the technical field of computers, and discloses a depth model management method and a server. In the embodiment of the invention, the management method of the depth model is applied to a server, and the method comprises the following steps: processing the service request of the client by adopting the standby depth model, and verifying the processing result of the standby depth model on the service request to obtain a verification result; the standby depth model is a newly acquired depth model parallel to the currently used depth model; judging whether the standby depth model reaches the standard or not according to the checking result; if the judgment result is negative, abandoning the standby depth model; and if so, discarding the currently used depth model. The embodiment of the invention also provides a server. By adopting the embodiment of the invention, the server can detect and verify the newly acquired depth model, thereby providing a foundation for ensuring the stability of the service quality.
Description
Technical Field
The embodiment of the invention relates to the technical field of computers, in particular to a management method of a depth model and a server.
Background
Deep learning is a new field in machine learning research, and the motivation is to establish and simulate a neural network for analyzing and learning by human brain, and to interpret data by simulating the mechanism of human brain. With the rapid development of computers and the internet, deep learning plays an increasingly important role in the aspects of big data processing, artificial intelligence analysis and the like. At present, model training and external service providing are often bundled together inside a server, namely, the model training is performed while the external service providing is performed, the server generates a new depth model through continuously training and adjusting parameters of the depth model, and when a new depth model exists, the new depth model directly replaces the currently used depth model to use the new depth model for service.
However, the inventors of the present patent application found that: the service accuracy of the depth model does not monotonically increase with increasing number of training sessions. In the prior art, the new depth model is directly used to replace the depth model currently used, so that the condition of service accuracy drop caused by model updating is easy to occur, and the stability of the service quality of the server is poor.
Disclosure of Invention
The embodiment of the invention aims to provide a management method of a depth model and a server, which can detect and verify the newly acquired depth model, avoid the condition of service accuracy drop caused by model updating and provide a foundation for ensuring the stability of the service quality of the server.
In order to solve the above technical problem, an embodiment of the present invention provides a method for managing a depth model, which is applied to a server, and the method includes:
processing the service request of the client by adopting the standby depth model, and verifying the processing result of the standby depth model on the service request to obtain a verification result; the standby depth model is a newly acquired depth model parallel to the currently used depth model;
judging whether the standby depth model reaches the standard or not according to the checking result; if the judgment result is negative, abandoning the standby depth model; and if so, discarding the currently used depth model.
An embodiment of the present invention further provides a server, including:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of managing a depth model described above.
Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which when executed by a processor implements the depth model management method described above.
Compared with the prior art, the method and the device for processing the service requests have the advantages that the server can take the newly acquired depth model as the standby depth model, and the currently used depth model and the standby depth model can process the service requests in parallel. The server can detect and verify whether the standby depth model reaches the standard, and if the standby depth model does not reach the standard, the server can still use the currently used depth model to perform external service; if the standby depth model reaches the standard, the server can abandon the currently used depth model, so that the standby depth model is used for external service. By the method, the condition that the service accuracy falls due to model updating is avoided, and a foundation is provided for ensuring the stability of the service quality of the server.
In addition, verifying the processing result of the standby depth model to the service request to obtain a verification result, specifically comprising: returning the processing result of the standby depth model to the service request to the client; receiving feedback information of a client aiming at a processing result; and acquiring a checking result according to the feedback information. Therefore, the check result is obtained based on the feedback information of the client side, the matching degree of the check result and the user requirement is high, the accuracy of judging whether the standby depth model reaches the standard subsequently can be improved, and a foundation is provided for improving the service quality of the server.
In addition, the management method of the depth model also processes the service request by using the currently used depth model and returns the processing result of the currently used depth model to the service request to the client. Therefore, the client can acquire as much related information as possible, the possibility that the content displayed by the client can meet the requirements of the user is effectively improved, and the stability of the service quality of the server can be ensured.
In addition, the management method of the depth model also judges whether the feedback information is negative feedback information; if the judgment result is yes, the step of processing the service request by using the currently used depth model is executed. Therefore, the server returns the processing result of the currently used depth model to the service request to the client only when the processing result of the standby depth model does not meet the user requirement, so that the stability of the service quality of the server can be ensured, and the trouble of pushing excessive invalid information to the user can be avoided.
In addition, whether the standby depth model reaches the standard or not is judged according to the checking result, and the method specifically comprises the following steps: if the verification result is the Nth verification result of the standby depth model, calculating the verification passing rate of the standby depth model; judging whether the check passing rate is greater than or equal to a preset passing rate or not; wherein N is a positive integer. Therefore, a specific implementation form for judging whether the standby depth model reaches the standard according to the checking result is provided, and the flexibility of the implementation mode of the invention is improved.
