JP2005339349A - Method and system for allocating coping method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically adjust variation of coping methods by every factor stratification from data content on the way of collection to perform randomizing allocation in an allocation system utilizing a computer. <P>SOLUTION: Information required for allocation is defined and the information to be defined is considered as a combination pattern table of stratification factors, coping methods and cures, a spreadsheet by every stratification factor, a function formula for specifying object factors and a probability variation formula of the coping methods. Allocation is performed to inputted data one by one utilizing these definitions. As for an allocation method, an imbalanced factor based on the defined function formula is specified in allocation ratio of an applicable spreadsheet by every stratification factor from a factor pattern included in the inputted data, allocation probability is calculated to the specified factor with the defined function formula and the probability variation formula using the allocation ratio of the spreadsheet as input and random allocation of the coping methods and the cure is performed based on the calculation result. Allocation results are accumulated in the spreadsheet and utilized in the next allocation request data. Thus, variation is automatically adjusted. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technique for assigning a treatment method, a solution, or a countermeasure to a subject in a research / survey field, a test subject, a survey subject, or a survey target using a computer. In particular, the present invention relates to a technique for dynamically allocating non-clinical trials, clinical trial treatments, solutions, or countermeasures in the medical, pharmaceutical, and bio fields using factor stratification.

  In the field of research and research, comparative studies are frequently conducted to confirm the effectiveness and effectiveness of countermeasures and treatments. In the comparison target test, first, factor information that is a condition of a sample and the number of populations that are the number of samples to be collected are set. That is, factor information and a population are set. Next, a sample is collected based on the set factor information and population, and the collected sample is randomly sampled. Finally, countermeasures, treatments, etc. are randomly assigned to the sampled samples.

  Patent Document 1 describes an example of a method for assigning a drug to be tested to a subject in a clinical test. In the example described in Patent Document 1, among the drugs to be administered, from the number of assigned subjects with the largest number of assigned subjects in the entire clinical trial, the assigned subjects with the smallest number of assigned subjects in the entire clinical trial. A test drug or a pseudo drug is assigned to the test candidate so that the difference obtained by subtracting the number is less than 2.

JP 2004-29894 A

  In the conventional allocation method, a treatment method, a drug, and the like cannot be allocated unless the subjects as a population are determined.

  However, in clinical trials examining the effects or efficacy of treatments or drugs for specific diseases, it is often difficult to collect as many patients as needed with predetermined factor information. In such cases, subject collection and treatment or drug assignment to the subject must be performed in parallel.

  It is an object of the present invention to provide a computer-based allocation system without setting factor information and the number of populations of samples to be collected, that is, while collecting samples, the variation of countermeasures according to factor layers. The object is to provide a system that performs automatic adjustment and performs random assignment.

  In order to automatically adjust the distribution of countermeasures by factor layer based on the data contents in the process of collection, assign the factor pattern from the factor information contained in the data to the unit of the allocation request of the collected data, and assign the currently assigned A core engine that adjusts the allocation ratio of each factor layer and implements allocation, that is, a core engine is necessary. By making this core engine mechanism independent of allocation request units, it is possible to promote balanced allocation in a timely manner even if the allocation request unit is not the population itself as in the prior art.

  However, in order to utilize the mechanism of the core engine, it cannot be used as a real problem unless there is a part that becomes the outer shell at the same time. Immediateness is also required to perform collection and allocation in parallel. The input data is not necessarily in the same format as the information required for allocation. In addition, each factor item is not necessarily stored in the same classification system. Therefore, it is necessary to have a device that becomes the gateway for the input data and the core engine. If this requirement is taken into consideration, a series of operations from input data to assignment becomes possible, but there are cases where the collected factor does not include the assumed factor pattern.

  When such a case is assumed, a simulation function of the allocation result and a factor-by-factor pattern based on the simulation result, that is, a function of changing the allocation pattern are required regardless of whether the allocation is executed or not.

  Finally, since it has a function capable of dynamically changing the allocation pattern, it is preferable to be able to specify the allocation pattern in which the allocation is performed from the management result, that is, from the auditing side, from the allocation result.

  However, in the method of calling the core engine for this minimum allocation request unit, it is considered that the processing load becomes considerably large when a large amount of data is handled.

In summary, the following measures are necessary.
・ Realization of a core engine that assigns factors by assigning factor patterns from factor information included in the data to the main issues and assignment request units, and adjusting the assignment ratio of the currently assigned factor layers.
・ Gateway function for input data and core engine.
・ Allocation result simulation function.
・ Audit trail function.
・ Measures to prevent deterioration of core engine performance.

Realization of the core engine To realize the core engine, firstly, means for automatically adjusting the balance between the coping method and the treatment method is necessary. The automatic adjustment function can be roughly divided into three processes.

  The first is a process of matching factor information contained in input data with each factor layer that has already been determined. The second is the process of automatically identifying the factors that are the most unbalanced among the matched factors. And the third is to increase the random occurrence probability of measures / treatments in order from the lowest to the proportion of treatment / treatment decisions for the identified unbalanced factors. This is a process for preferentially setting the allocation probability for a treatment method.

  In the first process, the factor information contained in the input data can be efficiently matched with each factor layer that has already been determined based on the combination of the assumed factor pattern and the coping / treatment pattern. A data generation pattern is generated in advance. Since all the patterns that can be generated can be covered by this method, each determined factor layer can be easily specified from the factor information included in the input data. This pattern table is easily generated by using classification information of stratification factors registered in advance in the master and classification information of countermeasures / treatments.

In the second process, it automatically identifies the factors that have the most unbalanced response / treatment decisions. First, a countermeasure / treatment pattern allocation ratio column is provided in the pattern table described above, and a difference detection calculation formula for detecting an imbalance is embedded with reference to the ratio column. Next, this difference detection calculation is performed for each factor pattern of the input data. Next, for each factor layer that was matched in the first process, that is, for each factor layer data row that matched the factor pattern contained in the input data, the difference detection calculation result contained in that row To determine the most unbalanced factor. For example, when the factor pattern of the input data is composed of factor 1 and factor 2 and factor 1 = 0 and factor 2 = 1, the difference detection calculation results of factor 1 = 0 and factor 2 = 1 are compared. In this example, if the calculation result of the difference detection of the factor 1 = 0 indicates 5% and the calculation result of the difference detection of the factor 2 = 1 indicates 20%, the unbalanced factor is 20% which indicates 20%. It becomes. When this procedure is generalized, it is expressed as a function formula for calculating the difference detection for each factor row and a function formula for specifying an unbalanced factor from the calculation of the difference detection for the target row group. Therefore, by providing this functional interface, it is possible to freely set the above-described calculation formula.

  Finally, in the third process, change the decision rate for the treatment and treatment of the identified factors. The fact that the balance between the coping method / therapeutic method is established means that the allocation ratio of each coping method / therapeutic method is equal to the average value when the ratio of the parameters is 100%. For example, when the number of treatment groups is 2, it is desirable that the allocation ratio of each countermeasure / treatment method is close to the average value of 50%. Similarly, when the number of treatment groups is 3, it is desirable that the allocation ratio of each countermeasure / treatment method is close to the average value of about 33%.

Therefore, when there are 3 treatment groups of A agent, B agent, and C agent, and the allocation ratio is A = 20%, B = 30%, C = 50%, pay attention to the minimum allocation ratio A = 20% Then, if the allocation ratio of agent A is set to a probability higher than the average occurrence probability of about 33%, the allocation to agent A will be promoted preferentially. A value close to about 33% of the average value can be obtained. For example, with respect to the probability setting, only the difference whose current allocation ratio is lower than the average value is added. The allocation ratio of agent A is obtained by the following formula.
(Assignment ratio) = (number of parameters% ÷ number of treatment groups-minimum treatment% assignment) + (number of parameters ÷ number of treatment groups)

If the present example is applied to this formula, the allocation ratio of agent A is as follows.
(Allocation ratio) = (100% ÷ 3-20%) + (100% ÷ 3) = about 46%
Next, the allocation ratio of B agent is calculated. Since the allocation ratio of Agent A is about 46%, the target parameters that can be allocated to Agent B are as follows.
(Allocation ratio) = 100% -46% = 54%. The number of treatment groups is 3-1 = 2.

If only the difference lower than the average value is added to this parameter as in the case of agent A, the following formula is established.
(Allocation ratio) = population% ÷ number of treatment groups – (number of populations x minimum treatment allocation ÷ 100%)) + (number of populations ÷ number of treatment groups)
If the present example is applied to this formula, the allocation ratio of the B agent is as follows.
(Allocation ratio) = (54% ÷ 2- (54% × 30 ÷ 100%)) + (54% ÷ 2) = Approx. 38%

Finally, the allocation ratio of C agent is calculated. Since the whole was regarded as 100%, the allocation ratio of the C agent is expressed by the following formula.
(Allocation ratio) = 100%-Agent A allocation%-Agent B allocation% = 100%-About 46%-About 38% = 16%
That is, the allocation probability of each treatment method is sequentially obtained by the following calculation formula, and the remaining ratio ratio may be applied as the allocation ratio of the remaining treatment method.
(Allocation ratio) = (number of parameters% ÷ number of treatment groups – (number of parameters × minimum treatment allocation% ÷ 100%))
+ (Number of populations ÷ number of treatment groups)

  When this procedure is generalized, it is expressed by a loop frame using the function expression and the result of the function expression. Therefore, by providing this functional interface and loop frame, the above-described calculation formula can be freely set.

  In order to calculate the probability of occurrence of allocation, assuming that the overall allocation ratio is 100% and the number of allocations is 100, the number of allocations is sequentially accumulated for each of the allocated treatment methods. Agent A = 46, Agent B = 46 + 38 = 84, Agent C = 46 + 38 + 16 = 100. In other words, by setting A agent for 1 to 46, B agent for 47 to 84, C agent for 85 to 100, the occurrence probability of A agent, B agent, and C agent can be set from 1 to 100. It can be regarded as an act of arbitrarily selecting one piece of data. Then, a 100 remainder system using a random function may be adopted for the act of arbitrarily selecting data.

  As described above, by following the three processes, it is possible to perform automatic adjustment control between the stratified factors and the leveling of the countermeasures and treatment methods within the stratified factors.

  In order to implement the three process controls, it is a condition that the allocated data is appropriately accumulated. In order to balance the stratified factors, it is necessary to store the allocation result for each factor element included in the input data. For example, it is assumed that an allocation request for one piece of input data is executed and assigned to agent A as a countermeasure / treatment. The number of processed data is one, but if factor 1 and factor 2 exist in the factor item of this input data, and each value is factor 1 = 0 and factor 2 = 1, respectively The cumulative number of methods and treatments is that factor 1 = 0 is agent A = + 1 and factor 2 = 1 is agent A = + 1. If this cumulative addition is performed for each allocation determination request event for the coping method / treatment method, it is possible to calculate the random occurrence probability of the coping method / therapeutic method using information obtained by feeding back the previous determination result for each request.

  By realizing the above data accumulation process, it is possible to establish the data base necessary for automatic adjustment control, and as a result, automatic adjustment of the leveling of coping and treatment methods between stratified factors and within stratified factors can be performed. I can do it.

  The process and calculation formula used for the allocation means should not be fixed. The calculation formulas for identifying factors and determining the ratio of countermeasures and treatment methods explained in the procedure of automatic adjustment control are released as a function formula interface so that users can set formulas by referring to items stored in the database. It may be. Thereby, the general setting of a specific factor and a ratio determination algorithm can be performed.

  In specifying the factors to be assigned and the measures and treatments, step-by-step usage using the results of the function formula is necessary. For example, when changing the allocation probability of a coping method / therapy, first change the allocation probability of the lowest coping method / therapy based on the overall parameter, 100%, and then the second coping method.・ The calculation of the allocation probability of a treatment method needs to be calculated using the remaining percentage as a parameter. If the processing in such a case is defined using a function expression, a loop algorithm that sets the result of the function expression as an input argument of the function expression is required. However, in this example, since the usage is limited to subtraction of parameters, it is possible to prepare in advance a statement instruction that transfers the result of the function expression to a variable used as an argument. Therefore, by preparing a loop frame incorporating a variable statement instruction and specifying the prepared variable as the argument of the function expression, it is possible to create a loop algorithm only by setting a single function expression. . By taking the above measures, the core engine function can be realized.

Input data and the gateway function of the core engine Next, the input data and the gateway function of the core engine. As one of the realization requirements, environmental factors with high variability of factor classification rules, that is, temporal factors and spatial factors Possible correspondence. In the case of these factor elements, the invariance of the value stored in the input data is high, and it is difficult to specify the factor elements directly. For example, in clinical trials and the like, there is a technique called facility block allocation, which is performed for the purpose of keeping the allocation balance of the coping method / therapeutic method equally for each facility as well as the stratification factor. In this case, it is assumed that there are A facility, B facility, and C facility, A facility and B facility are treated as I blocks, and C facility is treated as II blocks. Furthermore, it is assumed that a new subject is registered here and the subject is treated as a II block at the D facility. In this case, it is the allocation site that determines the block, not the data generation site. Accordingly, only the value of the D facility exists in the input data, and it is necessary to convert the D facility to the II block at the allocation site. Master conversion means and master maintenance means of this facility and block, in other words, if there is a master conversion means and master maintenance means for the spatial elements of the factor, the requirement requirements can be realized simply by adding a master record Means can be provided.

  This function is also effective for the time factor. For example, it can be applied to the requirement of what time zone the morning and afternoon factor elements correspond to.

  The system of the present invention incorporates into the system functional requirements for specifying an input data file generated in an allocation operation, selecting a factor item included in the data file, converting a factor item code, and extracting and selecting data according to the allocated state. It has a function. The input data file is specified by a procedure for selecting a file from the screen. The selection of the factor item is realized by a method of displaying a list of items of the selected input file on the screen and selecting a necessary item as a factor item. Also, the conversion of the factor item code is executed by means for describing a judgment formula in the selected factor item. Finally, in the data extraction / selection based on the assigned state, the state indicating whether or not the execution is completed at the time of the assignment is output to the item of the output data record of the assignment result. When reading input data, the field value of the allocation status specified from the screen and the allocation status of the output data record are compared, and the corresponding output data record is detected as a match or mismatch. Input data that matches the key is selected.

  In this way, it is possible to provide functions for performing assignment, displaying the result of assignment, and performing reassignment without processing or updating input data. By taking the above measures, the gateway function of the input data and the core engine can be realized.

Allocation result simulation function This is a means for realizing the allocation result simulation function. When the state before allocation is assumed, it is necessary to first simulate the allocation result by combining a plurality of stratified factors. This requires multiple input information patterns that define combinations of input data and stratification factors. Since the purpose is to compare simulation results, a plurality of simulation results are required. When the combination of stratification factors is determined from the simulation results, the next pattern is used to output to the allocation file for actual operation. Therefore, the input pattern used in the simulation and the output pattern defining the output file are combined. Considering this, it is possible to set multiple input data files, information resources related to output, and multiple output data files, and use a mechanism that can dynamically connect input information resources and output information resources during allocation. Function can be realized. In the simulation, there is a situation where the test data is used before the actual allocation data. Even in such a situation, if the test data and the production data are switched by the input file selection function described above, the simulation input information resource tested with the test data can be easily switched to the production data use. This suggests not only a simulation function but also versatile operability.

Even in the simulation function described above, the situation is slightly different when it is being assigned. During allocation, part of the input data has already been allocated. Therefore, even if the combination of stratification factors is changed, the allocation result cannot be changed. For example, in the case of a clinical trial, input data that has been assigned means that treatment has already started for the subject based on the assigned treatment. Therefore, when changing the combination of stratification factors, the allocation information of the countermeasures / treatments output in the assignment results and the input data including the combination of the stratification factors to be changed are combined with the assignment result as the key. It is necessary to reverse-generate the allocation summary table for another factor. Since the allocation result and the input data have a 1: 1 relationship, a combined file can be easily created by using a tool such as an SQL command. The problem is how to generate an allocation summary table of stratified factors from the allocation result key of this combined file. The stratified factor assignment summary table includes an assignment decision processing unit that decides the allocation of coping and treatment methods based on factor information, and the result value in the stratified factor assignment summary table based on the decided assignment key and factor information. Are accumulated. Therefore, the processing can be realized by skipping the processing of the allocation determination processing unit and providing a function for directly passing the allocation key and the key identification element of the factor information to the portion where the result values are accumulated in the allocation summary table.

  For example, when allocation is performed using the input information pattern A and the output information pattern A, it is necessary to switch to the pattern B. At this time, the environment of the input information pattern B and the output information pattern B by the combination of the new stratification factors is prepared, and input is first made from the output result of the output information pattern A by using the skip function and the input file selection function described above. The allocation summary table of the stratified factors for the information pattern B is reversely generated. After the test for this is performed, the output information pattern is returned to A, and the allocation for actual operation is performed.

  When such a simulation function is used, in the output data file of the allocation result, the data record assigned with the allocation pattern based on the stratification factor of the input information pattern A and the allocation based on the stratification factor of the input information pattern B The result is a mix of data records assigned by pattern.

Audit trail function Considering the audit trail, it is necessary to specify the allocation method from the data record. As a solution to this, according to the present invention, there is provided a mechanism for holding key information for binding to an input information resource used as a unit for determining a countermeasure / treatment method for an output data file. Since a form is formed in which input information patterns used as units for dealing with allocation results and treatment methods are linked, the subsequent input data and the definition information on factor allocation can be specified.

Measures to prevent performance degradation of the core engine Finally, it becomes a means to prevent performance degradation of the core engine, but the information necessary for calculating the probability of random occurrence is a unit that does not require data recalculation, that is, stratified factor pattern and countermeasure -By using a data key structure that can be stored on the storage medium in units of treatment patterns and the data block of the stored units can be read directly for each decision request event for treatment / treatment, treatment / treatment This makes it possible to avoid recalculation and re-reading of data blocks other than the factors required for the determination request and the countermeasure / treatment method.

  Implement the core engine implementation means, input data and core engine gateway function implementation means, allocation result simulation function and associated audit trail function implementation means, and core engine performance deterioration prevention means. Therefore, a system technique to perform randomization by automatically adjusting the variation in the countermeasures for each factor layer from the contents of the data being collected without pre-setting the factors of the target collected data and the number of populations. This can be realized by using a computer system including actual operation.

  According to the present invention, the factor information of the sample to be collected and the number of populations are not set, that is, while collecting the sample, the variation in the countermeasures according to the factor layer is automatically adjusted and randomly assigned. It can be carried out.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 shows a configuration of an allocation system according to the present invention. The allocation system includes an allocation processing device 200 and includes definition information 100, 101, and 102. The definition information includes a treatment group / countermeasure group table 100, a block table 101, and a stratification factor table 102. The definition information is set in the allocation system, the subject input data 910 is input, the allocation is performed according to the stratification factor, and the allocation result is output to the allocation table table 909. When the allocation result is used, the allocation table table 909 may be displayed alone, but the allocation table table 909 is key-coupled with the subject input data 910 and displayed on the display device 1000.

  Next, the configuration of the allocation processing device 200 will be described. The allocation processing device 200 includes an allocation information definition 201, an allocation execution definition 202, and an allocation execution 203. The configuration shown here is an example of an embodiment of the invention, and what is essential for the invention is the component, and other functional configurations may be adopted. In the present specification, the function or configuration may be configured as a program that can be executed by a computer.

  The assignment information definition 201 includes a treatment / countermeasure information definition 204, a block / stratification factor information definition 205, a pattern information definition 206, a subject input data conversion definition 207, and an assignment information definition 208. The treatment / coping information definition 204 and the block / stratification factor information definition 205 subdivide each element and register it as identification information in the master. The pattern information definition 206 defines a plurality of allocation information and output information for each pattern. The subject input data conversion definition 207 uses the pattern defined by the pattern information definition 206 to define a procedure for converting input data from the subject input data 910. The allocation information definition 208 defines allocation patterns such as stratified factors and coping / treatment methods.

  The function of the allocation execution definition 202 mainly controls allocation of input / output resources, which will be described later in detail. The allocation execution 203 includes subject input data conversion processing 213, input record allocation calculation 214, and allocation table data addition / update 215. The subject input data conversion processing 213 performs input data conversion processing using the subject input data conversion definition 207. The input record allocation calculation 214 executes the allocation of the converted input data based on the definition information of the allocation information definition 208. The allocation table data addition / update 215 outputs the allocation result to the allocation table table 909.

  FIG. 2 shows a configuration model of a database necessary for realizing the functions of the allocation system of FIG. In the database of this example, each database file is classified and described into an allocation information definition table group 900, an allocation execution definition table group 901, and an allocation execution table group 902.

  The characteristics of this database structure are specified below. First, allocation information is managed in units of allocation patterns by an allocation information definition table 904 and an allocation item management table 903. The information division table 905 shares resources of the basic elements of the definition 204 of treatment / handling information and the definition 205 of block / stratification factor information. In this way, allocation information can be managed in units of allocation patterns using basic elements.

By storing the tabulation table for each stratification factor in the database of the allocation tabulation table 906 and the allocation tabulation table 907, occurrence of data re-tabulation calculation is suppressed and performance deterioration is prevented. These allocation information definition table group 900 and the allocation table table 909 positioned in the allocation execution table group 902, the allocation pattern Seq930 key of the allocation management table 908 positioned in the allocation execution definition table group 901, and the output destination Since it is configured to be linkable by the pattern Seq931 key, operation variations described later are possible.

  FIG. 3 shows the input conversion function of the allocation system according to the present invention. FIG. 3 shows a flow in which the subject input data group 912 that is an input target is converted by the subject input data conversion definition 207 and the assignment subject input data 239 that is the conversion result is generated. The subject input data conversion definition 207 has four functions of an input file selection process 216, a block solution process 217, a stratified factor combination process 218, and a factor determination function process 219.

  The input file selection process 216 has a dynamic input file selection function. If the input file A is selected when the input file A 220 and the input file B 221 exist as input data targets, the selected information is stored in the input file name 944 of the allocation information definition table 904. Next, data mapping with the allocation information factor managed by the allocation system is performed using the selected input file. The data mapping function includes a block solution process 217, a stratified factor combination process 218, and a factor determination function process 219. In the block solution processing 217, a combination pattern of input data and an allocation information factor is registered in advance in the master, and the input data is converted into an allocation information factor using this master.

In the illustrated example, A facility 232 and B facility 233 are converted into I Block 236, and C facility 234 and D facility 235 are converted into II Block 237. The function of the stratified factor combination processing 218 is a function of selecting factor data from the input data element. In the illustrated example, each element of the input data, element 1 (224), element 2 (225), element 3 (226), element 4 (227) to factor 1 (229), factor 2 (230), factor 3 ( 231) is selected.

  This selection result is stored in the input item name 941 and output item name 935 of the assigned item management table 903 for each item. The factor determination function processing 219 converts the value for each element selected in the stratified factor combination processing 218 using a determination description statement when the value of the input data element does not match the classification element of the factor data of the system. In the illustrated example, the content specified in the factor determination instruction statement 238 is stored in the factor function specification 942 of the assignment item management table 903.

  The input data is converted by using these definition information, the assignment information definition table 904, and the assignment item management table 903 at the time of input conversion. In the example shown in the figure, two records included in the input file A220, a subject input data (S001) record 240 and a subject input data (S002) record 241 are assigned to a subject input data for assignment (S001) record 242 and for assignment, respectively. It is converted into subject input data (S002) record 243. In this example, the facilities 240 or the time zone 223 of the records 240 and 241 of the input file A220 are converted into the allocation block 228 of the records 242 and 243 of the assignment subject input data through the block solution processing 217, and the elements of the records 240 and 241 1 (224), element 2 (225), and element 3 (226) are converted into factor 1 (229), factor 2 (230), factor of records 242 and 243 through stratified factor combination processing 218 and factor determination function processing 219. 3 (231).

  FIG. 4 shows a dynamic stratification allocation function of the allocation system. The left half of FIG. 4 shows the allocation information definition 208 of the allocation system, and the right half shows the flow of processing. The assignment information definition 208 includes three blocks: a pattern generation 270 for all factors and countermeasures, a countermeasure aggregation 274 for each stratified factor, and a probability arithmetic expression definition 279 using the stratified factor item and the countermeasure aggregation. Including. In the all factor / countermeasure pattern generation 270 block, a stratified factor and coping / treatment pattern generation of a factor pattern summary table composed of stratified factors and coping / treatment methods is performed. In this specification, “factor pattern summary table” is described as “allocation summary table”, and hereinafter referred to as “assignment summary table”.

  The all factor / countermeasure pattern generation 270 includes a stratified factor combination pattern generation 271, a countermeasure combination pattern generation 272, and a stratified factor / countermeasure combination pattern generation 273. The stratified factor combination pattern generation 271 generates a stratified factor combination pattern based on the block information 947 and the stratified factor information 948 registered in the allocation information definition table 904. The coping method combination pattern generation 272 generates a coping method combination pattern based on the treatment / coping method information 946 registered in the allocation information definition table 904. The stratified factor / countermeasure combination pattern generation 273 further combines the stratified factor combination pattern and the countermeasure combination pattern to generate a stratified factor / countermeasure combination pattern. The allocation tabulation table that is the generation result is registered in the allocation tabulation table 906.

  The stratified factor coping table generation 274 includes a stratified factor coping method 275, a stratified factor coping method 276, a stratified factor coping rate 277, and a stratified factor coping method difference rate 278. Each element of is aggregated according to stratification factor, and the aggregation table according to stratification factor is generated. In this specification, the “total table for each stratified factor” is described as “allocation total table”, and is hereinafter referred to as “allocation total table”. The procedure of generation 274 of the countermeasure total table for each stratified factor is the same as the procedure of pattern generation 270 of all factors and countermeasures, and outputs the allocation total table as a generation result to the allocation total table 907.

  The definition 279 of the probability calculation expression using the stratified factor item and the countermeasure aggregation includes a definition 280 of the specific process of the allocation target factor and a definition 281 of the probability variation expression of the countermeasure. The assignment target factor specifying process definition 280 defines the assignment target factor specifying process, the coping method probability variation expression definition 281 defines the function function of the specified coping method decision probability, and assigns it to the allocation information definition Stored in the allocation function formula 945 of the table 904. This completes the main function of the assignment information definition 208.

  Next, the process described in the right half of FIG. 4 will be described. First, one subject input data 910 is read (step 610). The read input record is converted in accordance with the conversion definition described with reference to FIG. 3 (step 620). An allocation process is performed one by one based on the converted records (step 630).

  Step 630 of this allocation processing includes assignment target factor identification processing (step 631), countermeasure probability variation processing (step 632), countermeasure determination processing based on probability variation (step 633), and determination processing contents. This includes allocation execution based on allocation and allocation aggregation / allocation aggregation addition processing (step 634).

  First, in step 631, the allocation target factor is specified based on the factor information of the converted record. In step 632, the probability variation of the countermeasure for the specified allocation target factor is set. In step 633, a coping method / treatment method is randomly determined based on the probability setting. In step 634, allocation totalization / allocation total addition processing is performed based on the result. Finally, the allocation result is stored in the allocation table table 909 (step 640).

  FIG. 5 shows an example showing generation of a combination pattern of stratification factors and countermeasures of the allocation system. In FIG. 5, regarding the pattern generation 270 of all factors / countermeasures described in FIG. 4, the stratified factor combination pattern generation 271, the countermeasure combination pattern generation 272, the stratified factor / countermeasure combination pattern generation 273, Give a more detailed description.

First, the combination pattern generation 271 of stratification factors will be described. The stratification factor 300 is formed from block allocation information 301 and n factors indicated by factor 1 (302) and factor 2 (303). Identification information is defined under the block allocation information 301 and each factor. In the example of the figure, a block element identification 304 is defined under the block allocation information 301, and a factor 1 element identification 305 and a factor 2 element identification are respectively subordinate to the factor 1 (302) and the factor 2 (303). 306 is defined. This information is allocated information definition table 9 shown in FIG.
04 block information 947 and stratified factor information 948 are classified and managed, and detailed identification information is managed by an identification name 952 of the information division table 905. Therefore, a combination can be generated from this table storage information, and a combination pattern table 307 of stratified factors can be generated.

  Next, the countermeasure combination pattern generation 272 will be described. The coping / treatment 308 has a coping / treatment element 309 under its control. The coping / treatment element 309 is formed from n factors. Information on the coping method / treatment method is also managed by the allocation information definition table 904 shown in FIG.

  The combination pattern generation 273 of the stratified factor and the coping method is performed with the stratified factor pattern and the coping / treatment pattern shown above as inputs. The element key 321 of the allocation block, the element key 322 of the factor 1 and the element key 323 of the factor 2 included in the allocation summary table (at the time of pattern generation) 320 are created from the information of the stratified factor pattern. On the other hand, the allocation number 324 of the coping method / treatment method (A agent) and the allocation number 325 of the coping method / treatment method (B agent) are zero at the generation stage, so only the allocation frame column is created.

  The elements of the allocation frame column are present in the coping / treatment element (A agent) 310 and coping / treatment element (B agent) 311 of the coping / treatment element 309. A combination pattern of stratified factors and countermeasures is generated. The allocation calculation process corresponding to the core engine of the present invention will be described in detail below with reference to the drawings.

  6 and 7 show the process for specifying the allocation target factor of the allocation system. In this process, the process until the assignment target factor is specified for each input data assignment request event is described. In the example of FIG. 6, an allocation table being allocated, an allocation tabulation table (before factor identification) 340 and a subject input data for allocation (first item) record 330 are assumed. The following explanation is based on the simulation of this hypothetical data.

  First, data-specific key requirements for the allocation summary table (before factor specification) 340 are extracted from the data included in the assignment subject input data (first case) record 330. Key requirements are: assignment block 332, factor 1 (333), factor 2 (334) item elements and their values, assignment block value (first case) 336, factor 1 value (first case) 337 The value of factor 2 (first case) is 338.

  Based on this key requirement, the allocation target factor specifying process 350 is performed. That is, the allocation element is searched from the allocation block / factor element of the subject input data for allocation, and the allocation target factor is specified. The row of the key that matches the assignment element, assignment element-assignment block 341, assignment element-factor 1 (342), and assignment element-factor 2 (343) in the assignment summary table (before factor identification) 340 is searched. The result corresponds to allocation block (I Block) target data 357, allocation block (factor 1 = 0) target data 358, and allocation block (factor 2 = 1) target data 359 surrounded by an ellipse. For the data group of the data 357, 358, and 359, data having the worst allocation balance is specified.

In the present invention, a function that can freely set the calculation formula of the specific process as a functional formula is provided. Therefore, it is necessary to define the specific process before executing the specific process. As shown in the definition process 351 of the allocation target factor specifying process, an allocation function expression for specifying the allocation target factor is defined with reference to each item name and value included in the allocation tabulation. In the example of the definition process 351 of the assignment target factor specifying process in FIG. 6, the following formula is defined as an allocation function formula when the factor with the largest allocation ratio difference between the coping method and the treatment method is targeted.
Calculation formula for difference detection = Σ (| Treatment method allocation%-(100% ÷ number of treatment groups) |)
Assignable factor = Max (difference detection calculation result)

  An example of using the assignment function expression definition is described in the execution flow 352 of the assignment target factor specifying process in FIG. In order to improve the execution performance, it is better to store allocation balance information and difference detection information in the data before searching for the target factor. For this reason, the calculation formula for difference detection is performed at the time of addition of allocation aggregation / allocation aggregation.

  In the addition totalization / allocation total processing flow 354, a difference detection calculation is performed. First, an allocation key is extracted for each allocation element from the input data event (step 650).

  Next, using the difference detection calculation formula stored in the allocation function formula 945 of the allocation information definition table 904, the calculation result is stored in the allocation count table 907 (step 651). In the execution flow 355 of the allocation target factor specifying process, the allocation target factor is calculated. A target factor is detected using the allocation target factor stored in the allocation function formula 945 of the allocation information definition table 904 (step 652).

The result executed by the definition described above is shown as an execution result 353 of the assignment target factor specifying process. The execution result of the difference detection calculation formula is as follows in the case of “assignment block” − “I Block”.
(47%-(100% / 2)) + (53%-(100% / 2)) = 6%
The execution result of the allocation target factor is that the difference detection calculation result is the allocation block (6%), factor 1 (6
0%) and factor 2 (42%), factor 1 (60%) is the factor to be assigned. A result indicating the identification result of the factor 1 is represented by an allocation tabulation table (after the factor identification) 356.

FIG. 8 and FIG. 9 show the probability fluctuation calculation of the coping method of the allocation system. In the processing of FIG. 8, as shown in the countermeasure assignment probability process 360, a countermeasure / therapy is assigned to the identified assignment target factor. The input requirement of FIG. 6 is used as input data. Accordingly, the probability fluctuation calculation is performed from the allocation tabulation table (after the factor identification) 356 obtained from the result of FIG. This probability variation calculation can also be defined by a functional expression, as in the assignment target factor specifying process. As shown in the definition 361 of the probability variation formula of the coping method, an allocation function formula that determines the allocation probability of the coping method / treatment method is defined. In this example, the allocation function formula is defined by the following formula when the allocation probability of the countermeasure / treatment method for the allocation target factor is increased.
Assignment probability of factor to be assigned = (number of populations% number of treatment groups-(number of populations x minimum treatment amount assignment% 100%)) + (number of populations% number of treatment groups)

  An execution flow using the definition of the allocation function expression is expressed as an execution flow 362 of the probability variation expression of the countermeasure. First, the parameter% is set to 100% (step 655). Next, the allocation probability of the allocation target factor in the allocation tabulation table 907 is calculated using the allocation probability formula of the allocation target factor stored in the allocation function formula 945 of the allocation information definition table 904 (step 656).

  Next, the allocation probability of the allocation target factor is output (step 657). Thus far, the calculation of the allocation probability of one coping method / treatment method for the allocation target factor is completed. Thereafter, a subtraction process is performed on the number of parameters and the number of treatment groups for the next processing.

  First, the parameter% = the parameter% −the allocation probability of the allocation target factor is executed (step 658). Next, the number of treatment groups = the number of treatment groups-1 is executed (step 658). Then, it is determined whether or not the number of treatment groups is 1, and if it is not 1, calculation of the next countermeasure / treatment method is executed, and if it is 1, the process proceeds to step 661 (step 658). In step 661, the parameter% is output as the allocation probability of the target factor.

  FIG. 9 shows the results performed in this example. As shown by the execution result 363 of the probability variation of the coping method, the following results are obtained. The number of treatment groups is 2 because the treatment groups are two agents, agent A and agent B.

For the first time in the target group, the allocation probability = (100% ÷ 2- (100% × 20 ÷ 100%)) + (100% ÷ 2) = 80%
That is, the allocation probability of agent A is changed from 50% to 80%. At this stage, the population% is 100% -80% = 20%, and the number of treatment groups is 2-1 = 1.

Since the number of treatment groups is 1 for the second time of the target group, the parameter is directly used as the allocation probability of the target factor. That is, the allocation probability of agent B is changed from 80% to 20%.
Further, consider a case where the coping method / treatment method is 3 or more. As shown in the execution result 364 when the treatment group is 3 or more, the following results are obtained. The treatment group is A agent / B agent / C agent, and the number of treatment groups is three. Assume A = 20%, B = 30%, and C = 50%.

For the first time in the target group, the allocation probability = (100% ÷ 3- (100% × 20 ÷ 100%)) + (100% ÷ 3) = about 46%
That is, the allocation probability of agent A is changed from 20% to 46%. At this stage, the population% is 100% -46% = 54%, and the number of treatment groups is 3-1 = 2.

In the second time of the target group, the allocation probability = (54% ÷ 2- (54% × 30 ÷ 100%)) + (54% ÷ 2) = about 38%
That is, the allocation probability of agent B is changed from 30% to 38%. At this stage, the population% is 54% -38% = 16%, and the number of treatment groups is 2-1 = 1.

  Since the number of treatment groups is 1 for the third time in the target group, the parameter is used as it is as the allocation probability of the target factor. That is, the allocation probability of agent C is changed from 50% to 16%.

