A product selection test method and apparatus

Abstract

The application provides a product selection test method and device, which constructs an iteration cycle index set by using a non-repeated to-be-screened index set in a corresponding available index set and a historical index set of a to-be-tested product, screens the to-be-screened index in the iteration cycle index set according to a J-wheel index screening iteration operation, and finally obtains a recommended index set constructed by the available index set and the screened index. The automatic screening process not only saves a large amount of human cost and time resources, but also adopts an iteration screening matching index mode, so that the selected index is more accurate, and the accuracy of product selection is further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a product selection test method and device. BACKGROUND

[0002] When a product is first introduced or a product model or version is greatly changed, the product needs to be selected and tested first to evaluate its applicability in the application scenario of the enterprise, and then is listed in the information technology product catalog or enters the centralized procurement procedure.

[0003] At present, various enterprises rely on the experience of selection test engineers and combine a large amount of measured product data to perform product selection testing, which is not only time-consuming and laborious, but also the selected products are not ideal. SUMMARY

[0004] Therefore, the embodiments of the present application provide a product selection test method and device to solve the problem that the existing product selection is time-consuming and laborious and the selected products are not ideal.

[0005] To solve the above problems, the embodiments of the present application provide the following technical solutions:

[0006] The first aspect of the embodiments of the present application discloses a product selection test method, which comprises:

[0007] Obtaining a full-amount index set and an available index set generated based on product information of a product to be tested, the full-amount index set comprising a product set and a historical index set, each product in the product set and each historical index in the historical index set being in a many-to-many relationship;

[0008] Selecting non-repeated screening indexes in the available index set and the historical index set to form a set, obtaining an iteration cycle index set, and calculating a recommended score of each screening index in the iteration cycle index set;

[0009] Performing a J-round index screening iteration operation, selecting a required available index from the iteration cycle index set to add to the available index set, and the initial value of J is 1;

[0010] Obtaining the available index set after the J-round index screening iteration operation as a recommended index set of the product to be tested.

[0011] Optionally, the J-round index screening iteration operation comprises:

[0012] Sorting the screening indexes currently existing in the iteration cycle index set according to their respective recommended scores from high to low;

[0013] Selecting the top n screening indexes, determining whether there is an available index for the current product selection test in the n screening indexes, and n is a positive integer greater than 1.

[0014] If there are h available indexes, add the h available indexes as the product selection test of this time to the available index set, h < n;

[0015] Delete the n to-be-screened indexes from the iteration cycle index set, and determine whether the to-be-screened indexes currently existing in the iteration cycle index set meet the iteration exit condition;

[0016] If the iteration exit condition is not met, J is incremented by 1 to proceed to the next round of index screening iteration operation;

[0017] If there are no available indexes, determine whether the to-be-screened indexes currently existing in the iteration cycle index set meet the iteration exit condition;

[0018] If the iteration exit condition is not met, delete the n to-be-screened indexes from the iteration cycle index set, increment J by 1, and return to the step of selecting the top n indexes.

[0019] Optionally, the method further comprises:

[0020] If the number of to-be-screened indexes currently existing in the iteration cycle index set is less than n, exit the index screening iteration operation.

[0021] Optionally, the determination of whether the to-be-screened indexes currently existing in the iteration cycle index set meet the iteration exit condition comprises:

[0022] Obtain the total number Count1 of to-be-screened indexes currently existing in the iteration cycle index set;

[0023] Obtain the product Count3 of the total number Count2 of indexes in the union of the available index set and the historical index set and a set parameter δ, the set parameter δ being a positive number less than 1;

[0024] Determine whether the total number Count1 of indexes is less than the product Count3;

[0025] If yes, the iteration exit condition is met;

[0026] If no, the iteration exit condition is not met.

[0027] Optionally, the calculation of the recommendation score of each to-be-screened index in the iteration cycle index set comprises:

[0028] For each to-be-screened index i in the iteration cycle index set, calculate the corresponding index association score scoreX i and the product association score scoreY i .

[0029] obtain an index correlation weight a of each to-be-screened index i, the index correlation weight a is proportional to the index correlation weight;

[0030] for each to-be-screened index i, according to the respective corresponding index correlation weight a, the index correlation score scoreX i and the product correlation score scoreY i perform calculation to obtain the recommended score score i of the to-be-screened index i corresponding to the product.

