AI search recommendation method and system for purchasing

Abstract

The invention belongs to the technical field of intelligent recommendation, and particularly relates to an AI search recommendation method and system for purchasing, and the method comprises the steps: according to the regularity of the interval time period of historical orders containing commodities belonging to two large classes in all historical orders of a related user, determining the historical orders of the related user; determining the association degree of the two large classes in the historical orders of all related users, further determining whether the two large classes belong to association classes, and for the two large classes which do not belong to the association classes, further determining whether the two large classes belong to the association classes according to the regularity of the interval time periods of the historical orders of the commodities belonging to the two large classes in all the historical orders of all related users. Determining the stages of the two large classes in the historical order, and further determining whether the two large classes belong to advanced classes or not; and recommending the commodities to the user according to the association class and the advanced class of the large class to which the commodities belong in the reference order of the user. According to the invention, the recommendation accuracy and the user demand coverage range are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent recommendation technology. More specifically, the present invention relates to an AI search recommendation method and system for procurement. Background Art

[0002] The recommendation system uses data analysis technology to find out the products that users are most likely to be interested in and recommend them to them.

[0003] Item-based collaborative filtering (Item-CF) is a classic method in the field of recommendation systems. Its core idea is to quantify the similarity between items by calculating users' historical ratings or behavioral data (such as clicks, purchases) on items, and recommend products that users may be interested in based on similar items.

[0004] This method relies on the density of the user-item interaction matrix and usually uses cosine similarity or Pearson correlation coefficient to measure item association.

[0005] However, in practical applications, item-based collaborative filtering algorithms only judge similarity through co-purchase or ratings, resulting in recommendation results limited to products with similar functions, making it difficult to explore cross-category associations and advanced needs; in addition, item-based collaborative filtering algorithms only focus on static co-occurrence situations and lack temporal dynamic analysis, resulting in mismatched recommendation timing; therefore, item-based collaborative filtering algorithms have the problem of difficulty in exploring cross-category association needs and mismatched recommendation timing, which seriously affects the accuracy of recommendations. Summary of the invention

[0006] In order to solve the technical problem that the above-mentioned item-based collaborative filtering algorithm is difficult to mine cross-category related demands and the recommendation timing is not matched, which seriously affects the accuracy of recommendation, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides an AI search and recommendation method for procurement, including: obtaining the major category to which each commodity belongs through a procurement item classification catalog; taking any major category as the target major category, and denoting any major category other than the target major category as the reference major category; for all historical orders of a relevant user: denoting the historical orders containing commodities belonging to the target major category as target orders, determining whether there are associated orders among the target orders in all historical orders containing commodities belonging to the reference major category according to the interval time period between historical orders, and determining the degree of association between the target major category and the reference major category in the historical orders of the relevant user according to the number of target orders with associated orders and the magnitude and regularity of the duration of the interval time period between the target orders with associated orders and their associated orders; determining whether the target major category and the reference major category belong to associated categories according to the degree of association between the target major category and the reference major category in the historical orders of all relevant users; if the target major category and the reference major category do not belong to associated categories: determining the stage of the target major category and the reference major category in the historical orders according to the number of target orders with associated orders of all relevant users and the regularity of the duration of the interval time period between the target orders with associated orders and their associated orders, and further determining whether the target major category and the reference major category belong to advanced categories; recommending commodities to the user according to the associated categories and advanced categories of the major categories to which the commodities in the user's reference orders belong.

[0008] Preferably, the relevant user refers to a user who has both historical orders containing commodities belonging to the target major category and historical orders containing commodities belonging to the reference major category.

[0009] Preferably, the determining whether there are associated orders among the target orders in all historical orders containing commodities belonging to the reference major category according to the interval time period between historical orders includes: denoting the target major category as the target major category , and denoting the reference major category as the reference major category ; for the th target order , among all historical orders containing commodities belonging to the reference major category: if there are other target orders in the interval time period between all historical orders containing commodities belonging to the reference major category and the target order , then the target order has no associated order; otherwise, among the multiple historical orders in which there are no other target orders in the interval time period with the target order , the historical order with the shortest interval time period with the target order is denoted as the associated order of the target order .

