Goods source intelligent recall method and system based on multi-dimensional condition matrix

Abstract

The invention discloses an intelligent goods source recall method and system based on a multi-dimensional condition matrix, and the method comprises the steps: constructing a condition matrix according to a driver search condition, and independently constructing a precise / generalization condition for each dimension of the search condition; according to the method, goods sources are retrieved on the basis of goods source conditions of all dimensions, recall classification is carried out according to accurate / generalization conditions, and labels of recall paths are transmitted to downstream fine arrangement and rearrangement stages step by step to be used by a rearrangement strategy and list floor segmentation rendering. The problems that an existing goods source recall mechanism is low in recall rate, redundant in logic, high in strategy coupling degree, poor in maintainability and the like are solved, unified modeling and flexible configuration of recall logic are achieved, independent expansion and weight regulation and control of all dimensions are supported, the complexity of multi-path parallel recall in an original system is remarkably reduced, and the system performance is improved. The code readability and the maintenance efficiency are improved, the range of visible high-quality goods source pools of drivers is expanded, and the increase of orders carried out by the platform is driven.

Description

Technical Field

[0001] This invention relates to the field of online freight platform development technology, and more specifically, to a method and system for intelligent cargo retrieval based on a multi-dimensional condition matrix. Background Technology

[0002] When drivers access a freight platform app to search for cargo, the system needs to retrieve suitable cargo from tens of thousands of available listings—a process known as cargo retrieval. Besides retrieving suitable cargo, the system also needs to tag and categorize the cargo, such as: precise cargo, semi-precise cargo, and recommended cargo. The tagging results are then used for display on the product front end. Traditional methods typically employ fixed rules or simple generalization strategies for retrieval, making it difficult to balance breadth of retrieval with business flexibility. This results in shortcomings such as insufficient coverage of potential cargo, redundant and ambiguous retrieval logic across multiple channels, and difficulties in business control, leading to the omission of a large number of potential matching cargo and impacting driver-cargo transaction efficiency. Summary of the Invention

[0003] To address the aforementioned issues, the present invention aims to provide a cargo intelligent recall technology based on a multi-dimensional condition matrix, applicable to cargo search and recommendation scenarios in the cold chain zone of road freight platforms. This technology aims to solve problems such as low recall rate, logical redundancy, high strategy coupling, and poor maintainability in existing cargo recall mechanisms.

[0004] To achieve the above-mentioned technical objectives, an exemplary embodiment of this disclosure provides a first aspect of a method for intelligent retrieval of goods based on a multi-dimensional condition matrix, comprising the following steps: Construct a condition matrix based on driver search criteria, and independently construct precise / generalized conditions for each dimension of the search criteria; Based on all dimensions of the source conditions, the source is searched, and based on the precise / generalized conditions, the recall and regression classes are performed. The tags of the recall loop are passed down level by level to the downstream fine ranking and re-ranking stages for use by the re-ranking strategy and list floor segmentation rendering.

[0005] Furthermore, when constructing the condition matrix, the search conditions include four dimensions: origin, destination, vehicle length, and other filtering conditions. Among them, other filtering conditions include: cargo type, availability of price, price per kilometer, weight, volume, negotiable type, and full truckload and less-than-truckload (LTL) shipments. When constructing precise / generalized conditions, for the origin: Precise condition representation: Strictly matches search conditions, supplemented by basic geographic generalization; Generalized conditional representation: Introduces multi-source extension logic, including advanced geographic proximity, behavioral preferences, computational intent inference, reported intent, relay points along the search route and origin of the return journey, and dynamic contextual information; When constructing precise / generalized conditions, for the destination: Precise condition representation: Strictly matches search conditions, supplemented by basic geographic generalization; Generalization condition representation: Introduces multi-source extension logic, including advanced geographic proximity, behavioral preferences, computational intent inference, reported intent, search route relays, and return destination; When constructing precise / generalized conditions, for vehicle length: Precise criteria: Driver search vehicle length and registered vehicle length; Generalization conditional representation: driver calculates vehicle length based on intent, vehicle length based on behavior, and vehicle length based on reported intent; When constructing precise / generalized criteria, for other filtering conditions: Precise condition representation: Meeting the driver's screening criteria retains the user's explicit selection; Generalized conditions mean: no restrictions on screening conditions, relaxing the restrictions to improve recall coverage.