[0031] The second aspect of the embodiment of the application discloses a product selection test device, the device comprises:

[0032] an acquisition unit, configured to acquire a full-amount index set and an available index set generated based on product information of a to-be-tested product, the full-amount index set comprising a product set and a historical index set, each product in the product set and each historical index in the historical index set being in a many-to-many relationship;

[0033] a selection unit, configured to select to-be-screened indexes that are not repeated in the available index set and the historical index set to form a set, obtain an iteration cycle index set, and calculate a recommended score of each to-be-screened index in the iteration cycle index set;

[0034] an iteration screening unit, configured to perform J rounds of index screening iteration operations, select available indexes meeting requirements from the iteration cycle index set to add to the available index set, and set an initial value of J to 1;

[0035] a recommendation unit, configured to obtain the available index set after the J rounds of index screening iteration operations as a recommended index set of the to-be-tested product.

[0036] Optionally, the iteration screening unit performing the J rounds of index screening iteration operations comprises:

[0037] a sorting module, configured to sort to-be-screened indexes currently existing in the iteration cycle index set in descending order according to respective recommended scores;

[0038] a first judgment module, configured to select n to-be-screened indexes with the highest rankings, judge whether there are available indexes for the current product selection test in the n to-be-screened indexes, if there are h available indexes, execute an adding module, and if there are no available indexes, execute a second iteration module; wherein n is a positive integer greater than 1, and h < n;

[0039] the adding module, configured to add the h available indexes for the current product selection test to the available index set;

[0040] The first iteration module is configured to delete the n indicators to be screened from the iteration cycle indicator set, and execute the second judgment module. If the result of the second judgment module is that the exit iteration condition is not met, J is increased by 1, and the sorting module is executed.

[0041] The second iteration module is configured to execute the second judgment module. If the result of the second judgment module is that the exit iteration condition is not met, the n indicators to be screened are deleted from the iteration cycle indicator set, J is increased by 1, and the first iteration module is executed.

[0042] The second judgment module is configured to judge whether the indicators to be screened currently existing in the iteration cycle indicator set meet the exit iteration condition.

[0043] Optionally, the method further comprises:

[0044] The exit module is configured to exit the indicator screening iteration operation if the number of the indicators to be screened currently existing in the iteration cycle indicator set is less than n.

[0045] Optionally, the second judgment module is configured to:

[0046] Obtain the total number Count1 of the indicators to be screened currently existing in the iteration cycle indicator set, and obtain the product Count3 of the total number Count2 of the indicators in the union of the available indicator set and the historical indicator set and a set parameter δ, and judge whether the total number Count1 of the indicators is less than the product Count3. If yes, the exit iteration condition is met. If no, the exit iteration condition is not met.

[0047] The set parameter δ is a positive number less than 1.

[0048] Optionally, the selecting unit for calculating the recommended score of each indicator to be screened in the iteration cycle indicator set comprises:

[0049] The first calculation module is configured to calculate, for each indicator to be screened i in the iteration cycle indicator set, an indicator association score scoreX i and a product association score scoreY i .

[0050] The obtaining module is configured to obtain an indicator association weight α of each indicator to be screened i, the indicator association weight α being proportional to the indicator association weight.

[0051] The second calculation module is configured to, for each indicator to be screened i, calculate, according to the respective indicator association weight α, the indicator association score scoreX i and the product association score scoreY iThe recommended score corresponding to the index i to be screened is obtained by calculation. i .

[0052] Based on the product selection testing method and apparatus provided in the above embodiments of the present invention, the following steps are taken: A full set of indicators and a set of available indicators generated based on the product information of the product under test are obtained. The full set of indicators includes a product set and a historical set of indicators, where each product in the product set and each historical indicator in the historical set have a many-to-many relationship. Non-repeating indicators to be screened from the available set and the historical set are selected to form an iterative cycle indicator set, and the recommended score for each indicator to be screened in the iterative cycle indicator set is calculated. J rounds of indicator screening iterations are performed, selecting qualified available indicators from the iterative cycle indicator set and adding them to the available indicator set. The initial value of J is 1. The available indicator set after J rounds of indicator screening iterations is obtained as the recommended indicator set for the product under test. In this embodiment of the invention, an iterative cycle indicator set is constructed using the available set of indicators corresponding to the product under test and the set of non-repeating indicators to be screened from the historical set. The indicators to be screened in this iterative cycle indicator set are then screened according to J rounds of indicator screening iterations, ultimately resulting in a recommended indicator set constructed from the available set of indicators and the screened indicators. This automatic screening process not only saves a lot of manpower and time resources, but also uses an iterative screening method to match indicators, making the selected indicators more accurate and further improving the accuracy of product selection. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0054] Figure 1 This is an architecture diagram of a product selection testing system disclosed in an embodiment of the present invention;

[0055] Figure 2 This is a schematic diagram illustrating the execution flow of an index recommendation algorithm disclosed in an embodiment of the present invention;

[0056] Figure 3 This is a flowchart illustrating a product selection testing method disclosed in an embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram of the structure of a product selection testing device disclosed in an embodiment of the present invention. Detailed Implementation

[0058] With reference to the drawings and embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the scope of the present application.

[0059] In the present application, the term "comprising" or "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or further includes elements inherent to such a process, method, article or device. Without more limitations, the element defined by the phrase "including a" does not exclude the presence of additional identical elements in the process, method, article or device including the element.

[0060] As shown in Figure 1 The product selection test system includes an index management module 11, an index recommendation module 12 and a scene management module 13. The index management module 11, the index recommendation module 12 and the scene management module 13 can realize the closed-loop process of selection index input, recommendation and selection.

[0061] The index management module 11 is used to store all historical selection test data, and form a full index set based on the stored historical selection test data.

[0062] Before the index management module 11 stores any historical selection test data, the historical selection test data to be stored is subjected to unified formatting processing and content supplementing.

[0063] Specifically, the historical selection test data subjected to unified formatting processing and content supplementing in the index management module 11 at least includes product name, product description, test index name and test index description.

[0064] Specifically, the full index set is formed by a product set N and a historical index set M, wherein the product and the historical index are in a many-to-many relationship. One product can have multiple historical indexes, and one historical index can belong to multiple products at the same time.

[0065] The index management module 11 also supports basic functions such as query, input and deletion of historical selection test data.

[0066] The index recommendation module 12 can record the standardized product information of the product A to be tested, which at least includes the product name and product description. The index recommendation module 11 pushes the product information to the expert group and test engineers, and obtains the available index set q of the product A to be tested given by the expert group and test engineers.

[0067] The index recommendation module 12 performs content improvement on the available index set q based on the settings, and the improved content includes the index name and index description.

[0068] The index recommendation module 12 obtains the full index set formed based on historical selection test data in the index management module 11, compares the obtained available index set q with the historical index set M in the historical index set, obtains non-repeating indexes in the available index set and the historical index set, and constructs an iterative loop index set based on the non-repeating indexes.

[0069] The index recommendation module 12 screens the iterative loop index set according to the index recommendation algorithm, and adds the screened available indexes to the available index set to obtain the final recommended index set of the product A to be tested.

[0070] The scene management module 13 is used for users with administrator permission level to adjust or set the data, product classification, data processing, and index recommendation algorithm settings in the index management module 11 and the index recommendation module 12, etc. within the permission.

[0071] Suppose the available index set corresponding to the product A to be tested is q, the full index set includes the product set N and the historical index set M, and the available index set q may exist in the historical index set M or be a new index.

[0072] The intermediate index set M' = q U M, the number of indexes in the intermediate index set is represented by Count, and the number of indexes in the full index set is changed to Count(M'). At this time, Count(q) is much smaller than Count(M') and q is a subset of M', that is, Count(q) << Count(M') and q e M'.

[0073] In order to avoid that the number of available indexes cannot meet the test needs, other available indexes also need to be recommended. In addition to the selected available index set q, the number of selectable indexes becomes Count(M'-q), and the corresponding index set is the iterative loop index set (M'-q). Each index has a recommendation score, and score i is the score of the i-th index, 1 << i << Count(M'-q).

[0074] Based on this, a variable J is set as the round, and initialized as 1. Here, the execution flow of the index recommendation algorithm involved in the index recommendation module 11 disclosed in the embodiments of the present application is exemplified, as shown in the following table, mainly including: Figure 2

[0075] S201: Sort each index in the current iteration loop index set (M'-q) according to the respective recommended score from high to low.

[0076] S202: Determine whether the first n indexes are added to the indexes of the current product test. If h indexes are added, execute S203, and if no indexes are added, execute S205.

[0077] Wherein, n is a set variable, and h is less than n.

[0078] S203: Remove all n indexes from the iteration loop index set (M'-q).

[0079] After executing S203, the remaining index number of the iteration loop index set (M'-q) after removing n indexes is Count (M'-q)-J×n.

[0080] S204: Determine whether the remaining indexes in the current iteration loop index set (M'-q) satisfy the exit condition Count (M'-q)-J×n<δ×Count (M'), if yes, exit, and if no, return to execute S201 for J++ rounds.

[0081] Wherein, the parameter δ is 0.2.

[0082] S205: Determine whether the indexes in the current iteration loop index set (M'-q) satisfy the exit condition Count (M'-q)-J×n<δ×Count (M'), if yes, exit, and if no, execute S206.

[0083] S206: Remove all n indexes from the iteration loop index set (M'-q), and return to execute S202 for J++ rounds.

[0084] In the process of executing S206, after removing the n indexes of this round, there is no need to reorder, and J++ rounds of iteration are performed.

[0085] When the above index recommendation algorithm flow ends, the available index set of the product A under test can be confirmed as the union set of the indexes in the initial available index set q and the indexes h added in each round.

[0086] ​In the product selection test system disclosed in the embodiment of the application, the iteration cycle index set is constructed by using the non-repeated to-be-screened index set in the available index set and the historical index set corresponding to the to-be-tested product, the to-be-screened indexes in the iteration cycle index set are screened according to the J-round index screening iteration operation, and finally the recommended index set is obtained by using the available index set and the screened indexes. The automatic screening process not only saves a large amount of human cost and time resources, but also makes the selected indexes more accurate by using the iteration screening matching index mode, thereby further improving the accuracy of product selection.

[0087] The embodiment of the application further discloses a product selection test method which can be applied to the product selection test system disclosed in the embodiment of the application. Figure 3 As shown in the product selection test method, the product selection test method mainly comprises the following steps.

[0088] S301: An available index set is obtained based on the product information of a to-be-tested product.

[0089] In S301, the full-amount index set comprises a product set and a historical index set, and each product in the product set and each historical index in the historical index set are in a many-to-many relationship. One product can have multiple historical indexes, and one historical index can belong to multiple products.

[0090] S302: To-be-screened indexes that are not repeated in the available index set and the historical index set are selected and collected to obtain an iteration cycle index set, and a recommended score of each to-be-screened index in the iteration cycle index set is calculated.

[0091] In the specific execution of S302, the indexes in the available index set are compared with the indexes in the historical index set, the indexes that are not repeated in the available index set and the historical index set are selected as to-be-screened indexes, and the to-be-screened indexes are collected to obtain the iteration cycle index set.

[0092] For example, the indexes in the available index set are {a, b, c}, the indexes in the historical index set are {a, d, e, f}, and the indexes in the iteration cycle index set obtained by executing S302 are {b, c, d, e, f}.

[0093] S303: A J-round index screening iteration operation is performed, and indexes meeting the requirements are selected from the iteration cycle index set and added to the available index set, and the initial value of J is 1.

[0094] The recommended index algorithm shown in FIG. 4 can be used to specifically implement S303. Figure 2 The recommended index algorithm shown in FIG. 4 can be used to specifically implement S303.

[0095] S304: The available index set after the J-round index screening iteration operation is obtained as the recommended index set of the to-be-tested product.

[0096] In the product selection test method disclosed in the embodiments of the present application, the iteration cycle index set is constructed by using the non-repeated screening index set in the available index set and the historical index set corresponding to the product to be tested, the screening index in the iteration cycle index set is screened according to the J-round index screening iteration operation, and finally the recommended index set is obtained by using the available index set and the screened index. The automatic screening process not only saves a large amount of human cost and time resources, but also makes the selected index more accurate by using the iteration screening matching index, thereby further improving the accuracy of product selection.

[0097] Based on the product selection test method disclosed in the embodiments of the present application, the J-round index screening iteration operation involved in S303 is mainly as follows:

[0098] S401: The screening indexes currently existing in the iteration cycle index set are sorted in descending order of the respective recommended scores.

[0099] S402: The top n screening indexes are selected, and it is judged whether there is an available index for the product selection test in the n screening indexes. If there are h available indexes, S403 is executed; if there is no available index, S405 is executed.

[0100] Wherein, n is a positive integer greater than 1. h < n.

[0101] S403: The h available indexes for the product selection test are added to the available index set.

[0102] S404: The n screening indexes are deleted from the iteration cycle index set, and it is judged whether the screening indexes currently existing in the iteration cycle index set meet the iteration exit condition. If the iteration exit condition is not met, J is increased by 1 and S401 is executed for the next round of index screening iteration operation. If the iteration exit condition is met, the iteration is exited.

[0103] S405: It is judged whether the screening indexes currently existing in the iteration cycle index set meet the iteration exit condition. If the iteration exit condition is not met, S406 is executed. If the iteration exit condition is met, the iteration is exited.

[0104] S406: The n screening indexes are deleted from the iteration cycle index set, J is increased by 1, and S402 is executed to select the top n indexes from the remaining indexes in the iteration cycle index set.

[0105] The execution process and principle of S401 to S406 are the same as the execution process of the recommended index algorithm shown in Figure 2 The execution process of the recommended index algorithm shown in

[0106] In an embodiment of the present application, after the n indicators to be screened are deleted from the iteration loop indicator set in S404, before determining whether the indicators currently existing in the iteration loop indicator set meet the iteration exit condition, and after the n indicators to be screened are deleted from the iteration loop indicator set in S406, J is increased by 1 and the process returns to S402, the method further comprises:

[0107] Determining whether the number of indicators currently existing in the iteration loop indicator set is less than n. If yes, the indicator screening iteration operation is exited and it is not necessary to determine whether the iteration exit condition is met. If no, it is determined whether the iteration exit condition is met.

[0108] In an embodiment of the present application, the specific execution process of determining whether the indicators currently existing in the iteration loop indicator set meet the iteration exit condition in S404 and S405 comprises:

[0109] S501: Acquiring the total number Count1 of indicators currently existing in the iteration loop indicator set.

[0110] S502: Acquiring the product Count3 of the total number Count2 of indicators in the union of the available indicator set and the historical indicator set and a set parameter δ.

[0111] The set parameter δ is a positive number less than 1. Optionally, the set parameter δ is in the range of 20% to 40%.

[0112] S503: Determining whether the total number Count1 of indicators is less than the product Count3. If yes, the iteration exit condition is met. If no, the iteration exit condition is not met.

[0113] Based on the product selection test method disclosed in the above embodiments of the present application, the process of calculating the recommended score of each indicator to be screened in the iteration loop indicator set in S303 shown above mainly comprises:

[0114] First, the indicator correlation score scoreX i and the product correlation score scoreY i of each indicator to be screened i in the iteration loop indicator set are calculated.

[0115] The calculation process of the indicator correlation score scoreX i is as follows:

[0116]

[0117] q′ is the iteration loop indicator set, and the indicator j to be screened is selected from the iteration loop indicator set q′.

[0118] Cal1(i, j) describes the index relevance between the index i to be screened and the index j to be screened (1≤j≤Count(q’)), where the index X i is the sum of the relevance of the index i to all test indexes in the current iteration index set.

[0119] Where the method of obtaining the relevance uses the word segmentation algorithm in natural language processing, and compares the index description of the index i to be screened and the index j to be screened according to the word segmentation result. Cal1(i, j) is the number of same words after word segmentation comparison. After iterating the iteration index set q’, the sum is obtained, that is, X i .

[0120] In X i , the maximum value X max is selected, and X min is normalized between 0.1 and 1 to obtain scoreX i .

[0121]

[0122] The product relevance operator is Cal2, the transition variable is Y i , and the calculation process of the product relevance score scoreY i is as follows:

[0123] In the product set N in the full index set, the tth product, the corresponding index subset r in the intermediate index set M’=q∪M, considering the intersection between r and the iteration index set q’, the product relevance operator of the tth product is:

[0124] Cal2(i, t)=Count(r∩q’)×I (i，t) (I (i，t) ∈(0, 1))

[0125] Where M is the historical index set in the full index set, and q is the available index set corresponding to the product to be tested.

[0126] If the index i to be screened exists in the product t, then I (i，t) =1, if the index i does not exist in the product t, then I (i，t) =0. After iterating and summing the N products in the product set, Y i is obtained.

[0127]

[0128] In Y i , the maximum value Y max is selected, and Y min is normalized between 0.1 and 1 to obtain scoreYi .

[0129]

[0130] Then, the indicator association weight α of each indicator i to be screened is obtained. The value of the indicator association weight α is proportional to the indicator association weight. This indicator association weight α is a floating parameter. If the indicator association weight α increases, the indicator association weight increases, 1-α decreases, and the product association weight decreases.

[0131] Finally, for each index i to be screened, based on its corresponding index association weight α and index association score X... i The product-related score (scoreY) i The recommended score corresponding to the index i to be screened is obtained by calculation. i .

[0132] Final recommendation score:

[0133] score i =α×scoreX i +(1-α)×scoreY i

[0134] In the product selection testing method disclosed in this invention, the aforementioned indicator recommendation algorithm can intelligently recommend the most matching indicator values. This automatic screening process not only saves a significant amount of manpower and time resources, but also employs an iterative screening method to select matching indicators, making the selected indicators more accurate and further improving the accuracy of product selection.

[0135] Based on the product selection testing method disclosed in the above embodiments of the present invention, the present invention also discloses a product selection testing device, such as... Figure 4 As shown, the device includes an acquisition unit 41, a selection unit 42, an iterative filtering unit 43, and a recommendation unit 44.

[0136] The acquisition unit 41 is used to acquire a full set of indicators and an available set of indicators generated based on the product information of the product to be tested. The full set of indicators includes a product set and a historical set of indicators. Each product in the product set and the historical indicators in the historical set have a many-to-many relationship.

[0137] The selection unit 42 is used to select non-repeating indicators to be screened from the available indicator set and the historical indicator set to obtain an iterative cycle indicator set, and to calculate the recommended score for each indicator to be screened in the iterative cycle indicator set.

[0138] In an embodiment of the present application, the selecting unit 42 for calculating the recommended score of each to-be-screened index in the iterative cycle index set comprises:

[0139] The first calculating module is configured to calculate the corresponding index correlation score scoreX i and the product correlation score scoreY i for each to-be-screened index i in the iterative cycle index set.

[0140] The acquiring module is configured to acquire the index correlation weight α of each to-be-screened index i, the value of the index correlation weight α being proportional to the index correlation weight.

[0141] The second calculating module is configured to calculate, for each to-be-screened index i, the corresponding recommended score score i based on the respective corresponding index correlation weight α, the index correlation score scoreX i and the product correlation score scoreY i .

[0142] The iterative screening unit 43 is configured to perform J rounds of index screening iteration operations to select the required available indexes from the iterative cycle index set to add to the available index set, the initial value of J being 1.

[0143] In an embodiment of the present application, the iterative screening unit 43 for performing J rounds of index screening iteration operations comprises:

[0144] The sorting module is configured to sort the to-be-screened indexes currently existing in the iterative cycle index set in descending order of the respective recommended scores.

[0145] The first judging module is configured to select the top n to-be-screened indexes, judge whether there is an available index for the product selection test in the n to-be-screened indexes, and if there are h available indexes, execute the adding module; if there is no available index, execute the second iteration module; wherein n is a positive integer greater than 1, and h < n.

[0146] The adding module is configured to add the h available indexes for the product selection test to the available index set.

[0147] The first iteration module is configured to delete the n to-be-screened indexes from the iterative cycle index set, and execute the second judging module, if the result of the second judging module is that the iteration condition is not met, J is incremented by 1, and the sorting module is executed.

[0148] The second iteration module is configured to execute the second judgment module, and if the result of the second judgment module is that the exit iteration condition is not met, the n to-be-screened indexes are deleted from the iteration loop index set, J is increased by 1, and the first iteration module is executed.

[0149] The second judgment module is configured to judge whether the to-be-screened indexes currently existing in the iteration loop index set meet the exit iteration condition.

[0150] Optionally, the second judgment module is specifically configured to:

[0151] The total number Count1 of indexes of the to-be-screened indexes currently existing in the iteration loop index set is obtained, and the product Count3 of the total number Count2 of indexes in the union of the available index set and the historical index set and a set parameter δ is obtained, and it is judged whether the total number Count1 of indexes is less than the product Count3; if yes, the exit iteration condition is met; if no, the exit iteration condition is not met.

[0152] The set parameter δ is a positive number less than 1.

[0153] In an embodiment of the present application, the iteration screening unit 43 further comprises:

[0154] An exit module is configured to exit the index screening iteration operation if the number of to-be-screened indexes currently existing in the iteration loop index set is less than n.

[0155] A recommendation unit 44 is configured to obtain the available index set after J rounds of index screening iteration operation as the recommended index set of the to-be-tested product.

[0156] In the product selection test device disclosed in the embodiment of the present application, the iteration loop index set is constructed by using the to-be-screened index set which is not repeated in the available index set and the historical index set corresponding to the to-be-tested product, the to-be-screened indexes in the iteration loop index set are screened according to J rounds of index screening iteration operation, and finally the recommended index set is obtained by constructing the available index set and the screened indexes. The automatic screening process not only saves a large amount of human cost and time resources, but also makes the selected indexes more accurate by using the iteration screening matching indexes, and further improves the accuracy of product selection.

[0157] The various embodiments described in this specification are described in progressive order of complexity, from the simplest embodiment to more complex embodiments. Identify commonalities in the various embodiments so that the disclosure is not redundant. Each embodiment is directed to the differences between that embodiment and the other embodiments. In particular, the system or system embodiments are described more simply because they are substantially similar to the method embodiments. The system or system embodiments are described with reference to the method embodiments. The systems and system embodiments described above are merely illustrative of the principles of this application. Any feature in a drawing figure can be implemented in either hardware or software, or a combination thereof, and one skilled in the art would understand that the features of the examples described in this specification can be combined or interchanged, one of ordinary skill in the art can understand and implement the disclosure without undue experimentation.

[0158] Those skilled in the art will further appreciate that the units and algorithms described in connection with the examples disclosed herein can be embodied directly in hardware, in software, or in a combination of the two. For the sake of brevity, descriptions of these alternatives are not repeated here. In the interest of clarity, not all of the individual components of the examples described herein are shown in the figures. Those skilled in the art will appreciate that the functions of the various elements can be carried out by software, hardware, or a combination of the two. Depending upon the implementation chosen, certain implementations of the functionality described herein can be implemented in software, while other implementations can be implemented in hardware or firmware. Those skilled in the art will appreciate that the examples described herein can be practiced with a variety of computer systems other than that described. None of the elements are essential to the practice of the application unless the context demands otherwise.

[0159] The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A product selection test method characterized by, The method comprises: acquiring a full-quantity index set and an available index set generated based on product information of a product to be tested, the full-quantity index set comprising a product set and a historical index set, each product in the product set and each historical index in the historical index set being in a many-to-many relationship; selecting non-repeated screening indexes in the available index set and the historical index set to form a set of iteration cycle indexes, and calculating a recommended score of each screening index in the set of iteration cycle indexes; the recommended score of the screening index being determined based on a sum of a correlation degree of the screening index with all test indexes in the set of iteration cycle indexes and a sum of product correlation degrees of the screening index with all products in the product set; performing J-round index screening iteration operations to select required available indexes from the set of iteration cycle indexes in an iterative manner and add the available indexes to the available index set, an initial value of J being 1; acquiring the available index set after the J-round index screening iteration operations as a recommended index set of the product to be tested.

2. The method of claim 1, wherein, The J-round index screening iteration operations comprise: sorting the screening indexes currently existing in the set of iteration cycle indexes in descending order of their respective recommended scores; selecting n screening indexes with the highest ranks, determining whether there are available indexes for product selection testing in the n screening indexes, n being a positive integer greater than 1; if there are h available indexes, adding the h available indexes for product selection testing to the available index set, h < n; deleting the n screening indexes from the set of iteration cycle indexes, determining whether the screening indexes currently existing in the set of iteration cycle indexes meet an iteration exit condition; if the iteration exit condition is not met, increasing J by 1 to perform a next round of index screening iteration operations; if there are no available indexes, determining whether the screening indexes currently existing in the set of iteration cycle indexes meet the iteration exit condition; if the iteration exit condition is not met, deleting the n screening indexes from the set of iteration cycle indexes, increasing J by 1, and returning to the step of selecting n indexes with the highest ranks.

3. The method of claim 2, wherein, Further comprising: if the number of the screening indexes currently existing in the set of iteration cycle indexes is less than n, exiting the index screening iteration operations.

4. The method of claim 2, wherein, The determination of whether the screening indexes currently existing in the set of iteration cycle indexes meet the iteration exit condition comprises: acquiring a total number Count1 of the screening indexes currently existing in the set of iteration cycle indexes; acquiring a product Count3 of a total number Count2 of indexes in a union of the available index set and the historical index set and a set parameter δ, the set parameter δ being a positive number less than 1; determining whether the total number Count1 of the indexes is less than the product Count3; if yes, the iteration exit condition is met; if no, the iteration exit condition is not met.

5. The method according to any one of claims 1 to 4, characterized in that, The calculation of the recommended score of each screening index in the set of iteration cycle indexes comprises: calculating for each to-be-screened indicator i in the set of iteration loop indicators a corresponding indicator association score scoreX i and a product association score scoreY i ; acquiring an index correlation weight α of each screening index i, the index correlation weight α being proportional to an index correlation weight; For each to-be-screened index i, according to the respective corresponding index correlation weight a, the index correlation score scoreX i and the product correlation score scoreY i , a calculation is performed to obtain the recommended score score i corresponding to the to-be-screened index i.

6. A product selection testing apparatus characterized by comprising: The device comprises: The acquisition unit is configured to acquire a full-quantity indicator set and an available indicator set generated based on product information of a product to be tested, the full-quantity indicator set comprising a product set and a historical indicator set, each product in the product set and each historical indicator in the historical indicator set being in a many-to-many relationship; The selection unit is configured to select, from the available indicator set and the historical indicator set, non-repeated to-be-screened indicators to form a set, to obtain an iterative loop indicator set, and to calculate a recommended score of each to-be-screened indicator in the iterative loop indicator set; the recommended score of the to-be-screened indicator is determined based on a sum of an association degree of the to-be-screened indicator with all test indicators in the iterative loop indicator set and a sum of product association degrees of the to-be-screened indicator with all products in the product set; The iterative screening unit is configured to perform J rounds of indicator screening iteration operations to select, in an iterative manner, required available indicators from the iterative loop indicator set to add to the available indicator set, an initial value of J being 1; The recommendation unit is configured to acquire the available indicator set after the J rounds of indicator screening iteration operations as a recommended indicator set of the product to be tested.

7. The apparatus of claim 6, wherein, The iterative screening unit configured to perform the J rounds of indicator screening iteration operations comprises: The sorting module is configured to sort, according to respective recommended scores, to-be-screened indicators currently existing in the iterative loop indicator set from high to low; The first judgment module is configured to select n to-be-screened indicators with the highest rankings, to judge whether there are available indicators for product selection testing this time in the n to-be-screened indicators, to execute the adding module if there are h available indicators, and to execute the second iteration module if there are no available indicators; n is a positive integer greater than 1, and h < n; The adding module is configured to add the h available indicators for product selection testing this time to the available indicator set; The first iteration module is configured to delete the n to-be-screened indicators from the iterative loop indicator set, to execute the second judgment module, to increase J by 1 if a result of the second judgment module is that an iteration exit condition is not met, and to execute the sorting module; The second iteration module is configured to execute the second judgment module, to delete the n to-be-screened indicators from the iterative loop indicator set if a result of the second judgment module is that the iteration exit condition is not met, to increase J by 1, and to execute the first iteration module; The second judgment module is configured to judge whether to-be-screened indicators currently existing in the iterative loop indicator set meet an iteration exit condition.

8. The apparatus of claim 7, wherein, Further comprising: The exit module is configured to exit the indicator screening iteration operation if a number of to-be-screened indicators currently existing in the iterative loop indicator set is less than n.

9. The apparatus of claim 7, wherein, The second judgment module is configured to: acquire a total number Count1 of to-be-screened indicators currently existing in the iterative loop indicator set, acquire a product of a total number Count2 of indicators in a union of the available indicator set and the historical indicator set and a set parameter δ, and judge whether the total number Count1 of to-be-screened indicators is less than the product Count3; if yes, the iteration exit condition is met; if no, the iteration exit condition is not met; wherein the set parameter δ is a positive number less than 1.

10. The apparatus of any one of claims 6 to 9, wherein, The selecting unit that calculates the recommended score of each to-be-screened index in the iteration cycle index set comprises: a first calculation module configured to calculate, for each to-be-screened indicator i in the set of iteration cycle indicators, a corresponding indicator correlation score scoreX i and a product correlation score scoreY i ; An acquisition module is configured to acquire an index correlation weight α of the each to-be-screened index i, wherein the index correlation weight α is proportional to the index correlation weight. a second calculating module, configured to, for each to-be-screened index i, calculate a recommended score scorei of the to-be-screened index i according to the index correlation weight α corresponding to the to-be-screened index i, the index correlation score scoreX corresponding to the to-be-screened index i, and the product correlation score scoreY i i i .​​

Images