  FIG. 10 and FIG. 11 show the countermeasure determination based on the result of the probability variation calculation of the allocation system. As input data for the description, the input requirement shown in FIG. 6 is used. Therefore, the countermeasure determination is performed from the execution result 363 of the probability variation of the countermeasure obtained from the result of FIG. The execution result 363 of probability fluctuation is expressed as an internal calculation array 372 of allocation probabilities. The process leading to the coping method is divided into a random number range setting process and a coping / treatment method determining process based on the set random number range, which are shown in FIGS. 10 and 11, respectively.

  As shown in FIG. 10, the random number range setting process will be described using a random number range setting flow 370. First, the lower limit position and the upper limit position are cleared to zero (step 670). At this stage, the allocation probability 374 of the countermeasure / therapy is extracted from the allocation probability internal calculation array 372 (step 671).

  Next, upper limit position = (assignment probability−1) + lower limit position is executed (step 672). As the allocation probability unit, the lower limit position and the upper limit position are output to the random number range internal calculation array 377 (step 673). Finally, lower limit position = upper limit position + 1 is performed (step 674). The processing from step 671 is repeated until the data of the internal calculation array 377 is completed. When this process is executed using the data of the internal calculation array 377 in the random number range, the processing is as follows.

  First, at step 670, lower limit position = 0 and upper limit position = 0 are set. Here, step 671 is executed, and the probability record (first case) 375 of the coping method / treatment method and the agent A are read. Step 672 is executed using the read data. In step 672, since the upper limit position = (allocation probability-1) + lower limit position, the allocation probability is 80, and the lower limit position is 0 from the previous calculation result, and (80-1) + 0 = 79.

  Accordingly, the coping / treatment method at this time is agent A, the lower limit position is 0, and the upper limit position is 79, and the output result of step 673 is the random number range record of the coping / treatment method (first case) 381.

  Next, in step 674, lower limit position = upper limit position + 1 is executed, and 79 + 1 = 80 is prepared, and the lower limit position for the second case is prepared. Similarly, the probability record (second case) of the coping method / therapeutic method is read and the upper limit position = (assignment probability-1) + the lower limit position is calculated, and (20-1) + 80 = 99 is obtained. The treatment method is B agent, the lower limit position is 80, and the upper limit position is 99. The output result of the second case is a random number range record (second case) 382 of the countermeasure / treatment method. This completes the random number range setting process.

  As shown in FIG. 11, the coping / treatment method determination process will be described using a coping / treatment determination flow 371. First, a random number is generated using a random number function provided by the computer system (step 680). Next, a remainder system of 100 is taken for this random number result (step 681).

  Finally, a countermeasure / treatment method in which the following relationship is established among the lower limit position 379 of the random number generation of the countermeasure / treatment method of the internal calculation array 377 of the random number range, the calculated remainder, and the upper limit position 380 is determined. (Step 682).

Lower limit position 379 = <Remainder of search calculation = <Upper limit position 380
As shown in the coping method / therapeutic method determination result 383, assuming that the remainder is 65, the range is 0 = <65 = <79, so the coping method / therapeutic method is agent A.

  12 and 13 show the addition process to the summary table for each stratification factor of the allocation system. As input data for the description, the input requirement shown in FIG. 6 is used. Therefore, the addition process will be described from the coping method / treatment method obtained from the result of FIG. 11 and the input data of FIG. In the addition process, the addition to the assignment summary table, which is the summary table for each stratified factor, via the assignment summary table, which is the solution table of the factor key, based on the input data and the results of coping / treatment methods, and Assign and calculate the difference detection value.

  First, the input data is assigned subject input data (first record) record 330, and the coping / treatment method result is the coping / treatment decision result 383. Then, a table corresponding to the allocation tabulation table is an allocation tabulation table 392 at the time of addition accompanying event processing, and a table corresponding to the allocation tabulation table is an allocation tabulation table 396 at the time of addition accompanying event processing. Allocation summary table (I Block) for handling during event processing (I Block) Measures and treatment method excerpt 393, Allocation summary table for event processing (addition 1) Factor and treatment method excerpt 394, Allocation summary table (factor 2 = 1) for handling during event processing Excerpt 395 for coping and treatment methods clearly shows how individual allocation results shown in the allocation summary table are added to the allocation summary table This is shown for illustrative purposes.

  Next, this process will be described in order. As shown in FIG. 12, first, the factor key designation requirement is extracted from the data of the assignment subject input data (first case) record 330. The factor key designation requirements are the allocation block value (first case) 336, the factor 1 value (first case) 337, and the factor 2 value (first case) 338. Using this factor key designation requirement and the agent A of the determination result 383 of the countermeasure / treatment method as a key, the addition position of the allocation totaling table 392 at the time of addition accompanying event processing is specified. Since the allocation occurs for one case, the number of additions is 1 as shown in the allocation addition event 391. The addition / calculation item associated with the first event processing indicated by the identified key requirement is 400. The associated total items are updated using these 400 events as a trigger.

  First, a factor pattern 401 is added. Next, an allocation summary table is added to each factor included in the event of the input data.

As shown in FIG. 13, in this case, the factors are allocation block = I Block, factor 1 = 0, factor 2
= 1 and 1 is added to 408, 413, and 418 because the coping method / therapeutic method is agent A. The total of 408, 413, and 418 is a unit of allocation block = I Block, factor 1 = 0, factor 2 = 1 in the allocation summary table as indicated by 393, 394, and 395, respectively. Next, each corresponding row of the allocation summary table is calculated using the addition event of 408, 413, and 418 as a trigger. Taking 408 as an example, if the order is followed, first, the countermeasure / treatment method 409 is added, and then the allocation ratios 410 and 411 are recalculated based on the addition contents. Finally, a difference detection result indicated by 412 is calculated based on a function formula defined as a recalculation result. This calculation process is similarly performed for each row 413, 418, 414, 415, 416, 417, 419, 420, 421, 422. Finally, the allocation result is added as allocation table additional data 390.

  Through the processes of FIGS. 6 to 13 described above, the allocation calculation process for one data record, in other words, the allocation request 1 event is completed.

  14 and 15 show a second example of the allocation calculation process shown in FIGS. 6 to 13. As the input data, the input requirement shown in FIG. 6 is used. Therefore, description will be made using the results of the allocation tabulation in FIG. 12 and the allocation tabulation in FIG.

As shown in FIG. 14, as input data, a new assignment subject input data (second record) record 450 is assumed as second data. Since the factor elements of this record are allocation block = I Block, factor 1 = 0, factor 2 = 2, the allocation target is the allocation block (I Block) target data 458, allocation block (factor 1 = 0) target data 459, and allocation block (factor 2 = 2) target data 460.

  Next, the allocation target factor specifying process is executed. As shown in the execution result (second case) 461 of the assignment target factor specifying process, an assignment function expression similar to that for the first case is used. Therefore, similarly, since the allocation block (0%), the factor 1 (54%), and the factor 2 (72%), the factor 2 (72%) is specified as the target. The result is an allocation summary table (after specifying the second factor) 462.

  Next, the probability variation of the countermeasure is executed. As shown in the execution result (second case) 463 of the probability variation of the coping method, the same allocation function formula as the first case is used. The number of treatment groups is 2 because the treatment groups are two agents, agent A and agent B.

For the first time in the target group, the allocation probability = (100% ÷ 2− (100% × 14 ÷ 100%)) + (100% ÷ 2) = 86%
That is, the allocation probability of agent A is changed from 50% to 86%. At this stage, the population% is 100% -86% = 14%, and the number of treatment groups is 2-1 = 1.

  Since the number of treatment groups is 1 for the second time of the target group, the parameter is directly used as the allocation probability of the target factor. That is, the allocation probability of agent B is changed from 86% to 14%. Further, assuming that the coping method / treatment method calculated based on the probability of changing the coping method based on the probability variation and the execution result (second case) 464 is B agent, the coping method / treatment method Can obtain the result of B agent.

As shown in FIG. 15, the item 470 is identified from the assignment result of the B agent and the factor element of the assignment subject input data (second case) record 450, assignment block = I Block, factor 1 = 0, factor 2 = 2. The Using the update of this item 470 as a trigger, as shown in the allocation totalization / allocation total addition process (second case) 465, as in the first case, addition to the allocation totalization / allocation total, 471 to 477 are executed one after another. Then, the allocation table additional data (second item) 456 is performed.

  FIGS. 16 and 17 show the access process to the summary table database for each stratification factor of the allocation system. As the input data, the input requirement shown in FIG. 6 is used. Here, a description will be given focusing on the processing module that directly takes charge of database access based on the events that occur accompanying the calculation of the allocation summary table and the assignment summary table.

  As described above, according to the allocation request event of the input data, first, key resolution is performed in the allocation tabulation table, and addition processing of the allocation tabulation table is executed based on the result. Therefore, FIG. 16 shows an access summary table and database access approach, and FIG. 17 shows an access summary table database access approach. The access approach is described below step by step.

  First, referring to FIG. When the combination of the factor pattern of the allocation tabulation table and the countermeasure / treatment method is solved, addition 500 to the allocation tabulation table is executed. Update the corresponding record in the table with the added allocation table. In order to directly access the corresponding record of the allocation totaling table 906, the allocation totaling table 906 needs to have a factor pattern and a data specifying key for a countermeasure / treatment method. In this example, an assignment generation key 513 for specifying a factor pattern and a treatment / handling method ID 514 for specifying a countermeasure / treatment are held as data specifying keys.

  Further, as shown in the assignment generation key rule 505, the key area of the assignment generation key 513 is a key for specifying one factor, “information classification (stratified factor information)” — “detailed classification (factor classification)” — “key” Since a structure that concatenates and holds “ID (block key ID)” is adopted, it is not affected by the excess or deficiency of the number of factors.

  This will be described with reference to FIG. Next, the allocation summary table is updated based on the addition event of the allocation summary table. This function uses the allocation key retrieval 501 first. This processing module disassembles the keys of the assignment generation key 513 (FIG. 16) by factor and takes out the keys. In the illustrated allocation key extraction 501, “allocation block” − “I Block” is extracted as an example. The assignment element key 519 solves the assignment elements displayed in the column of the assignment summary table, but the countermeasure / treatment method corresponding to the row and the summary column cannot be specified. As this specific requirement, according to the column requirement in the value column 526 of the treatment / remedy ID, that is, depending on the requirement of access to the treatment / treatment, or access to the aggregation, By storing the treatment method ID or the total ID, it is possible to directly access the corresponding record in the allocation total table 907. An example of the access is indicated by addition 502 to the corresponding element / total element. Next, based on the addition result, it is necessary to execute the allocation (%) calculation 503 and the difference detection calculation 504 of the corresponding element / aggregated element. The difference in these requirements is the difference in the attribute for the field of whether the addition number is held in the cumulative or the ratio, and the specific unit of the key is the same. Therefore, the data field indicated by the previous key is held in the cumulative requirement, the ratio requirement, and the requirement unit, thereby enabling data storage and access. The data fields correspond to the stratified factor countermeasures total 521, the stratified factor countermeasures total 522, the stratified factor countermeasure ratio 523, and the stratified factor countermeasure difference ratio 524, respectively.

FIG. 18 shows the execution control subfunction of the allocation system. Here, among the functions of the allocation execution definition 202, the functional configuration of the allocation pattern designation 209 and the output destination pattern designation 210 will be mainly described.

  One feature of the assignment system of the present invention is that it has a plurality of input information resources related to input data and assignment definition information and output information represented by an assignment table, and can be dynamically linked. In realizing this function, input information is managed in the form of an allocation pattern, and output information is managed in the form of an output pattern. The information managed by the allocation pattern includes information regarding input data to be allocated, subject input data 910, information regarding allocation definition, allocation information definition table 904, allocation item management table 903, information regarding aggregation by stratified factor, allocation aggregation table Reference numeral 906 denotes an allocation tabulation table 907. Access to these information resources is performed via the allocation information definition 201.

  On the other hand, the information managed by the output pattern is composed of information related to the allocation result and the allocation table table 909 alone. When performing allocation execution, the allocation execution definition 202 executes each of the allocation pattern specification 209 and the output destination pattern specification 210, specifies the allocation pattern to be used and the output destination pattern, and solves the execution resource. Resource allocation is possible. In the example shown in the figure, two allocation patterns A (565) and B (566) are created in this allocation pattern, and output pattern A (567), output pattern B (568), and output pattern C are output patterns. Three of (569) are created.

  The other two functions specified in the function of the allocation execution definition 202, the allocation status specification 211, and the allocation calculation skip specification 212 are optional functions for extracting data by specifying the allocation status when the former inputs data, The latter is an optional function that controls the implementation after the addition process to the allocation summary table and the allocation summary table, skipping the allocation determination process of the allocation process.

  FIG. 19 is a flowchart showing the execution control subfunction of the allocation system. Here, the description will focus on branch control according to the designation requirements of the allocation pattern designation 209, output destination pattern designation 210, allocation status designation 211, and allocation calculation skip designation 212 described in the function of the allocation execution definition 202 shown in FIG. Note that the part designated or described as designation is a description based on the designation requirements 209 to 212. The process flow is described below in order.

First, the allocation information definition 904 is read in accordance with the specified allocation pattern Seq and allocation pattern specification 209 (step 700). Similarly, the allocation item management 903 is also read in accordance with the specified allocation pattern Seq and allocation pattern specification 209 (step 710). Up to this point, the definition resource has been read, and the input data file must be specified next. Since the specification of the input data file is managed by the input conversion definition of the assignment information definition, the input data file is determined from the read assignment information definition 904 and assignment item management 903 (step 720).

Next, the specified allocation state, that is, the allocation status designation 211 and the allocation status of the allocation table are compared with the record of the confirmed input data file, and the records are read if they match (step 730). The read input data is converted according to the input conversion definition (step 740). As preparation for the allocation process, an allocation tabulation / allocation tabulation table is specified based on the specified allocation pattern Seq (step 750). Before executing the allocation calculation of the input record shown in the next step 770, the allocation calculation processing is controlled according to the allocation calculation skip designation 212. If the allocation calculation skip specification 212 is specified, that is, if the allocation calculation skip specification 212 is YES, the processing of step 771, step 772, and step 773 is skipped, the allocation table factor information 580, the allocation table allocation result ( (Treatment / Treatment Method) Based on the information of 581, the allocation totalization and the allocation total addition processing (step 774) are directly executed (step 760).

  In the case of NO, all processes included in step 770 are executed in order. Execution of allocation target factor specifying processing (step 771), coping method probability variation processing (step 772), coping method determination processing based on probability variation (step 773), allocation aggregation and allocation aggregation Addition processing (step 774) is executed. Finally, the allocation result is written in the allocation table table 909 according to the specified output destination pattern Seq and output destination pattern specification 210 (step 780). The above steps 730 to 780 are repeated until the record of the subject input data 910 is read.

  FIG. 20 shows the allocation status function of the allocation system. Here, the operation in the operation example using the function of the allocation state designation 211 shown in FIGS. 18 and 19 will be described. As one of the objects of the present invention, it is described that measures are taken to enable operation without involving the allocation request source data and without being aware of the allocation. This requirement includes requirements that can be used without being aware of the assigned state on the input data side in addition to input data selection and format conversion. The example shown in the figure explains whether or not this requirement is specifically satisfied.

  The configuration of the figure will be described below. At the top of the figure is the subject input data 910 that has already been assigned some input data records. Then, the input requirement before the allocation is located on the left side with the allocation execution 203 in the center, and the output result after the allocation is located on the right side. At the bottom is an example of data introduction using input data and output results.

  To explain in order, first, in the input requirement on the left side, the assignment state designation 550 is unassigned, and an instruction to set only unassigned subject input data 910 as an assignment target is set. Next, looking at the allocation table record (before allocation) 552 located at the bottom, there are two allocation results, S001 and S002 already exist in the subject ID 222, S001 has been allocated, and S002 has not been allocated. It has become. Here, when turning to the subject input data 910 again, S001, S002, and S003 exist in the subject ID 222. Since S003 does not exist in the allocation result of the allocation table record (before allocation) 552, it is necessary to use the allocation target data. For the remaining S001 and S002, only S002 that matches the unallocated assignment state designation 550 is targeted. Therefore, it is necessary to input S002 and S003 as assignment targets and assign them.

  It will be checked whether the process for obtaining this result is processed in the flow shown in the allocation execution 203. First, one input record is read (step 790). The first S001 is read.

  Next, the allocation status designation 550 and the allocation status of the allocation table table 909 are compared, and if they match, or if there is no record in the allocation table table 909, a record is read (step 791). The first case does not satisfy this condition, so it is inconsistent.

  The second input record is read (step 790). The second S002 is read. In the second case, S002 indicated by the subject ID 222 of the assignment state designation 550 and the subject ID 222 of the assignment table record (prior to assignment) 552 match, so that it becomes an assignment target, the assignment calculation of the input record in step 792, and the assignment in step 793 The allocation status of the table / table is updated. Similarly, when step 791 is executed for S003 which is the third input data, since S003 indicated by the subject ID 222 does not exist in the allocation table record (before allocation) 552, it becomes the allocation target, and similarly, steps 792 and 793 are performed. Is executed. As a result, S002 and S003 become assignment targets, and the result of the assignment table record (after assignment) 553 can be obtained.

  As described above, the assignment can be performed without intervention of the input data of the assignment request source and without being aware of the assignment. Finally, subject input data 910 and assignment table table 909 are key-joined by subject ID 222 as shown in the dotted line portion at the bottom of the figure, subject input / assignment result combination record 555, thereby displaying subject input data display 1000. It becomes possible to display the data before and after allocation together with the input data.

  21, FIG. 22 and FIG. 23 show the simulation function of the allocation system. Here, simulation functions that can be realized using the functions of the allocation pattern designation 209, the output destination pattern designation 210, and the allocation calculation skip designation 212 shown in FIGS. 18 and 19 will be described. In the example shown in the figure, in the process of data collection where there is already input data that has already been assigned, after changing the assignment information accompanying the change of the factor pattern, the data collection is resumed and the assignment is performed using the new factor pattern. The simulation function will be described through an operation example. In this example, this flow is roughly divided into three blocks as shown in FIGS. 21, 22, and 23.

  The block before stratification factor correction (for actual operation) 560 in FIG. 21 shows the operation before the factor pattern change. The block 561 after stratification factor correction (for regeneration) 561 in FIG. 22 creates a stratification factor-specific aggregation table (allocation aggregation table and allocation aggregation table) from already allocated data after the factor pattern change. It is an example. In the allocation method considering the stratified factor balance implemented in the present invention, it is necessary to add the allocation results so far to the tabulated table for each stratified factor before accepting a new allocation request. Therefore, before allocating new input data after changing the assignment, this table, that is, the tabulation table for each stratified factor generated by changing the factor pattern based on the already assigned data (assignment aggregation) Table and allocation summary table). The block 562 after modification of the allocation target stratification factor (for actual operation) in FIG. 23 performs allocation to new input data using the changed factor pattern.

  As shown in FIG. 21, before stratification factor correction for assignment (for actual operation) 560, subject input data 563 is input, assignment pattern A (565) is designated in assignment pattern designation 209, and output destination pattern designation is performed. The output pattern A (567) is designated in 210, and the assignment is performed. The result is output to the allocation table data 570 for actual operation set in the output pattern A (567).

  Next, as shown in FIG. 22, after the stratified factor correction (for regeneration) 561 to be allocated, it is necessary to newly set a factor pattern, so the allocation information definition 208 is assigned to a new allocation pattern B (566). Set using. In the illustrated example, regeneration is defined by factor 1 and factor 2. Then, input data to be added for the summary table for each stratified factor is prepared. The input data is the allocation table data 570 for actual operation, which is the allocation result of the allocation target stratified factor correction (for actual operation) 562, and the input data after the allocation target stratification factor correction (for actual operation) 562 is performed. A certain subject input data 563 is combined to prepare combined data 571 of the subject input data and the allocation table data. Next, specify the output file. The output purpose is not the actual allocation but the addition result up to that time is added to the tabulation table for each stratification factor, so the output pattern B (568) which is the test output destination is designated. In this state, input / output resource settings are complete. Then, the allocation calculation skip designation 212 is designated and the allocation is executed. The allocation calculation skip described with reference to FIG. 19 works, and the allocation execution 203 based on the factor information included in the combined data 571 of the subject input data and the allocation table data and the allocation result add the allocation aggregation / allocation aggregation as shown in the figure. Only executed.

With the above procedure, the allocation tabulation / allocation tabulation after changing the factor pattern is created from the allocation result, and the preparation for accepting allocation of new input data is completed. In the usage form given here, the result is output to a file other than the actual allocation, so that it is possible to divert the simulation to another factor pattern.

  As shown in FIG. 23, when the preparation for receiving the assignment of new input data is completed, the stratified factor correction (for actual operation) 562 to be assigned is executed. Here, since it is necessary to output the allocation result data subsequent to the allocation result of actual operation, the output destination pattern designation 210 is returned to the output pattern A (567) and executed. As shown in the example of the figure, for the new subject input data 564, an allocation table record 575 after correction of the stratification factor is added to the allocation table record 574 before correction of the stratification factor in the allocation table table 909. Is output. The allocation pattern field 573 of records 574 and 575 stores an allocation pattern, allocation pattern A (565), and allocation pattern B (566) when each allocation is performed.

The allocation system according to the present invention has the following advantages.
Each time an allocation request is made, the accumulated stratification factor information is allocated as a population, and the result is a circulation model configuration that uses the feedback and uses it, so it is always available to the population without human intervention. Dynamic allocation with optimal factor variation can be performed.

  In addition, a structure that allows the user to define the calculation formulas related to the factor content of the stratified factors, the number of factors, the structure that does not restrict the coping and treatment methods, the identification of the target factors, and the allocation ratio as functional expressions Can be used for stratified allocation of various research and research themes.

  In addition, in determining the allocation, instead of determining the most unbalanced factors and allocation method, we adopt a preferential probability variation technique, so the randomness required in the field of research and research (randomness) Can be maintained until the end.

  In addition, since a configuration is used in which individual values are block mapped to factor types, it is easy to set stratified factors for spatial elements including diffusion values, temporal elements including continuous values, and variable factor elements. Can be done.

  In addition, since the stratified factors, the method of generating countermeasures and treatment methods by automatic generation from the master, and the configuration of the loop algorithm function for setting the function formula, it was done by conventional program creation Frame requirements are reduced, and high-quality allocation processing can be created in a short time and at a low cost.

  In addition, because it adopts a configuration that allows input data to be fetched without being processed or updated and the assignment status can be managed, data can be easily exchanged with other systems and data registration / collection systems related to research and surveys.

  Further, since a configuration of a plurality of input information patterns, a plurality of output information patterns, and an input / output pattern dynamic combination function is employed, an allocation simulation is possible.

  Furthermore, since the configuration of the regeneration function for the tabulated by stratified factor from the output result is adopted, the allocation pattern after the allocation can be dynamically changed.

  And since the input information pattern used as the unit for determining the method of dealing with and assigning assignment results adopts a link configuration, it is possible to identify the input data after the fact and the definition information related to factor assignment, and the audit trail. Is possible.

  In addition, the allocation core engine employs a configuration that prevents performance degradation associated with large amounts of data, making it possible to handle large-scale tests and surveys.

It is explanatory drawing which shows the whole function structure of the allocation system concerning one Embodiment of this invention. It is a database model figure of the allocation system concerning one Embodiment of this invention. It is explanatory drawing which shows the input conversion function of the allocation system concerning one Embodiment of this invention. It is explanatory drawing which shows the allocation function according to the dynamic layer of the allocation system concerning one Embodiment of this invention. It is explanatory drawing of the combination pattern production | generation of the stratification factor and countermeasure of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the specific process of the allocation object factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the flow of the specific process of the allocation object factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the probability fluctuation | variation calculation of the countermeasure of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the result of the probability fluctuation | variation calculation of the countermeasure of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the countermeasure determination based on the result of the probability fluctuation | variation calculation of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the flow of the countermeasure determination based on the result of the probability fluctuation | variation calculation of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the 1st example of the addition process to the total table according to the classification | category factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the 2nd example of the addition process to the total table according to the classification | category factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the Example of the allocation process of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the Example of the allocation process of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the access process to the allocation totalization table database according to the classification | category factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the access process to the allocation totalization table database according to the stratification factor of the allocation system concerning one Embodiment of this invention. It is explanatory drawing which shows the execution control subfunction of the allocation system concerning one Embodiment of this invention. It is a flowchart which shows the execution control subfunction of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the allocation state function of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the simulation function of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the simulation function of the allocation system concerning one Embodiment of this invention. It is explanatory drawing regarding the simulation function of the allocation system concerning one Embodiment of this invention.

Explanation of symbols

100 Treatment Group / Countermeasure Group Table 101 Block Table 102 Stratified Factor Table 200 Allocation Processing 201 Allocation Information Definition 202 Allocation Execution Definition 203 Allocation Execution 204 Definition of Treatment / Corporation Information 205 Definition of Block / Strategic Factor Information 206 Pattern Information Definition 207 Subject input data conversion definition 208 Assignment information definition 209 Assignment pattern designation 210 Output destination pattern designation 211 Assignment status designation 212 Assignment calculation skip designation 213 Subject input data conversion processing 214 Assignment calculation of input record 215 Addition / update of assignment table data 216 Input file selection processing 217 Block solution processing 218 Stratified factor combination processing 219 Factor determination function processing 220 Input file A
221 Input file B
222 Subject ID
223 facilities or time zone 224 element 1
225 Element 2
226 Element 3
227 Element 4
228 Allocation block 229 Factor 1
230 Factor 2
231 Factor 3
232 A facility 233 B facility 234 C facility 235 D facility 236 I Block
237 II Block
238 Factor Determination Command Statement 239 Assignment Subject Input Data 240 Subject Input Data (S001) Record 241 Subject Input Data (S002) Record 242 Assignment Subject Input Data (S001) Record 243 Assignment Subject Input Data (S002) Record 270 All Generation of factor / countermeasure pattern 271 Generation of combination pattern of stratified factor 272 Generation of combined pattern of remedy 273 Generation of combined pattern of stratified factor and coping 274 Generation of counter total table for stratified factor 275 Countermeasure of stratified factor Aggregation by law 276 Aggregation of countermeasures by stratified factor 277 Ratio of countermeasures by stratified factor 278 Ratio of countermeasure differences by stratified factor 279 Definition of probability formula using stratified factor items and countermeasure aggregation 280 Definition of specific processing 281 Probability of countermeasure Definition of the dynamic type 300 layers by factor 301 block assignment information 302 Factor 1
303 Factor 2
304 Block element identification 305 Factor 1 element identification 306 Factor 2 element identification 307 Combination pattern table of stratified factors 308 Countermeasure / treatment 309 Countermeasure / therapy element 310 Countermeasure / therapy element (agent A)
311 Elements of coping and treatment (agent B)
320 Allocation summary table (at the time of pattern generation)
321 Element key of allocation block 322 Factor key of factor 1 323 Element key of factor 2 324 Countermeasure / treatment method (agent A) allocation number 325 Countermeasure / treatment method (agent B) allocation number 326 Countermeasure / therapy method allocation Number 330 Subject input data for allocation (first item) record 331 Subject ID
332 Allocation block 333 Factor 1
334 Factor 2
335 Subject ID value (first case)
336 Value of the allocation block (first case)
337 Value of factor 1 (first case)
338 Factor 2 value (first case)
340 Allocation summary table (before factor identification)
341 Allocation element-Allocation block 342 Allocation element-Factor 1
343 Assigned Factor-Factor 2
344 Countermeasures / Treatment (Agent A) by element allocation 345 Countermeasures / Treatment (agent B) by element allocation 346 Countermeasures / therapy / elemental allocation by total 347 Countermeasures / therapy (A) Allocation rate by element of drug) 348 Allocation rate by element of coping method / treatment method (B agent) 349 Allocation difference rate of coping method / treatment method (B agent) 350 Identification process of allocation target factor 351 Identification process of allocation target factor Definition flow 352 Execution flow of specific process for allocation target factor 353 Execution flow of specific process for allocation target factor 354 Addition processing flow of allocation totalization / allocation total 355 Execution flow of specific processing for allocation target factor 356 Allocation totalization table (after factor identification) )
357 Allocation block (I Block) target data 358 Allocation block (factor 1 = 0) target data 359 Allocation block (factor 2 = 1) target data 360 Countermeasure allocation probability processing 361 Countermeasure probability variation formula definition 362 Countermeasure Execution flow of probability variation formula 363 Execution result of probability variation of coping method 364 Execution result when treatment group is 3 or more 370 Random number range setting flow 371 Coping method / treatment method determination flow 372 Internal calculation array of allocation probability 373 Countermeasure / Treatment Drugs 374 Countermeasure / Treatment Allotment Probability 375 Countermeasure / Treatment Probability Record (First Case)
376 Probability record of coping / treatment (second)
377 Internal calculation array of random number range 378 Countermeasure / treatment drug 379 Lower limit of random number generation of countermeasure / therapy 380 Upper limit position of random number generation of countermeasure / treatment 381 Random number range record of countermeasure / treatment ( 1st)
382 Random-range record for coping / treatment (second)
383 Responses / Treatment Decision Results 390 Additional data for allocation table 391 Allocation addition event 392 Allocation summary table for additions associated with event processing 393 Allocation summary table for additions associated with event processing (I Block) Excerpt 394 Excerpt of the allocation summary table (factor 1 = 0) for coping / event methods associated with event processing 395 Allocation summary table (factor 2 = 1) for coping / treatment methods relating to event processing Excerpt of 396 Allocation summary table during addition associated with event processing 400 to 430 Addition / calculation items associated with event processing 450 Subject input data for allocation (second record) 451 Subject ID value (second record)
452 Value of assigned block (second case)
453 Factor 1 value (second case)
454 Factor 2 value (second case)
455 Result of coping / treatment (second)
456 Additional allocation data (second)
457 Allocation Summary Table (when the first item is completed)
458 Allocation block (I Block) target data 459 Allocation block (factor 1 = 0) target data 460 Allocation block (factor 2 = 2) target data 461 Execution result of allocation target factor specifying process (second case)
462 Allocation summary table (after specifying the second factor)
463 Execution result of probability variation of countermeasure (second case)
464 Decision process flow and execution result based on probability variation (second case)
465 Addition and allocation summation processing (second case)
466 Allocation summary table for event processing (2nd)
467 Allotment summary table at the time of event processing (second)
470-490 Items to be added / calculated with event processing (second)
500 Addition to allocation summary table 501 Extraction of allocation key 502 Addition to corresponding element / aggregation element 503 Calculation of allocation (%) of corresponding element / aggregation element 504 Difference detection calculation 505 Rules for allocation generation key 506 Allocation aggregation table access data 507 Allocation summary table access A agent data (at the time of aggregation)
508 Allocation summary table access B agent data (at the time of aggregation)
509 Assignment summary table access summary data (at the time of summary)
510 Allocation summary table access agent A data (when calculating allocation rate)
511 Allocation summary table access B agent data (at the time of allocation rate calculation)
512 Assignment summary table access summary data (at the time of assignment rate calculation)
513 Assignment generation key 514 Treatment / handling ID
515 Number of assigned tests
516 Value field of assignment generation key 517 Value field of treatment method / handling method ID 518 Value field of number of assigned tests 519 Assigned element key
520 Treatment / Coping ID
521 Totaling by stratification factors
522 Countermeasures for stratified factors
523 Ratio of countermeasures for stratified factors
524 Countermeasure difference ratio of stratified factors
525 Value field of assigned element key 526 Value field of treatment / handling ID
527 Value column for totaling by stratification factor
528 Value column for countermeasures aggregation for stratified factors 529 Value column for countermeasure ratios for stratified factors
530 Value field of difference method for dealing with stratified factors 550 Specification of allocation status 551 Subject input record 552 Allocation table record (before allocation)
553 Allocation table record (after allocation)
554 Additional allocation table record 555 Combined record of subject input / allocation result 560 Before correction of stratified factor to be allocated (for production)
561 After modification of stratified factors to be allocated (for regeneration)
562 After stratification factor correction of allocation target (for production use)
563 Subject input data 564 New subject input data 565 Allocation pattern A
566 Allocation pattern B
567 Output pattern A
568 Output pattern B
569 Output pattern C
570 Allocation table data for actual operation 571 Combined data of subject input data and allocation table data 572 Allocation table data for testing 573 Allocation pattern column 574 Allocation table record before correction of stratification factor 575 Assignment after correction of stratification factor Table record 580 Allocation table factor information 581 Allocation table allocation results (coping / treatment)
900 Assignment information definition table group 901 Assignment execution definition table group 902 Assignment execution table group 903 Assignment item management table 904 Assignment information definition table 905 Information division table
906 Allocation total table 907 Allocation total table 908 Allocation management table 909 Allocation table table 910 Subject input data 911 Allocated subject input data 912 Subject input data group 930 Allocation pattern Seq
931 Output destination pattern Seq
932 Information classification 933 Output table name 934 Output display order 935 Output item name 936 Display attribute classification 937 Protection attribute classification 937 Assignment randomization usage classification 939 Assignment randomization designation order 940 Input order 941 Input item name 942 Factor function designation 943 Information Seq
944 Input file name 945 Assignment function formula 946 Treatment / handling information 947 Block information 948 Stratification factor information 949 Allocation status information 950 Detailed classification 951 Key ID
952 Distinguished name 953 Assignment summary table Seq
954 Allocation generation key 955 Treatment / coping ID
956 Number of assigned tests 957 Assignment summary table Seq
958 Assigned element key 959 Treatment / handling ID
960 Aggregation by stratification factor by coping method 961 Aggregation by coping method by stratification factor 962 Coping ratio by stratification factor 963 Coping difference ratio by stratification factor 964 Management line number 965 Allocation table Seq
966 Subject ID
967 Block ID
968 Treatment / Countermeasure ID
969 Allocation status ID
1000 Subject input data display

Claims (8)

  In a coping method assignment method for randomly allocating coping methods to test subjects, an input step for inputting input data relating to a plurality of test subjects, and an input data conversion step for transforming the input data into allocation data having factor information , Comparing the factor information of the test object to which the countermeasure is assigned with the factor information of the test object to which the countermeasure is already assigned, and detecting the test object in which both match, and the test object to which the above countermeasure is assigned Among the detected test objects, an unbalance detection step for detecting an unbalanced test object with an unbalanced allocation rate of the countermeasure, and a test object with a low allocation ratio of the countermeasure among the unbalanced test objects. On the other hand, the allocation probability of the above countermeasure is set high, and the above countermeasure is applied to a test object having a high allocation ratio of the above countermeasure among the above unbalanced test objects. Allocation method of Action including the allocation probability adjustment step of setting a low assignment probabilities, the.   2. The method of assigning a countermeasure according to claim 1, wherein factor information of a test object to which the countermeasure is assigned and factor information of a test object other than the test object are set in the assignment probability setting process of the countermeasure in the assignment probability adjusting step. A method of assigning countermeasures characterized by reflecting.   The coping method allocation method according to claim 1, wherein the result of the allocation probability adjustment step is restored by obtaining the coping method determination result from information held in the output data of the coping method allocation result. How to assign the corrective action.   2. The allocation method of the coping method according to claim 1, wherein in the allocation probability adjusting step, the allocation probability of the coping method is obtained by a function expression.   2. The coping method allocation method according to claim 1, wherein the detection step, the imbalance detection step, and the allocation probability adjustment step are executed for each test object included in the input data.   A computer-readable program for causing a computer to execute the countermeasure allocation method according to claim 1.   In the input data conversion unit that converts input data to allocation data and the unit of allocation request for collected data, factor patterns are allocated from the factor information included in the allocation data, and the allocation ratio for each factor layer that is currently allocated Core engine function unit that performs allocation by adjusting, simulation unit that performs simulation of allocation results, and information necessary for calculating random occurrence probability is stored in units of stratified factor pattern and countermeasure pattern, and data in the stored unit Means for reading a block for each action determination request event, and a countermeasure allocation system.   8. The coping method assignment system according to claim 7, wherein the coping method assignment result is linked to an input information pattern, thereby enabling the assignment method to be specified from the coping method assignment result. The allocation system of the coping method to do.

Images