In addition, if the verification result is the Nth time of the standby depth model and the verification result is verification passing, whether continuous M times of verification passing exist before the verification result is judged; wherein M is a positive integer. Therefore, a specific implementation form for judging whether the standby depth model reaches the standard according to the checking result is provided, and the flexibility of the implementation mode of the invention is improved.
In addition, before processing the service request of the client by adopting the standby depth model, the method further comprises the following steps: and receiving the depth model pushed by the training end server, and taking the received depth model as a standby depth model. Therefore, the model training end and the model server end are respectively arranged on the two servers, so that the waste of computing resources can be avoided, and the utilization rate of the computing resources is improved.
In addition, the training end server periodically pushes the depth model, and a foundation is provided for realizing rapid updating of the model and ensuring stable increase of service quality.
Drawings
One or more embodiments are illustrated by way of example in the accompanying drawings, which correspond to the figures in which like reference numerals refer to similar elements and which are not to scale unless otherwise specified.
Fig. 1 is a detailed flowchart of a management method of a depth model according to a first embodiment;
fig. 2 is a detailed flowchart of a management method of a depth model according to a fifth embodiment;
fig. 3 is a schematic structural diagram of a correspondence relationship between a model training terminal and a model service terminal according to a fifth embodiment;
fig. 4 is a schematic diagram of a server according to a sixth embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that numerous technical details are set forth in order to provide a better understanding of the present application in various embodiments of the present invention. However, the technical solution claimed in the present application can be implemented without these technical details and various changes and modifications based on the following embodiments.
The first embodiment of the invention relates to a management method of a depth model, and the specific flow is shown in fig. 1. The method for managing a depth model in the present embodiment is applied to a server, and is specifically described below:
and 101, processing the service request of the client by adopting the standby depth model, and verifying the processing result of the standby depth model on the service request to obtain a verification result.
Specifically, the server has a currently used depth model, and when the server acquires a new depth model, the server does not use the newly acquired depth model to replace the currently used depth model, but uses the acquired new depth model as a standby depth model, so that the standby depth model and the currently used depth model process the service request of the client side in parallel.
In an embodiment, the server may pre-store the verification data, so that when the server receives the service request of the client, the server may process the service request of the client by using the backup depth model, so as to verify the processing result of the backup depth model according to the preset verification data, thereby obtaining the verification result.
In another embodiment, the server obtains the verification result based on the feedback information of the client, and the matching degree of the verification result and the user requirement is higher, so that the accuracy of subsequently judging whether the standby depth model reaches the standard can be improved, and a foundation is provided for improving the service quality of the server.
Specifically, the server returns the processing result of the standby depth model to the service request to the client, receives feedback information of the client for the processing result, and acquires the verification result according to the feedback information. For example, taking the server as an intelligent customer service server as an example, when the user inputs the tag corresponding to the question on the client, the client sends the currently acquired tag as a service request to the server. And then, the standby depth model in the server retrieves and acquires the problem solutions related to the labels according to the labels in the service request, and returns the acquired problem solutions as processing results to the client, so that the client can conveniently display the received processing results to the user for viewing. It should be noted that when the result of the presentation processing is yes, the client also displays the inquiry information (e.g., whether the above answer solves the wording of your question) and provides the selection buttons "yes" and "no". If the user selects the 'yes' selection button, the feedback information is yes, and if the user selects the 'no' selection button, the feedback information is no. In this way, if the server detects that the feedback information is yes, the server considers that the check result of the standby depth model is that the check is passed; and if the server detects that the feedback information is negative, the server considers that the check result of the standby depth model is that the check is not passed.
However, the above feedback information is yes or no, and is merely an exemplary illustration for easy understanding, and the feedback information may exist in other forms in actual operation. That is to say, when the server obtains the verification result of the backup depth model according to the feedback information, the server may obtain the verification result by determining whether the feedback information is negative feedback information. If the feedback information is negative feedback information, the server considers that the verification result is that the verification is not passed; if the feedback information is not negative feedback information, the server considers that the verification result is verification pass.
And step 102, judging whether the standby depth model reaches the standard or not according to the verification result. If yes, go to step 103, otherwise go to step 104.
Specifically, the server determines whether the verification result is the nth verification result for the standby depth model. If the verification result is the Nth verification result of the standby depth model, the server calculates the verification passing rate of the standby depth model and judges whether the verification passing rate is larger than or equal to the preset passing rate. If the check pass rate is greater than or equal to the preset pass rate, the output result of step 102 is yes, otherwise, the output result of step 102 is no. The numerical value and the preset passing rate of N can be preset by technicians and stored in the server.