[0010] Preferably, the determining the degree of association between the target major category and the reference major category in the historical orders of the relevant user includes: ; where is the degree of association between the target major category and the reference major category in the historical orders of the relevant users, is the mean value of the durations of the intervals between all target orders with associated orders and their associated orders, is the number of target orders with associated orders, is the number of all target orders, is the number of all historical orders containing items belonging to the reference major category, is the standard deviation of the durations of the intervals between all target orders with associated orders and their associated orders, characterizes the regularity of the durations of the intervals between target orders with associated orders and their associated orders, is the Sigmoid function.

[0011] Preferably, determining whether the target major category and the reference major category belong to the associated category according to the degree of association between the target major category and the reference major category in the historical orders of all relevant users includes: arranging the degrees of association between the target major category and the reference major category in the historical orders of all relevant users in ascending order to obtain the first quartile after arrangement: if the first quartile is greater than the preset association threshold, the target major category and the reference major category belong to the associated category; if the first quartile is less than or equal to the preset association threshold, the target major category and the reference major category do not belong to the associated category.

[0012] Preferably, determining the periodicity of the target major category and the reference major category in the historical orders includes: ; where is the periodicity of the target major category and the reference major category in the historical orders, is the number of relevant users within of the duration of the interval between the target order with an associated order and its associated order, is the range of the periodicity duration of the target major category and the reference major category in the historical orders, is the number of all relevant users, is the standard deviation of the duration corresponding to the range of the periodicity duration, is the natural exponential function, is the Sigmoid function, characterizes the regularity of the durations of the intervals between target orders with associated orders of all relevant users and their associated orders.

[0013] Preferably, the range of the periodicity duration and the standard deviation of the duration corresponding to the range of the periodicity duration include: obtaining the intervals between the target orders with associated orders of all relevant users and their associated orders, arranging the durations of all the intervals in ascending order, and calculating the duration within The standard deviation of the durations of all interval time periods within the range , represents the percentile after sorting, represents the percentile after sorting; is an integer within the range, and obtain the corresponding range with the smallest value of , as the phased duration range of the target major category and the reference major category in historical orders, and use it as the duration standard deviation corresponding to the phased duration range.

[0014] Preferably, determining whether the target major category and the reference major category belong to the advanced category includes: if the phase in the historical orders of the target major category and the reference major category is greater than a preset phase threshold, then the target major category and the reference major category belong to the advanced category.

[0015] Preferably, the method for obtaining the reference orders of the user includes: taking the maximum value of the upper limits of the durations of the phased duration ranges of all two major categories with an advanced relationship in historical orders as the advanced reference duration of the user ; obtaining, from all the historical orders of the user, the historical orders whose order placement times are within the range, and denoting them as the first reference orders of the user, where is the current time; taking the average value of the durations of the interval time periods between the target orders with associated orders and their associated orders in all the historical orders of the user as the associated reference duration of the user ; obtaining, from all the historical orders of the user, the historical orders whose order placement times are within the

[0016] In a second aspect, the present invention provides an AI search and recommendation system for procurement, including a processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned AI search and recommendation method for procurement is implemented.

[0017] By adopting the above technical solution, generating a computer program for the above-mentioned AI search and recommendation method for procurement and storing it in the memory to be loaded and executed by the processor, thereby manufacturing a terminal device according to the memory and the processor, which is convenient to use.

[0018] The beneficial effects of the present invention are as follows: Through the collaboration of large-category-level association mining, time dynamics modeling, and group statistical decision-making, the present invention not only solves the problem of homogeneous recommendations in traditional methods by distinguishing between associative and progressive relationships, short-term and long-term needs, but also makes up for the lack of time dynamics through a dynamic time window mechanism. Ultimately, it comprehensively improves the diversity and timeliness of recommendation results, significantly enhancing the recommendation accuracy and the breadth of user needs coverage. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flowchart schematically showing an AI search and recommendation method for procurement in the present invention; Figure 2 is a flowchart schematically showing step S2. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

[0021] Next, the specific embodiments of the present invention will be described in detail in conjunction with the accompanying drawings.

[0022] An embodiment of the present invention discloses an AI search and recommendation method for procurement. Referring to Figure 1 , it includes steps S1 - S3: S1. Obtain the large category to which each commodity belongs through the procurement item classification catalog.

[0023] It should be noted that determining the large category to which each commodity belongs lays a foundation for subsequent mining of the associative and phased relationships between large categories.

[0024] Specifically, the procurement item classification catalog is a detailed item of the specific classification of procurement objects and is the concretization of the procurement scope. Therefore, the procurement item classification catalog includes the large category to which each commodity belongs.

[0025] S2. Determine whether the relationships between all large categories belong to the associative category and the progressive category.

[0026] It should be noted that the item-based collaborative filtering algorithm only models similarity based on static co-occurrence relationships (such as pairs of commodities purchased by the same user), ignoring the time series characteristics of user behavior (such as purchase intervals and phased demands), resulting in homogeneous recommendation results, that is, the recommendation results are limited to commodities with similar functions, unable to distinguish between associative commodities with associative relationships (for example: toothpaste and toothbrush) and phased progressive commodities (for example: baby stroller and toddler walker), and it is difficult to mine cross-category associative and progressive demands.

[0027] Therefore, for all major categories, taking any major category as the target major category and any major category other than the target major category as the reference major category, determine whether the relationship between the target major category and the reference major category belongs to the associated category and the advanced category.

[0028] It should be noted that by traversing every two different major categories and taking them as the target major category and the reference major category respectively, all possible combinations of major categories are systematically covered, avoiding missing potential associated relationships or advanced relationships, and providing a clear comparison benchmark for the subsequent steps.

[0029] The flowchart of step S2 is referred to Figure 2 , including step S201 to step S205, specifically: S201. Determine relevant users.

[0030] Specifically, relevant users refer to users who have both historical orders containing products belonging to the target major category and historical orders containing products belonging to the reference major category.

[0031] S202. Determine the degree of association between the target major category and the reference major category in the historical orders of each relevant user.

[0032] Specifically, for all historical orders of a relevant user: mark the historical orders containing products belonging to the target major category as target orders; determine whether there are associated orders for the target orders among all historical orders containing products belonging to the reference major category according to the interval time period between two historical orders.

[0033] Among them, the interval time period between two historical orders refers to the time period from the order placement time of the first historical order to the order placement time of the second historical order.

[0034] Among them, mark the target major category as the target major category , and mark the reference major category as the reference major category ; for the th target order , among all historical orders containing products belonging to the reference major category: (1) If there are other target orders in the interval time period between all historical orders containing products belonging to the reference major category and the target order , then the target order has no associated order.

[0035] (2) Otherwise, among the multiple historical orders in which there are no other target orders in the interval time period with the target order , mark the historical order with the shortest interval time period with the target order as the associated order of the target order .

[0036] It should be noted that by determining that there are other target orders in the interval time period to screen the target orders with associated orders, it is possible to avoid misjudging repeated purchase behaviors as related purchase behaviors and capture the most direct needs of users.

[0037] It should be noted that the unit of the duration of the interval time period is days.

[0038] Furthermore, according to the number of target orders with associated orders and the magnitude and regularity of the duration of the interval time period between the target orders with associated orders and their associated orders, determine the degree of association between the target major category and the reference major category in the historical orders of the relevant user; then the calculation formula for the degree of association between the target major category and the reference major category in the historical orders of the relevant user is: ; In the formula, is the degree of association between the target major category and the reference major category in the historical orders of the relevant user, is the mean value of the duration of the interval time period between all target orders with associated orders and their associated orders, is the number of target orders with associated orders, is the number of all target orders, is the number of all historical orders containing items belonging to the reference major category, is the standard deviation of the duration of the interval time period between all target orders with associated orders and their associated orders, characterizes the regularity of the duration of the interval time period between the target orders with associated orders and their associated orders, is the Sigmoid function.

[0039] It should be noted that the calculation formula for the degree of association consists of three terms, namely the time mean term , the quantity ratio term and the time regularity term . Among them, for items with an associated relationship, the time interval between purchases is relatively short. Therefore, the smaller the time mean term, the greater the degree of association between the target major category and the reference major category in the historical orders of the relevant user; items with an associated relationship are not sporadic in the orders. Therefore, the larger the quantity ratio term, the greater the degree of association between the target major category and the reference major category in the historical orders of the relevant user; the same user has personal purchase habits. Therefore, for items with an associated relationship, the time interval between purchases has regularity. The time regularity between the target order and its associated order is measured by the standard deviation of the duration. The smaller the standard deviation, the stronger the time regularity, and the greater the degree of association between the target major category and the reference major category in the historical orders of the relevant user.

[0040] S203. Determine whether the target major category and the reference major category belong to associated categories.

[0041] Specifically, determine whether the target major category and the reference major category belong to associated categories according to the degree of association between the target major category and the reference major category in the historical orders of all relevant users, including: arranging the degrees of association between the target major category and the reference major category in the historical orders of all relevant users in ascending order to obtain the first quartile after arrangement: if the first quartile is greater than a preset association threshold, the target major category and the reference major category belong to associated categories; if the first quartile is less than or equal to the preset association threshold, the target major category and the reference major category do not belong to associated categories.

[0042] Among them, the specific value of the association threshold can be set according to the actual application scenario and requirements, and the value range of the association threshold is [0.5, 0.75]. In the present invention, the association threshold is set to 0.6.

[0043] It should be noted that by determining the association through the first quartile, it is possible to avoid the dominance of the results by a small number of highly associated users. At the same time, by filtering out weakly associated combinations through the preset association threshold, only strong associated combinations with high frequency and regular time are retained.

[0044] Furthermore, it should be noted that the determination of associated categories effectively identifies short-term associated demands through time proximity screening and multi-dimensional weight calculation (time mean, quantity ratio, regularity).

[0045] S204. If the target major category and the reference major category do not belong to associated categories, determine the stage of the target major category and the reference major category in the historical orders.

[0046] Specifically, if the target major category and the reference major category do not belong to associated categories: determine the stage of the target major category and the reference major category in the historical orders according to the quantity of target orders with associated orders among all relevant users and the regularity of the time interval duration between the target orders with associated orders and their associated orders, including: 1. Obtain the time intervals between the target orders with associated orders and their associated orders of all relevant users, arrange the durations of all time intervals in ascending order, and calculate the standard deviation of the durations of all time intervals within the range , represents the th percentile after sorting, represents the th percentile after sorting.

[0047] 2. Among them, is For integers within the range, obtain the one that makes take the minimum value of The corresponding range , as the phased duration range of the target major category and the reference major category in historical orders, and As the standard deviation of the duration corresponding to the phased duration range.

[0048] It should be noted that based on the initial interval , gradually expand the upper and lower limits, calculate the standard deviation of the duration of the interval time period within each expanded interval, and select the expanded interval with the smallest difference in the standard deviation of the duration of the interval time period between adjacent expanded intervals as the phased duration range.

[0049] 3. The formula for the phase of the target major category and the reference major category in historical orders is: ; In the formula, is the phase of the target major category and the reference major category in historical orders, is the number of relevant users whose duration of the interval time period between the target order with associated orders and its associated orders is within , is the phased duration range of the target major category and the reference major category in historical orders, , are respectively the lower limit and the upper limit of the duration of the phased duration range of the target major category and the reference major category in historical orders, is the number of all relevant users, is the standard deviation of the duration corresponding to the phased duration range, is the natural exponential function, is the Sigmoid function, characterizes the regularity of the duration of the interval time period between the target order with associated orders and its associated orders of all relevant users.

[0050] It should be noted that the formula for the phase consists of two terms, namely the quantity ratio term and the time regularity term . Among them, in all historical orders of each user, goods with an advanced relationship basically only appear once. Therefore, the number of relevant users whose duration of the interval time period between the target order with associated orders and its associated orders is within The closer the ratio to the number of all relevant users is to 1, the greater the stage in the historical orders of the relevant users for the target major category and the reference major category; there is a regular time for the staged upgrade between products. Therefore, the time regularity is measured by the standard deviation of the duration corresponding to the duration range of the stage. The smaller the standard deviation, the stronger the time regularity, and the greater the stage in the historical orders of the relevant users for the target major category and the reference major category.

[0051] S205. Determine whether the target major category and the reference major category belong to the advanced category.

[0052] Specifically, if the stage in the historical orders of the target major category and the reference major category is greater than the preset stage threshold, then the target major category and the reference major category belong to the advanced category.

[0053] Among them, the specific value of the stage threshold can be set according to the actual application scenario and requirements, and the value range of the stage threshold is [0.6, 0.8]. In the present invention, the stage threshold is set to 0.66.

[0054] It should be noted that the determination of the advanced category captures the staged upgrade requirements in the user life cycle through the extended percentile interval and standard deviation analysis, and supports the triggering of long-cycle recommendations.

[0055] Furthermore, it should be noted that step S2 innovatively introduces time series analysis. Through the correlation degree calculation formula and the stage calculation formula, combined with the statistical behavior of group users and dynamic threshold determination, it accurately distinguishes the associated category and the advanced category.

[0056] S3. Recommend products to the user according to the associated category and the advanced category of the major categories to which the products in the user's reference order belong.

[0057] It should be noted that the item-based collaborative filtering algorithm only focuses on the static co-occurrence situation and lacks the analysis of time dynamics, resulting in mismatched recommendation timing.

[0058] Therefore, according to the advanced reference duration and the associated reference duration of the user, the reference order of the user is determined, and then products are recommended to the user according to the associated category and the advanced category of the major categories to which the products in the user's reference order belong.

[0059] The specific steps are as follows: 1. Denote the maximum value of the upper limit of the duration range of the stage in the historical orders of all two major categories with an advanced relationship as the advanced reference duration of the user ; obtain the historical orders within the range of the order placement time in all historical orders of the user and denote them as the first reference order of the user, where is the current moment.

[0060] 2. Obtain the major category to which the goods in the user's first reference order belong: If there is an advanced category for this major category, further determine the advanced reference duration Whether it is within the range. If the advanced reference duration is within the range, then trigger the advanced category recommendation, and use the advanced category of this major category as the recommended category for the user. is the phased duration range of this major category and its advanced category in the historical orders.

[0061] 3. Take the average value of the durations of the interval time periods between the target orders with associated orders and their associated orders in all the user's historical orders as the user's associated reference duration ; Obtain the historical orders in all the user's historical orders whose order placement times are within the range, and record them as the user's second reference orders. is the current time.

[0062] 4. Obtain the major category to which the goods in the user's second reference order belong: If there is an associated category for this major category, trigger the associated category recommendation, and use the associated category of this major category as the recommended category for the user.

[0063] 5. Recommend goods belonging to the recommended category to the user.

[0064] It should be noted that step S3 generates dynamic recommendations based on the time window constraint of the user's recent behavior. Among them, the advanced category recommendation takes the upper limit of the group phased duration as a reference and combines the extended interval matching to ensure that the recommendation timing conforms to the user's phased demand law; the associated category recommendation takes the individual average interval duration as a window, focuses on the immediate associated demand, and avoids recommending too early or too late.

[0065] The embodiment of the present invention also discloses an AI search and recommendation system for procurement, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an AI search and recommendation method for procurement according to the present invention is implemented.

[0066] The above system also includes other components well known to those skilled in the art such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

Claims

1. An AI search and recommendation method for procurement, characterized in that Including: Obtain the major category to which each commodity belongs through the procurement item classification catalog; Take any major category as the target major category, and record any major category other than the target major category as the reference major category; For all historical orders of a relevant user: Record the historical orders containing commodities belonging to the target major category as target orders. According to the interval time period between historical orders, determine whether there are associated orders among all historical orders containing commodities belonging to the reference major category for the target orders. According to the quantity of target orders with associated orders and the magnitude and regularity of the duration of the interval time period between the target orders with associated orders and their associated orders, determine the degree of association between the target major category and the reference major category in the historical orders of this relevant user; According to the degree of association between the target major category and the reference major category in the historical orders of all relevant users, determine whether the target major category and the reference major category belong to associated categories; If the target major category and the reference major category do not belong to associated categories: According to the quantity of target orders with associated orders among all relevant users and the regularity of the duration of the interval time period between the target orders with associated orders and their associated orders, determine the stage characteristics of the target major category and the reference major category in the historical orders, and further determine whether the target major category and the reference major category belong to advanced categories; Recommend commodities to the user according to the associated categories and advanced categories of the major categories to which the commodities in the user's reference orders belong.

2. The AI search and recommendation method for procurement according to claim 1, wherein, The relevant user refers to a user who has both historical orders containing commodities belonging to the target major category and historical orders containing commodities belonging to the reference major category.

3. The AI search and recommendation method for procurement according to claim 1, wherein, The determination of whether there are associated orders among all historical orders containing commodities belonging to the reference major category for the target orders according to the interval time period between historical orders includes: Denote the target major category as the target major category , and denote the reference major category as the reference major category ; For the th target order , among all historical orders that contain items belonging to the reference major category: If there are other target orders during the interval period between all historical orders that contain items belonging to the reference major category and the target order , then the target order has no associated order; otherwise, among multiple historical orders in which there are no other target orders during the interval period with the target order , the historical order with the shortest interval period with the target order is recorded as the associated order of the target order .

4. An AI search and recommendation method for procurement according to claim 1, characterized in that, The determination of the degree of association between the target major category and the reference major category in the historical orders of this relevant user includes: ； In the formula, is the correlation degree between the target major category and the reference major category in the historical orders of the relevant user, is the average value of the durations of the interval time periods between all target orders with associated orders and their associated orders, is the number of target orders with associated orders, is the number of all target orders, is the number of all historical orders containing products belonging to the reference major category, is the standard deviation of the durations of the interval time periods between all target orders with associated orders and their associated orders, characterizes the regularity of the durations of the interval time periods between target orders with associated orders and their associated orders, is the Sigmoid function.

5. The AI search and recommendation method for procurement according to claim 1, wherein The determination of whether the target major category and the reference major category belong to associated categories according to the degree of association between the target major category and the reference major category in the historical orders of all relevant users includes: Arrange the degree of association between the target major category and the reference major category in the historical orders of all relevant users in ascending order to obtain the first quartile after arrangement: If the first quartile is greater than the preset association threshold, then the target major category and the reference major category belong to associated categories; If the first quartile is less than or equal to the preset association threshold, then the target major category and the reference major category do not belong to associated categories.

6. The AI search and recommendation method for procurement according to claim 5, wherein, The determination of the stage characteristics of the target major category and the reference major category in the historical orders includes: ； In the formula, is the stage of the target major category and the reference major category in historical orders, is the duration of the interval between the target order with associated orders and its associated orders within for the number of relevant users, is the stage duration range of the target major category and the reference major category in historical orders, is the number of all relevant users, is the standard deviation of the duration corresponding to the stage duration range, is the natural exponential function, is the Sigmoid function, characterizes the regularity of the duration of the interval between the target order with associated orders and its associated orders for all relevant users.

7. An AI search and recommendation method for procurement according to claim 6, characterized in that, The stage duration range and the duration standard deviation corresponding to the stage duration range include: Obtain the interval time period between the target order with associated orders and its associated orders for all relevant users, sort the durations of all interval time periods in ascending order, and calculate the standard deviation of the durations of all interval time periods within the range , represents the percentile after sorting, and represents the For an integer within the range, obtain the corresponding range for which the value of is minimized, as the phased duration range of the target major category and the reference major category in historical orders, and use as the standard deviation of the duration corresponding to the phased duration range.

8. An AI search and recommendation method for procurement according to claim 1, characterized in that The determination of whether the target major category and the reference major category belong to advanced categories includes: If the stage characteristics of the target major category and the reference major category in the historical orders are greater than the preset stage threshold, then the target major category and the reference major category belong to advanced categories.

9. An AI search and recommendation method for procurement according to claim 1, characterized in that, The method for obtaining the user's reference orders includes: Denote the maximum value among the upper limits of the time ranges of the phased durations of all two major categories with an advanced relationship in the historical orders as the advanced reference duration of the user. ; Obtain the historical orders of the user in which the order placement time is within the range, and denote them as the first reference orders of the user. is the current time; The average of the durations of the time intervals between a target order with associated orders and its associated orders among all the user's historical orders is used as the user's associated reference duration. ; Obtain the historical orders among all the user's historical orders whose order placement times are within and record them as the user's second reference orders.

10. An AI search and recommendation system for procurement, characterized in that, Including: A processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an AI search and recommendation method for procurement according to any one of claims 1-9 is implemented.

Images