TWI579719B - Solution searching system and method for operating a solution searching system - Google Patents

Description

Solution search system and solution search system operation method

The present invention relates to a solution search system, and more particularly to a solution search system that utilizes massive data and data exploration techniques.

In addition to being closely related to R&D technology, the success of a product requires a large number of tests to ensure product stability, especially for high-stability, high-reliability technology products such as industrial instruments, mobile devices, workstations, and individuals. For products such as computers or servers, the standards for quality control testing are more stringent. When a product is detected, it must be done by copying the problem, collecting and analyzing the relevant information, identifying the cause of the problem, suggesting possible solutions, and verifying the proposed solution... to ensure that the detected problem is Properly resolved, these processes may not only be very time consuming, but may even lead to missed opportunities for the product to enter the market, and implementation must rely on the personal experience and professionalism of the engineer; that is, whether the engineer's experience and professionalism are sufficient will be greatly It affects the time required to propose a solution, and it may also affect whether the proposed solution can completely solve the problem, resulting in the quality of the solution is difficult to grasp. In addition, because personal experience is not easy to pass on, even if the problem to be solved is the same or similar, different engineers may have to repeat the above process to get a solution. This practice is not only inefficient, but also ensures that the engineer can find the most appropriate. s solution.

In addition, for the same type of products, the proportion of the same or similar problems is very high. In the past, although the solution was recorded or archived, due to the wide variety of problems, the amount of information involved is quite large. The way engineers describe problems may be inconsistent, so it is difficult to systematically store them. As a result, engineers are still not easy to find relevant solutions, and it is difficult to achieve the purpose of sharing experiences with engineers.

An embodiment of the present invention provides a solution search system including a computing server, a plurality of model servers, a huge database, a database server, a hub server, and an associated database.

The computing server receives and generates a problem description file corresponding to the model input file of the prediction model according to the number of the prediction model corresponding to the problem description file and the problem description file. Each model server corresponds to a prediction model, and each model server is configured to generate a solution code by using a prediction model corresponding to the model server and a model input file when receiving the model input file. The database server reads at least one solution from the huge database based on the solution code generated by the model server. The associated database stores the identity of the user, the predictive models available to the user, and the small and/or temporary data required for the hub server to operate.

The hub server provides a plurality of usable prediction models to the user according to the user's identity, and the data exploration algorithms corresponding to the plurality of usable prediction models are different. When the user selects the first prediction model from the predictive models that can be used and transmits the first problem description file, the hub server transmits the numbers of the first problem description file and the first prediction model to the computing server. The hub server transmits the first model input file corresponding to the first prediction model generated by the first problem description file generated by the computing server to the first model server corresponding to the first prediction model. The hub server transmits the first solution code generated by the first model server to the database server, and outputs the first solution that the database server reads from the huge database.

Another embodiment of the present invention provides an operation method of a solution search system including a computing server, a plurality of model servers, a huge database, a database server, a hub server, and an associated database. . The operation method of the solution search system includes the hub server providing a plurality of usable prediction models to the user according to the identity of the user recorded in the associated database, and selecting the first one among the predictive models that the user can use. When predicting the model and transmitting the first problem description file to the hub server, the hub server transmits the number of the first problem description file and the first prediction model to the computing server, and the computing server describes the file and the first prediction according to the first problem. The number of the model generates a first problem description file corresponding to the first model input file of the first prediction model, and the hub server transmits the first model input file generated by the computing server to the first model corresponding to the first prediction model a server, the first model server generates a first solution code by using the first prediction model and the first model input file, and the hub server transmits the first solution code generated by the first model server to the database server, The database server reads at least one first solution from the huge database according to the first solution code. Case, the central server and database server reads the output from the massive database of at least a first solution. The invention is related to a high-efficiency solution search system, which can effectively solve the defects of the core of the current Internet of Things architecture that cannot accurately hit the problem.