Compared with the prior art, the method and the device for processing the service requests have the advantages that the server can use the newly acquired depth model as the standby depth model, and the currently used depth model and the standby depth model are enabled to process the service requests in parallel. The server can detect and verify whether the standby depth model reaches the standard, and if the standby depth model does not reach the standard, the server can still use the currently used depth model to perform external service; if the standby depth model reaches the standard, the server can abandon the currently used depth model, so that the standby depth model is used for external service. By the method, the condition that the service accuracy falls due to model updating is avoided, and a foundation is provided for ensuring the stability of the service quality of the server.
A second embodiment of the present invention relates to a depth model management method. The second embodiment is improved on the basis of the first embodiment, and the main improvement lies in that: in the second embodiment of the present invention, the server also returns the processing result of the currently used depth model to the client, which effectively ensures the stability of the server service quality.
Specifically, after receiving a service request from the client, the server may further process the service request using the currently used depth model, and return a processing result of the currently used depth model on the service request to the client. Therefore, the client can acquire as much related information as possible, the possibility that the content displayed by the client can meet the requirements of the user is effectively improved, and the stability of the service quality of the server can be ensured. For example, taking the server as an intelligent customer service server as an example, the client displays the processing result of the standby depth model and the processing result of the current depth model, so that the possibility that the question input by the user on the client is solved is higher.
In this embodiment, the server feeds back the processing result of the standby depth model (hereinafter referred to as a first processing result) and the processing result of the current depth model (hereinafter referred to as a second processing result) to the client together, so that the user can select a processing result with a high matching degree with the user's own needs by himself/herself when viewing the content displayed by the client. And the client can also take the selection result of the user as feedback information and upload the feedback information to the server. And if the user clicks the first processing result for checking, the client selects the first processing result as feedback information and uploads the feedback information to the server, and the server acquires a verification result of the standby depth model as verification pass according to the feedback information. And if the user clicks the second processing result for checking, the client uploads the selected second processing result as feedback information to the server, and the server obtains a check result of the standby depth model as a check failure according to the feedback information. The client can determine the selection result of the user according to the click operation of the user, the duration of each processing result and other information.
The third embodiment of the present invention relates to a depth model management method. The third embodiment is substantially the same as the second embodiment, and mainly differs in that: the server pushes the first processing result and the second processing result at different time. The following is specifically described:
specifically, when the server receives a service request of the client, the server processes the service request of the client by using the standby depth model, and after a processing result of the standby depth model on the service request is returned to the client, the server waits for receiving feedback information of the client. The server judges whether the feedback information is negative feedback information or not after receiving the feedback information of the client, and under the condition that the feedback information is judged to be negative feedback information, the server processes the service request by using the currently used depth model and returns the processing result of the currently used depth model to the service request to the client. Therefore, the server can timely return the processing result of the currently used depth model to the service request to the client under the condition that the processing result of the standby depth model does not meet the requirement of the user, the stability of the service quality of the server can be ensured, and the trouble caused by pushing of too much invalid information to the user can be avoided.
A fourth embodiment of the present invention relates to a depth model management method. The fourth embodiment is substantially the same as the first embodiment, and mainly differs therefrom in that: and judging whether the standby depth model reaches the standard according to the verification result in different modes. The following is specifically described:
in this embodiment, the method for the server to determine whether the standby depth model reaches the standard according to the verification result is as follows: and the server detects whether the verification result is the Nth verification result of the standby depth model. And if the verification result is the Nth verification result of the standby depth model, the server judges whether the verification result passes the verification. And if the verification result is that the verification passes, the server detects whether continuous M times of verification passes before the verification result. If the server detects that continuous M times of verification pass exist before the verification result is obtained, the server judges that the standby depth model reaches the standard; and if the server detects that continuous M times of verification pass do not exist before the verification result is detected, the server judges that the standby depth model does not reach the standard. The value of N, M may be preset by a technician and stored on the server.
A fifth embodiment of the present invention relates to a method for managing a depth model, and a specific flow is shown in fig. 2. The fifth embodiment is an improvement of any of the above embodiments, and the main improvement is that: in the fifth embodiment of the present invention, the model training end and the model service end are respectively disposed on two servers, which can avoid the waste of computing resources and improve the utilization rate of computing resources. The following is specifically described:
Specifically, a technician may set a periodic push depth model of the training-side server, thereby providing a foundation for realizing rapid update of the model and ensuring stable increase of service quality. For example, the period may be one week.