[0006] Preferably, when performing cargo retrieval, cargo that meets the conditions is retrieved from the cargo pool based on four dimensions of cargo conditions. The four dimensions are used in combination with precise / generalized conditions, and a total of eight retrievals are performed. Among them, cargo retrieval is performed through the cargo ES index, which stores all available cargo in the ES inverted index and supports the following retrieval fields: origin, destination, vehicle length, cargo type, price, weight, volume, negotiable type, and full truckload and less-than-truckload. When there is an update to the information of each shipment, the Kafka message of cargo update is listened for, and the content of the index is updated accordingly.

[0007] Preferably, when classifying recalls, after searching for the source of goods, each shipment has 2^4 = 16 possible tag combinations based on the search criteria: precise / generalized × origin / destination / vehicle length / other, which are then classified into 16 recall routes.

[0008] Based on the same inventive concept, a second aspect of this invention provides a cargo intelligent recall system based on a multi-dimensional condition matrix, comprising: The condition construction module is used to construct a condition matrix based on driver search conditions, and to independently construct precise / generalized conditions for each dimension of the search conditions. The recall module is used to search for goods based on all dimensions of goods conditions, and to perform recall regression based on precise / generalized conditions. The tags of the recall loop are passed down level by level to the downstream fine ranking and re-ranking stages for use by the re-ranking strategy and list floor segmentation rendering.

[0009] Based on the same inventive concept, a third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method of the first aspect.

[0010] Based on the same inventive concept, a fourth aspect of the present invention provides a non-transitory computer-readable storage medium that stores computer instructions for causing a computer to perform the method as described in the first aspect.

[0011] Based on the same inventive concept, the fifth aspect of the present invention provides a computer program product, including computer program instructions, which, when executed on a computer, cause the computer to perform the method as described in the first aspect.

[0012] The present invention discloses the following technical effects: This invention solves the problems of low recall rate, logical redundancy, high strategy coupling and poor maintainability of existing cargo sourcing mechanisms, and expands the scope of the high-quality cargo sourcing pool visible to drivers, thereby driving the growth of fulfillment orders on the platform. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic diagram of the method described in this invention; Figure 2 This is a schematic diagram of the hardware structure of the electronic device described in this invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0016] like Figures 1-2As shown, this invention provides a method for intelligent cargo retrieval based on a multi-dimensional condition matrix. The invention proposes a structured and configurable four-dimensional condition matrix retrieval framework, which decouples the retrieval conditions into four core dimensions: origin, destination, vehicle length, and other screening conditions (such as cargo type, full truckload / less-than-truckload, price, etc.). Each dimension is further divided into two levels: "precise" and "recommended". A condition matrix in the form of a Cartesian product is constructed to achieve systematic and modular cargo retrieval.

[0017] The recall framework mentioned in this invention consists of three parts: constructing a condition matrix, sourcing retrieval, and recall classification. Part 1: Design of the Condition Matrix: Construct a condition matrix based on driver search conditions. The search conditions include four dimensions: origin, destination, vehicle length, and other filtering conditions. Each dimension independently constructs precise / generalized conditions.

[0018] 1. Origin Dimension: Precision: It basically strictly matches the search conditions, supplemented by basic geographic generalization, such as the city-level administrative region corresponding to the origin and the N-kilometer range of the driver's location.

[0019] Generalization: Introducing multi-source extension logic, including advanced geographical proximity, such as adjacent areas of provinces / cities, and Xkm around the center point; behavioral preferences, such as drivers' frequently visited departure points and destinations of accepted orders; calculation of intent inference, such as drivers' departure points with the most clicks and calls in the past 24 hours and in real time; reported intent, such as drivers' intended departure points reported on WeChat; departure points of relays and return trips along the search route; and dynamic contextual information, such as the current location of people / vehicles.