1 is a schematic diagram of a solution search system 100 in accordance with an embodiment of the present invention. The solution search system 100 includes a computing server 110, N model servers 120 ₁ to 120 _N , a huge database 130 , a database server 140 , a hub server 150 , and an associated database 160 , where N is greater than 1. A positive integer.

In some embodiments of the present invention, the user may organize the problem encountered when operating a product system into a problem description file, and the problem description file may use text form to record information related to the product system problem, and the content thereof. It can include a description of the system's problem, observed phenomena and results, the subsystems associated with the product system problem, and the process in which the problem occurred (ie, how the problem can be replicated), but is not limited to the above information.

When a user attempts to use the solution search system 100 to search for a solution, the hub server 150 can provide a corresponding plurality of usable prediction models to the user according to the user identity recorded in the associated database 160. FIG. 5 is a table of the identity of the user stored in the associated database 160 according to an embodiment of the present invention.

The identity record in Figure 5 records the user account of the solution search system 100 and its corresponding password. In Fig. 5, each user's password is preset to "0000". However, in other embodiments of the present invention, other preset passwords may be used, and the user may change the password by himself. In addition, the identity record table of FIG. 5 also records the physical identity of each user. For example, the users User1, User2, and User3 are classified as general users, and the users Admin1 and Admin2 have the identity of the administrator. In some embodiments of the present invention, a user having a general user identity can utilize a predictable model of an authorized group to which the user belongs, and a user having an administrator identity is an authorizer of the authorized group. A predictive model that can be used in an authorized group can also be established. FIG. 6 is an authorized group and related materials stored in the associated database 160 according to an embodiment of the present invention. In Figure 6, the authorization groups G1 and G2 are established for the user Admin1, which respectively correspond to the product 1 and product 2 of the company A, and the authorization group G3 is established for the user Admin2, which corresponds to the company B. Product 3. For example, user Admin1 may be responsible for managing customer company A, while company A uses products such as server 1, such as server, and product 2, such as a personal computer, due to the functional attributes of products 1 and 2 may have The difference, the problems encountered and the solutions may be different, so the user Admin1 can establish the authorization groups G1 and G2 for products 1 and 2 respectively, and the authorization groups G1 and G2 will use the phases respectively. Different prediction models. In this way, in the process of the solution search system 100 establishing the prediction model, it is possible to avoid the problem that the prediction accuracy of the prediction model is low because the problems of the products of different attributes are mixed.

Furthermore, the user Admin2 may be responsible for managing the customer company B. Since the products used by the company A and the company B are also different, the user Admin2 also establishes the authorization group G3 and establishes the authorization group G3. The prediction model used, in this way, can also avoid the problem that the prediction accuracy of the prediction model is low due to the mixing of the problems of the products with different attributes.

In addition, in FIG. 5, the user Super is classified as a super administrator, and the user having the super administrator identity can establish a manager identity corresponding to the user, and can also freely use or establish each authorized group. Forecast model. In some embodiments of the present invention, the solution search system 100 can preset only one super administrator, but is not limited thereto.

In the above embodiment, if the hub server 150 compares the recorded data in the associated database 160 to the first user User1 as a general user, the authorized group to which the primary server 150 belongs is G1, and the authorized group G1 can The prediction models used are prediction models M1, M 2 and M3. Therefore, when the first user User1 uses the solution search system 100 to search for a solution, the hub server 150 can provide the available prediction models M1, M2, and M3 to the first user User1 for selection.

In some embodiments of the present invention, the prediction models M1, M2, and M3 may correspond to different data exploration algorithms, such as Bayes, CBayes, or SGD, respectively. That is to say, the user can select a preferred prediction model according to the needs.

When the user User1 selects the first prediction model M1 from the predictable models M1, M2, and M3 that can be used and transmits the first problem description file A1, the hub server 150 may describe the first problem file A1 and the first prediction model. The number of M1 is transmitted to the arithmetic server 110.

After the operation server 110 receives the number of the file A1 and the first prediction model M1 according to the first problem, the number of the file A1 and the first prediction model M1 can be described according to the first problem, and the first problem description file A1 is generated corresponding to The first model of the first prediction model M1 is entered into the file B1.

In some embodiments of the present invention, the first problem description file A1 may use a fixed format to list information related to system problems, such as but not limited to text file formats such as csv, json, and xml, so that the computing server 110 can be compared. The content of the first problem description file A1 is accurately interpreted to generate the first model input file B1. In other embodiments of the present invention, the first problem description file A1 may also use information in a non-fixed format text column related to system problems.

In an embodiment of the present invention, the computing server 110 may describe a text description attribute file of the file A1 according to the first problem. The attribute description file may be composed of a plurality of keywords, each of which is composed of a pair of attribute name and attribute value, and is one of the present inventions. In the embodiment, it can be described in the text format of json. When the first problem description file A1 uses non-fixed format text bar information related to system problems, the computing server 110 may also use a regular expression to identify the keyword name and the value of the obtained keyword. Moreover, the computing server 110 can use the standard word comparison table to compare the text of the first problem description file A1 to generate a keyword description file. Figure 7 is a partial view of a standard word comparison table according to an embodiment of the present invention. Syntactic vocabulary and vocabulary can be standardized through the standard word comparison table, so that the meaning of the keyword description file can be expressed correctly and efficiently. In addition, in order to avoid the confusion of the semantics of the keyword description file, the values of all keywords can be represented in lowercase.

After completing the keyword description file, the computing server 110 may preferentially select keywords with higher weights or user preset preferences from the keyword description file as predictors to generate predictor files, and then according to the predictive factor file. And the predictive model generates a first model input file B1. For example, the computing server 110 can adjust the predictor file according to the characteristics of the predictive model. For example, in the CBayes model, regardless of the dependency of the numbers, the numbers in the predictor file can be Partially deleted to generate the first model input file B1. However, different prediction models have different requirements for the format of the input file, and the present invention is not limited to the above embodiment.

Because the content of the standard word comparison table, that is, the number and type of standard words it contains, directly affects the content of the keyword description file and the predictive factor file, and also affects the prediction accuracy of the prediction model, so in the present invention In some embodiments, the prediction models M1, M2, and M3 may also correspond to different standard word comparison tables, respectively. That is to say, when the user with the manager's identity establishes the prediction model, in addition to selecting different data exploration algorithms, different standard word comparison tables can be established for each prediction model to conform to different product characteristics.

Since each model server 120 ₁ to 120 _N utilizes different prediction models for analysis and prediction, after the computing server 110 generates the first model input file B1, the hub server 150 can transmit the first model input file B1 to the model. The model server corresponding to the first prediction model M1 among the servers 120 ₁ to 120 _N , for example, if the prediction model corresponding to the first model server 120 ₁ is M1, the hub server 150 will be the first The model input file B1 is transmitted to the first model server 120 ₁ .

In some embodiments of the present invention, the correlation database 160 may also store a comparison table of the model servers corresponding to each prediction model for the central server 150 to refer to.

When the first model server 120 ₁ receives the first model input file B1, the first model server 120 ₁ can analyze the first model input file B1 by using the prediction model M1 corresponding to the first model server 120 ₁ . The first problem describes which types of problem description files the file A1 may be similar to, resulting in a solution code C1. The hub server 150 then transmits the first solution code C1 generated by the first model server 120 ₁ to the database server 140. The database server 140 can read at least one solution D1 from the huge database 130 according to the first solution code C1 generated by the first model server 120 ₁ , that is, each solution code may correspond to More than one solution. Finally, the hub server 150 outputs the first solution D1 that the database server 140 reads from the huge database 130.

In an embodiment of the present invention, the database server 140 and the huge database 130 can be a database server and a huge database supporting systems such as Hadoop Distribute File System (HDFS), Hadoop Map/Reduce, and Hive. Or, it can support other database systems suitable for processing huge amounts of data to meet the needs of the solution search system 100 for quickly processing and storing large amounts of data. In addition, the associated database 160 can be a general file system, such as MySql, PostgreSql, etc., in addition to storing the above-mentioned user identity and its corresponding predictive model number. A small amount and/or temporary data required for the hub server 150 to operate can be stored.

Through the above-described solution search system 100 of the embodiment of the present invention, engineers can share their experiences in solving system problems in the past, and can easily find possible solutions to shorten product development time and improve the quality of the solution. In addition, by setting the user's identity, the solution search system 100 can further search for solutions to problems encountered by different types of products by using independent prediction models, thereby avoiding the prior art being solved due to different product attributes. The problem of low accuracy of the scheme search system.

2 is a schematic diagram of a solution search system 200 according to an embodiment of the present invention. The solution search system 200 and the solution search system 100 can operate according to the same principle. However, the solution search system 200 further includes a modeling server 170. That is, the solution search system 200 enables the user to build a predictive model. When the user wants to establish a prediction model through the solution search system 200, the hub server 150 first identifies the user's identity, for example, by confirming the contents of FIG. 5, confirming that the user Admin1 has the identity of the manager, the hub The server 150 can provide a plurality of usable model parameter sets for the user Admin1, such as the model parameter sets P1 to P3. Each of the usable model parameter sets P1 to P3 can respectively correspond to different data exploration algorithms and/or different standard word comparison tables, so the user Admin1 can select the characteristics of the established prediction model according to the requirements. In addition, in order for the modeling server 170 to be able to build the model, the user Admin1 must also provide a plurality of solved second problem description files A2, and the second problem description file A2 that has been solved can be the same as the first problem description file A1. The format, and each of the resolved second problem description files A2 can contain fields related to records and system problems, such as descriptions of system problems and phenomena, subsystems to which system problems belong, processes that occur, etc. In addition to the bits, the cause description field, solution field, and solution code for the system problem can be included. In this way, the modeling server 170 can further use the corresponding data exploration algorithm to find the relationship between each of the solved second problem description files A2 and its solution, and establish a prediction model.

For example, when the user Admin1 selects the first usable model parameter group from the available model parameter groups P1 to P3, for example, the model parameter group P1 that can be used, and transmits a plurality of solved second problem descriptions. In the case of the file A2, the hub server 150 can transmit the plurality of solved second problem description files A2 and the first usable model parameter group P1 to the computing server 110. After receiving the plurality of solved second problem description files A2 and the first usable model parameter group P1, the computing server 110 describes the file A2 and the first usable model according to the plurality of solved second problems. The data exploration algorithm and the standard word comparison table corresponding to the parameter group P1 generate a second solution D2 corresponding to each solved second problem description file A2 and a corresponding second problem description file A2 corresponding to the solution. The second model enters the file B2.

The hub server 150 can transmit the second solution D2 corresponding to each of the solved second problem description files A2 to the database server 140, so that the database server 140 is included according to each second solution A2. The second solution code C2 stores each second solution D2 to the huge database 130. At the same time, the hub server 150 can describe the second solution code C2 corresponding to the second problem description file A2 and the second model input file B2 corresponding to each of the solved second problem description files A2. The first usable model parameter set P1 is transmitted to the modeling server 170, so that the modeling server 170 describes the second model input file B2 according to each of the solved second problem description files A2, and each solved first The second problem description file corresponds to the second solution code C2 corresponding to the file A2 and the data exploration algorithm and the standard word comparison table corresponding to the first usable model parameter group P1 to establish a usable prediction model.

In an embodiment of the present invention, the computing server 110 can describe the characters of the plurality of key fields in the file A2 according to each second question, such as the subsystem to which the recording problem belongs (for example, bios, bmc, power, thermal, mechanical). , hardware, os, driver) field, system problem description field and text description in the solution field, generating a second solution D2 corresponding to each second problem description file A2, each second Solution D2 includes a second solution code C2, a solution field, and a description field. In an embodiment of the present invention, the solution field can be used to record the address stored in the solution step file, such as a network address, instead of directly recording the solution steps, so that the solution step is too complicated. The related information is too large, and the burden on the computing server 110 is unnecessarily increased.

In addition, although each second problem description file A2 may have originally included its corresponding solution code, the computing server 110 may further describe other fields in the file A2 according to the second problem, such as a description field, Information such as subsystem fields to adjust the second solution code included in each of the final second solutions. For example, the solution code can be generated according to the content of the cause description field to generate a solution code, so that all solutions with the same cause description can have the same solution code.

For example, in an embodiment of the present invention, if there is a second problem description, the solution code of the file includes a plurality of subcodes, such as bios.mrc, where bios represents the second problem description file and the basic input/output system (basic input/output system). Correlation, and mrc indicates that the second problem description file system is related to the memory reference code in the basic input/output system, and the operation server 110 can describe the second problem in the file according to the information of other fields. The solution code bios.mrc is extended to bios.mrc.i2c, indicating that the second problem description file is associated with an inter-integrated circuit (I2C) of the memory reference code in the basic input/output system. That is, when the number of subcodes included in the solution code is larger, it means that the problem is classified as finer. Since the number of subcodes included in the solution code may affect the search speed and accuracy of the solution search system 200, it can be adjusted according to the needs of each system.

Each of the second solutions D2 is stored in the huge database 130 by the database server 140 according to the second solution code C2 included in each of the second solutions D2, and the modeling server 170 completes the establishment of the prediction model. The solution search system 200 can utilize the predictive model established by the modeling server 170 to search for possible solutions to system problems. In some embodiments of the present invention, the hub server 150 may transmit the established prediction model to a specified model server of the model servers 120 ₁ to 120 _N and compare the prediction model with the model server. It is stored in the associated database 160. In this way, when a user wants to use the prediction model to search for a solution, the hub server 150 can transmit the problem description file input by the user to the corresponding model server.

In the foregoing embodiment, when the database server 140 reads the first solution D1 from the huge database 130 according to the first solution code C1 generated by the model server 120, the database server 140 reads The first solution D1 may be a second solution of the second solution code C2 in the plurality of second solutions D2 that is consistent with the first solution code C1, and corresponds to the first solution code C1. The number of subcodes included in the second solution code C2 is not less than the number of subcodes included in the first solution code C1.

In addition, in an embodiment of the present invention, in the solution search system 200, the hub server 150 and the computing server 110, the database server 140, the model server 120, and the modeling server 170 are permeable to each other. And application programming interface (API) to transfer files and communications. In an embodiment of the present invention, a User Datagram Protocol (UDP) or a Transmission Control Protocol (TCP) can be used to achieve a remote procedure call between application interfaces (Remote Procedure Call). , RPC), in this way, the solution search system 200 can be constructed in a decentralized manner, which is beneficial to system expansion and maintenance.

Since the solution search system 200 of the present invention enables engineers to share their experiences in solving system problems in the past through a huge database and data mining algorithms, it is easy to find possible solutions to reduce the time to solve product problems. And improve the quality of the solution. In addition, by setting the user's identity, the solution search system 200 can be further configured to establish independent prediction models and search for solutions to problems encountered by different types of products, thereby avoiding the prior art because of different product attributes. The problem of low accuracy of the solution search system.

FIG. 3 is a flow diagram of an operation method 300 of the solution search system 100 in accordance with an embodiment of the present invention. The operation method 300 of the solution search system includes steps S310 to S380:

S310: The hub server 150 provides a plurality of usable prediction models M1, M2, and M3 to the first user User1 according to the identity of the user User1 recorded in the association database 160.

S320: When the first user User1 selects the first prediction model M1 from the predictable models M1, M2, and M3 that can be used and transmits the first problem description file A1 to the hub server 150, the hub server 150 describes the first problem. The number of the file A1 and the first prediction model M1 is transmitted to the computing server 110;

S330: The operation server 110 describes the number of the file A1 and the first prediction model M1 according to the first problem, and generates a first problem description file A1 corresponding to the first model input file B1 of the first prediction model M1;

S340: a first hub server model input 150 of the operational arising from the server 110 to transfer files B1 model server ₁₂₀₁ to 120 _N corresponding to the first prediction model M1 of the first model server _1201;

S350. The first model server 120 ₁ generates a first solution code C1 using the first prediction model M1 and the first model input file B1;

S360: The hub server 150 transmits the first solution code C1 generated by the first model server 120 ₁ to the database server 140;

S370: The database server 140 reads at least one first solution D1 from the huge database 130 according to the first solution code C1;

S380: The hub server 150 outputs the at least one first solution D1 that the database server 140 reads from the huge database 130.

4 is a flow chart showing an operation method 400 of the solution search system 200 in another embodiment of the present invention. The method 400 of operation of the solution search system 200 includes steps S410 through S460:

S410: The hub server 150 provides a plurality of available model parameter groups P1, P2, and P3 to the second user Admin1 according to the identity of the second user Admin1 recorded in the associated database 160.

S420: When the second user Admin1 selects the first usable model parameter group P1 from the available model parameter groups P1, P2 and P3 and transmits the plurality of solved second problem description files A2, the hub server 150 Transmitting a plurality of solved second problem description files A2 and the first usable model parameter group P1 to the computing server 110;

S430: The computing server 110 generates a second solution D2 corresponding to each of the solved second problem description files A2 according to the plurality of solved second problem description files A2 and the first usable model parameter group P1. A second problem that has been solved describes the second model input file B2 corresponding to the file A2;

S440: The database server 140 stores each second solution D2 according to the second solution code C2 included in each second solution D2;

S450: The hub server 150 describes the second solution code C2 corresponding to the second problem description file A2 and the second model input file B2 corresponding to each of the solved second problem description files A2. a usable model parameter set P1 is transmitted to the modeling server 170;

S460: The modeling server 170 describes the second model input file B2 of the file A2 according to each of the solved second problem descriptions, and the second solution code C2 corresponding to the second problem description file A2 and the first solution. A data mining algorithm corresponding to the model parameter set P1 can be used to establish a predictive model that can be used.

Through the operation methods 300 and 400 of the solution searching system 100, 200 of the above embodiment of the present invention, the algorithm of the huge database and the data exploration can be utilized to enable the engineers to share their experiences in solving system problems in the past, and can be easily searched. Probable solutions to reduce time to resolve product issues and improve solution quality.

In summary, the solution search system and the solution search system operation method of the embodiments of the present invention can utilize a huge database and a data exploration algorithm to assist users in sharing their past experience in solving problems, while using When discovering system problems, it is easy to find possible solutions to reduce the time to resolve product issues. In this way, it is possible to avoid the problems in the prior art because the related solutions are not easy to search, and the efficiency and quality of solving the system problems are difficult to control. In addition, by setting the user's identity, the solution search system can further enable the independent prediction model to search for solutions to problems encountered by different types of products, thus avoiding the prior art solution search due to different product attributes. The problem of low system accuracy. The above are only the preferred embodiments of the present invention, and all changes and modifications made to the scope of the present invention should be within the scope of the present invention.

100, 200‧‧‧ Solution Search System
110‧‧‧ Computing Server
120 ₁ to 120 _N ‧‧‧Model Server
130‧‧‧ huge database
140‧‧‧Database Server
150‧‧‧Central Server
160‧‧‧Related database
170‧‧‧Modeling Server
User1, Admin1‧‧‧ user
A1‧‧‧First problem description file
B1‧‧‧ first model input file
C1‧‧‧First Solution Code
D1‧‧‧ first solution
A2‧‧‧Second problem description file
B2‧‧‧Second model input file
C2‧‧‧Second solution code
D2‧‧‧Second solution
P1, P2, P3‧‧‧ model parameter set
300, 400‧‧‧ method
Steps S310 to S380, S410 to S460‧‧

FIG. 1 is a schematic diagram of a solution search system according to an embodiment of the present invention. 2 is a schematic diagram of a solution searching system according to another embodiment of the present invention. FIG. 3 is a flow chart of an operation method of the solution searching system according to an embodiment of the present invention. 4 is a flow chart of an operation method of a solution search system according to another embodiment of the present invention. Figure 5 is a table showing the identity of a user stored in an associated database according to an embodiment of the present invention. FIG. 6 is an authorized group and related materials stored in the associated database according to an embodiment of the present invention. Figure 7 is a partial view of a standard word comparison table according to an embodiment of the present invention.

300‧‧‧ method

S310 to S380‧‧‧ steps

Claims (10)

A solution search system, comprising: a computing server, configured to receive and according to a problem description file and a number of a prediction model corresponding to the problem description file, generating a problem description file corresponding to one of the prediction model model inputs Archive; a plurality of model servers, each model server corresponding to a prediction model, each model server is configured to use the prediction model corresponding to the model server and the model when receiving a model input file The input file generates a solution code; a huge database for storing the solution; a database server coupled to the huge database for generating the solution code according to the model server The huge database is configured to read at least one solution; a hub server coupled to the computing server, the model server, and the database server for: providing a plurality of characters according to a first user The predictive model that can be used is selected by the first user, and the predictive models that can be used correspond to different data explorations respectively. The first user's identity is a general user; when the first user selects a first prediction model from the predictable models that can be used and transmits a first problem description file, the first question is Transmitting the number of the description file and the first prediction model to the computing server; transmitting, by the computing server, the first problem description file corresponding to the first model input file of the first prediction model to the models a first model server corresponding to the first prediction model in the server; transmitting, by the first model server, one of the first solution codes to the database server; and outputting the database server by The at least one first solution is read by the mega-database; and an associated database is coupled to the hub server for storing the identity of the first user, the first user corresponding to the Some predictable models that can be used and the small and/or temporary data required for the hub server to operate. The solution search system of claim 1, further comprising a modeling server coupled to the hub server, wherein the hub server is further configured to: provide a plurality of available according to the identity of a second user The model parameter set is selected by the second user, and the available model parameter sets respectively correspond to different data exploration algorithms and/or different standard word comparison tables, and the second user's identity a manager; when the second user selects a first usable model parameter set from the set of available model parameters and transmits a plurality of solved second problem description files, the solved Transmitting a second problem description file and the first usable model parameter set to the computing server, so that the computing server generates each of the solved second problem description files and the first usable model parameter set A second problem solved by the second problem description file and a second model input file corresponding to each of the solved second problem description files; each second question that has been solved Transmitting the second solution corresponding to the profile to the database server, so that the database server stores each second solution according to one of the second solution codes included in each second solution; and The second solution input file corresponding to the second solution code corresponding to each of the solved second problem description files and the second model input file corresponding to each of the solved second problem description files and the first usable model parameter set Transmitting to the modeling server, so that the modeling server inputs the second model input file according to each of the solved second problem description files, and the second corresponding to each solved second problem description file The solution code and the first set of model parameters that can be used establish a predictive model that can be used. The solution search system of claim 2, wherein: the hub server is further configured to establish and/or change the identity of the user and set the first user to use according to the control of the second user. The predictive model is usable; and the hub server is further configured to establish and/or change the identity of the second user and set the second user according to the control of the third user having one of the super administrators. The set of model parameters that can be used can be established. The solution search system of claim 2, wherein: each of the resolved second problem description files includes a solution field, where the solution field stores an address stored in a resolution step file; a second solution includes the second solution code and a solution field corresponding to each of the solved second problem description files; and the computing server describes at least one key in the file according to each second question The text in the field selects a plurality of category keywords to be combined into the second solution code corresponding to each second problem description file. The solution search system of claim 1, wherein: the computing server selects a plurality of keywords from the text of the first problem description file to generate a keyword description file, and generates a file according to the keyword description file. The predictive factor file, and the first model input file is generated according to the predictive factor file and the data exploration algorithm corresponding to the first predictive model. A method for operating a solution search system, the solution search system comprising a computing server, a plurality of model servers, a huge database, a database server, a hub server, and an associated database, The method includes: the hub server provides a plurality of usable prediction models to the first user according to the identity of the first user recorded in the relational database, and the predictive models that can be used respectively correspond to To a different data exploration algorithm, the first user's identity is a general user; when the first user selects a first prediction model from the available prediction models and transmits a first problem description file And to the hub server, the hub server transmits the first problem description file and the number of the first prediction model to the operation server; the operation server describes the file and the first prediction model according to the first problem a number that generates the first problem description file corresponding to the first model input file of the first prediction model; the hub server stores the operation Transmitting the first model input file generated by the server to the first model server corresponding to the first prediction model in the model server; the first model server utilizing the first prediction model and the first The model input file generates a first solution code; the hub server transmits the first solution code generated by the first model server to the database server; the database server is according to the first solution The code reads the at least one first solution from the huge database; and the hub server outputs the at least one first solution that the database server reads from the huge database. The method for operating a solution search system according to claim 6, wherein the solution search system further comprises a modeling server, the method further comprising: the hub server according to one of the records in the associated database The user's identity provides a plurality of usable model parameter sets for the second user to select, the available model parameter sets respectively correspond to different data exploration algorithms, and the second user's identity management When the second user selects a first usable model parameter set from the set of available model parameters and transmits a plurality of solved second problem description files, the hub server Solving the second problem description file and the first usable model parameter set are transmitted to the computing server; the computing server describes the file and the first usable model parameter set according to the second problem solved Generating a second solution corresponding to the second solution corresponding to each of the solved second problem description files and the second model input file corresponding to each of the solved second problem description files; The database server stores each second solution according to one of the second solution codes included in each second solution; the hub server associates the second file corresponding to each of the resolved second problem description files The solution code and the second model input file corresponding to each of the solved second problem description files are transmitted to the modeling server by the first usable model parameter set; and the modeling server is configured according to each The second model input file of the second problem description file that has been solved, the second solution code corresponding to each of the solved second problem description files, and the first usable model parameter set are used to establish a usable Forecast model. The operating method of the solution search system of claim 7, further comprising: the hub server establishing or changing the identity of the first user according to the control of the second user and setting the user to use the a predictive model that can be used; and the hub server establishes or modifies the identity of the second user and sets the second user to establish based on the control of a third user having a super administrator status A model parameter set that can be used. The method for operating a solution search system according to claim 7, wherein: each of the resolved second problem description files includes a solution field, and the solution field stores one bit stored in the solution step file. Each second solution includes the second solution code and a solution field corresponding to each of the resolved second problem description files; and the computing server is described according to the second problem that has been solved The step of generating the second solution corresponding to each of the solved second problem description files by the file and the first usable model parameter set includes the operation server describing a plurality of key columns in the file according to each second question The text in the bit selects at least one category keyword to be combined into the second solution code corresponding to each second problem description file. The operation method of the solution search system of claim 6, wherein the operation server generates the first problem description file corresponding to the first prediction model according to the first problem description file and the number of the first prediction model. The step of inputting the file into the first model includes: the computing server selecting a plurality of keywords from the text of the first problem description file to generate a keyword description file; the computing server generates a file according to the keyword description file The predictive factor file, and the computing server generates the first model input file according to the predictive factor file and the data exploration algorithm corresponding to the first predictive model.