More specifically, if the model training end and the model server end are both arranged on one server, each server is an independent individual and performs model training and external service, and the model training often consumes a large amount of computing resources. However, the training performed by the servers with the same function is the same, and the model training between different servers is not shared, so that when the model training end and the model service end are both arranged in one server, the utilization rate of the computing resources of the server is low. In this embodiment, the server of the depth model management method is a server of the model server. Therefore, by adopting the mode that the model training end and the model server end are respectively arranged on the two servers, the condition that one model training end corresponds to a plurality of model server ends can be realized, as shown in fig. 3, so that repeated training of different servers on the model can be reduced, the utilization rate of computing resources is ensured, and the lightweight of the model server end is realized.
The steps of the above methods are divided for clarity, and the implementation may be combined into one step or split some steps, and the steps are divided into multiple steps, so long as the same logical relationship is included, which are all within the protection scope of the present patent; it is within the scope of the patent to add insignificant modifications to the algorithms or processes or to introduce insignificant design changes to the core design without changing the algorithms or processes.
A sixth embodiment of the present invention relates to a server, as shown in fig. 4, including: at least one processor 301; and a memory 302 communicatively coupled to the at least one processor 301. The memory 302 stores instructions executable by the at least one processor 301, and the instructions are executed by the at least one processor 301, so that the at least one processor 301 can execute the method for managing the depth model in the above method embodiments.
Where the memory 302 and the processor 301 are coupled in a bus, the bus may comprise any number of interconnected buses and bridges, the buses coupling one or more of the various circuits of the processor 301 and the memory 302. The bus may also connect various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. A bus interface provides an interface between the bus and the transceiver. The transceiver may be one element or a plurality of elements, such as a plurality of receivers and transmitters, providing a means for communicating with various other apparatus over a transmission medium. The data processed by the processor is transmitted over a wireless medium via an antenna, which further receives the data and transmits the data to the processor.
The processor 301 is responsible for managing the bus and general processing and may also provide various functions including timing, peripheral interfaces, voltage regulation, power management, and other control functions. And the memory 302 may be used to store data used by the processor in performing operations.
Compared with the prior art, the implementation mode of the invention avoids the condition of falling of service accuracy caused by model updating and provides a foundation for ensuring the stability of the service quality of the server.
A seventh embodiment of the present invention relates to a computer-readable storage medium storing a computer program. The computer program realizes the above-described method embodiments when executed by a processor.
Compared with the prior art, the implementation mode of the invention avoids the condition of falling of service accuracy caused by model updating and provides a foundation for ensuring the stability of the service quality of the server.
That is, as can be understood by those skilled in the art, all or part of the steps in the method for implementing the above embodiments may be implemented by a program instructing related hardware, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be understood by those of ordinary skill in the art that the foregoing embodiments are specific embodiments for practicing the invention, and that various changes in form and details may be made therein without departing from the spirit and scope of the invention in practice.
Claims (10)
1. A method for managing a depth model, which is applied to a server, comprises the following steps:
processing a service request of a client by adopting a standby depth model, and verifying a processing result of the standby depth model on the service request to obtain a verification result; the standby depth model is a newly acquired depth model which processes a service request of a client side in parallel with a currently used depth model;
judging whether the standby depth model reaches the standard or not according to the checking result; if the judgment result is negative, discarding the standby depth model; and if so, discarding the currently used depth model.
2. The method for managing the depth model according to claim 1, wherein the verifying the processing result of the service request by the backup depth model to obtain a verification result specifically includes:
returning the processing result of the standby depth model to the service request to the client;
receiving feedback information of the client aiming at the processing result;
and acquiring the verification result according to the feedback information.
3. The method for managing a depth model of claim 2, further comprising:
and processing the service request by utilizing the currently used depth model, and returning a processing result of the currently used depth model to the service request to the client.
4. The method for managing a depth model of claim 3, further comprising:
judging whether the feedback information is negative feedback information or not;
and if the judgment result is yes, executing the step of processing the service request by using the currently used depth model.
5. The method for managing the depth model of claim 1, wherein the determining whether the backup depth model meets the standard according to the checking result specifically includes:
if the verification result is the Nth verification result of the standby depth model, calculating the verification passing rate of the standby depth model;
judging whether the check passing rate is greater than or equal to a preset passing rate or not; wherein N is a positive integer.
6. The method for managing the depth model of claim 1, wherein the determining whether the backup depth model meets the standard according to the checking result specifically includes:
if the verification result is the Nth verification result of the standby depth model and the verification result is verification passing, judging whether continuous M times of verification passing exist before the verification result; wherein M is a positive integer.