[0020] 2. Destination dimension: Precise: It strictly matches the search criteria and is supplemented by basic geographical generalization, such as the municipal administrative region corresponding to the destination district and the neighboring cities of the city.

[0021] Generalization: Introducing multi-source extension logic, including advanced geographical proximity, such as the province where the city is located, neighboring cities of the province, etc.; behavioral preferences, such as the destinations that drivers frequently travel to and the location of their home address; calculation intent inference, such as the destinations that drivers click or call most frequently in the past 24 hours and in real time; reported intent, such as the intended departure point reported by the driver on WeChat Work; and destinations along the search route for relays and return trips.

[0022] 3. Vehicle length dimension: Precision: Drivers can search for and register drivers.

[0023] Generalization: Drivers calculate intent to infer vehicle length, such as vehicle lengths with more clicks and calls in the past 24 hours and in real time; behavioral vehicle lengths, such as vehicle lengths with more transactions in the past few months; and vehicle lengths that report intent, such as vehicle lengths that drivers report on WeChat for business.

[0024] 4. Other filtering criteria: Other screening criteria include: cargo type (frozen goods / fresh meat / food and beverages / vegetables, etc.), availability of price, price per kilometer (below 3 / 3-4 / 4-5 / above 5, etc.), weight (0-4 tons / 4-6 tons / 7-8 tons / above 8 tons, etc.), volume (0-7 cubic meters / 7-20 cubic meters / 20-30 cubic meters / above 30 cubic meters, etc.), negotiable price (fixed price / negotiable price / negotiable price), and full truckload and less-than-truckload (LTL).

[0025] Precision: Strictly meet the driver screening criteria and retain explicit user selection.

[0026] Generalization: Removes restrictions on screening criteria, relaxing limitations to improve recall coverage.

[0027] Part Two: Design for Cargo Source Retrieval: Based on four dimensions of cargo source conditions, cargo matching the conditions is retrieved from the cargo source pool (Cargo Source ES Index) eight times in total, representing four dimensions × precision / generalization. The Cargo Source ES Index is an independently built engineering capability for cold chain logistics, storing all available cargo sources in an ES inverted index, supporting retrieval fields such as: origin, destination, vehicle length, cargo type, price, weight, volume, negotiable type, full truckload and less-than-truckload, etc. When information is updated for each shipment, the system listens for cargo source update Kafka messages and updates the index accordingly.

[0028] Part Three: Design for Recall Routes: After the cargo retrieval, each shipment, based on the retrieval criteria: precise / generalized × origin / destination / vehicle length / other, has a total of 2^4 = 16 tag combinations, which are categorized into 16 recall routes. The tags of the recall routes are progressively passed down to the downstream sorting and reordering stages for use in reordering strategies and list floor segmentation rendering.

[0029] This invention achieves unified modeling and flexible configuration of recall logic through this matrix design, supports independent expansion and weight adjustment of each dimension, significantly reduces the complexity of multi-path parallel recall in the original system, and improves code readability and maintenance efficiency.

[0030] Practical applications show that after the invention was launched in the platform's cold chain logistics zone, the cargo recall rate increased from 88% to 95%, effectively expanding the pool of high-quality cargo visible to drivers; at the same time, it boosted the number of cold chain logistics fulfillment orders by 0.5%, significantly optimizing the platform's matching efficiency and user experience.

[0031] Figure 2This diagram illustrates a more specific hardware structure of an electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0032] It should be noted that, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the device described above may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0033] The electronic devices described above are used to implement a corresponding intelligent retrieval method for goods based on a multidimensional condition matrix in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0034] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, which are used to cause the computer to execute a resource intelligent recall method based on a multi-dimensional condition matrix as described in any of the above embodiments.

[0035] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0036] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute a resource intelligent recall method based on a multidimensional condition matrix as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0037] Based on the same inventive concept, corresponding to any of the above-described embodiments, this application also provides a computer program product, which includes a computer program. In some embodiments, the computer program instructions can be executed by one or more processors of a computer to cause the computer and / or the processor to perform the intelligent retrieval method for goods based on a multidimensional condition matrix described in the above embodiments. Corresponding to the execution entity for each step in each embodiment of the intelligent retrieval method for goods based on a multidimensional condition matrix, the processor executing the corresponding step can belong to the corresponding execution entity.

[0038] To simplify the description and discussion, and to avoid making the embodiments of this application difficult to understand, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings.

[0039] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. A method for intelligent sourcing retrieval based on a multidimensional condition matrix, characterized in that, Includes the following steps: Construct a condition matrix based on driver search criteria, and independently construct precise / generalized conditions for each dimension of the search criteria; Based on all dimensions of the source conditions, the source is searched, and based on the precise / generalized conditions, the recall and regression classes are performed. The tags of the recall loop are passed down level by level to the downstream fine ranking and re-ranking stages for use by the re-ranking strategy and list floor segmentation rendering.

2. The intelligent sourcing recall method based on a multidimensional condition matrix according to claim 1, characterized in that: When constructing the condition matrix, the search conditions include four dimensions: origin, destination, vehicle length, and other filtering conditions. Among them, other filtering conditions include: cargo type, availability, price per kilometer, weight, volume, negotiable type, and full truckload and less-than-truckload (LTL) shipments. When constructing precise / generalized conditions, for the aforementioned starting point: Precise condition representation: Strictly matches search conditions, supplemented by basic geographic generalization; Generalized conditional representation: Introduces multi-source extension logic, including advanced geographic proximity, behavioral preferences, computational intent inference, reported intent, relay points along the search route and origin of the return journey, and dynamic contextual information; When constructing precise / generalized conditions, for the stated destination: Precise condition representation: Strictly matches search conditions, supplemented by basic geographic generalization; Generalization condition representation: Introduces multi-source extension logic, including advanced geographic proximity, behavioral preferences, computational intent inference, reported intent, search route relays, and return destination; When constructing precise / generalized conditions, for the vehicle length: Precise criteria: Driver search vehicle length and registered vehicle length; Generalization conditional representation: driver calculates vehicle length based on intent, vehicle length based on behavior, and vehicle length based on reported intent; When constructing precise / generalized criteria, for other filtering conditions: Precise condition representation: Meeting the driver's screening criteria retains the user's explicit selection; Generalized conditions mean: no restrictions on screening conditions, relaxing the restrictions to improve recall coverage.

3. The intelligent sourcing recall method based on a multidimensional condition matrix according to claim 2, characterized in that: When performing cargo retrieval, cargo matching the criteria is retrieved from the cargo pool based on four dimensions of cargo conditions, for a total of eight retrievals. Among these, cargo retrieval is performed through the cargo ES index, which stores all available cargo in an inverted index. Supported retrieval fields include: origin, destination, vehicle length, cargo type, price, weight, volume, negotiable type, and full truckload and less-than-truckload. When there is an information update for each shipment, the Kafka message for cargo update is listened to, and the index content is updated accordingly.

4. The intelligent sourcing recall method based on a multidimensional condition matrix according to claim 3, characterized in that: When classifying recalls, after searching for the source of goods, each shipment is classified into 16 recall routes based on the following search criteria: precise / generalized × origin / destination / vehicle length / other.

5. A source retrieval intelligent system based on a multidimensional condition matrix, used to implement the source retrieval intelligent method based on a multidimensional condition matrix as described in any one of claims 1 to 4, characterized in that, include: The condition construction module is used to construct a condition matrix based on driver search conditions, and to independently construct precise / generalized conditions for each dimension of the search conditions. The recall module is used to search for goods based on all dimensions of goods conditions, and to perform recall regression based on precise / generalized conditions. The tags of the recall loop are passed down level by level to the downstream fine ranking and re-ranking stages for use by the re-ranking strategy and list floor segmentation rendering.

6. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing a computer to perform the method of any one of claims 1 to 4.

8. A computer program product, characterized in that, It includes computer program instructions that, when run on a computer, cause the computer to perform the method as described in any one of claims 1 to 4.

Images