7. The method for managing the depth model of claim 1, wherein before processing the service request of the client using the backup depth model, the method further comprises:
and receiving a depth model pushed by a training end server, and taking the received depth model as the standby depth model.
8. The method for managing the depth model of claim 7, wherein the training server periodically pushes the depth model.
9. A server, comprising:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method of managing a depth model as claimed in any one of claims 1 to 8.
10. A computer-readable storage medium, storing a computer program, wherein the computer program, when executed by a processor, implements the method of managing a depth model of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810739543.3A CN109242109B (en) | 2018-07-06 | 2018-07-06 | Management method of depth model and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810739543.3A CN109242109B (en) | 2018-07-06 | 2018-07-06 | Management method of depth model and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109242109A CN109242109A (en) | 2019-01-18 |
| CN109242109B true CN109242109B (en) | 2022-05-10 |
Family
ID=65071881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810739543.3A Active CN109242109B (en) | 2018-07-06 | 2018-07-06 | Management method of depth model and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109242109B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106610854A (en) * | 2015-10-26 | 2017-05-03 | 阿里巴巴集团控股有限公司 | Model update method and device |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
| CN107563280A (en) * | 2017-07-24 | 2018-01-09 | 南京道熵信息技术有限公司 | Face identification method and device based on multi-model |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001183A (en) * | 2007-01-10 | 2007-07-18 | 网之易信息技术（北京）有限公司 | Test method and system for network application software |
| CN104731607B (en) * | 2013-12-18 | 2018-10-30 | 华为技术有限公司 | Terminal Lifelong Learning processing method, device and system |
| CN106227792B (en) * | 2016-07-20 | 2019-10-15 | 百度在线网络技术（北京）有限公司 | Method and apparatus for pushed information |
| CN106448670B (en) * | 2016-10-21 | 2019-11-19 | 竹间智能科技（上海）有限公司 | Conversational system is automatically replied based on deep learning and intensified learning |
| CN107273436A (en) * | 2017-05-24 | 2017-10-20 | 北京京东尚科信息技术有限公司 | The training method and trainer of a kind of recommended models |
| CN107330522B (en) * | 2017-07-04 | 2021-06-08 | 北京百度网讯科技有限公司 | Method, device and system for updating deep learning model |
-
2018
- 2018-07-06 CN CN201810739543.3A patent/CN109242109B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106610854A (en) * | 2015-10-26 | 2017-05-03 | 阿里巴巴集团控股有限公司 | Model update method and device |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
| CN107563280A (en) * | 2017-07-24 | 2018-01-09 | 南京道熵信息技术有限公司 | Face identification method and device based on multi-model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109242109A (en) | 2019-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11687832B1 (en) | Training a model using parameter server shards | |
| EP3836037A1 (en) | Method and system for executing machine learning process | |
| CN108833458B (en) | Application recommendation method, device, medium and equipment | |
| US20190124202A1 (en) | Call center system having reduced communication latency | |
| CN109685536B (en) | Method and apparatus for outputting information | |
| US11475055B2 (en) | Artificial intelligence based method and apparatus for determining regional information | |
| CN109344044B (en) | Method and system for calculating rendering time of first screen of page | |
| CN111125518B (en) | Household appliance information recommendation system and method | |
| CN103713935A (en) | Method and device for managing Hadoop cluster resources in online manner | |
| CN109144846B (en) | Test method and device for testing server | |
| JP2019503522A (en) | Optimized digital component analysis system | |
| WO2019085356A1 (en) | Method for displaying net promoter score and terminal thereof | |
| CN109242109B (en) | Management method of depth model and server | |
| US10482403B2 (en) | Methods and systems for designing of tasks for crowdsourcing | |
| CN114201626B (en) | Multimedia recommendation method and device, electronic equipment and storage medium | |
| US10084853B2 (en) | Distributed processing systems | |
| CN116205697A (en) | Resource recommendation method and resource recommendation device for virtual public cloud | |
| CN111598390B (en) | Method, device, equipment and readable storage medium for evaluating high availability of server | |
| CN115086194A (en) | Data transmission method for cloud application, computing equipment and computer storage medium | |
| CN111382263B (en) | Data processing method, device and system | |
| CN109885504B (en) | Recommendation system test method, device, medium and electronic equipment | |
| CN113902132A (en) | Negative feedback behavior prediction model training method, message pushing method and equipment | |
| CN116185782B (en) | Service monitoring method and device for social software | |
| CN111782794A (en) | Question-answer response method and device | |
| WO2014117566A1 (en) | Ranking method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |