CN111507790A

CN111507790A - Processing method of clothing matching information, data object processing method, system and equipment

Info

Publication number: CN111507790A
Application number: CN201910101037.6A
Authority: CN
Inventors: 曹阳; 章人可; 戴能
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2019-01-31
Filing date: 2019-01-31
Publication date: 2020-08-07
Also published as: WO2020156306A1

Abstract

The embodiment of the application provides a clothing matching information processing method, a data object processing method, a system and equipment. The processing method of the clothing matching information comprises the following steps: acquiring at least one type of data object set different from the type of the first data object; extracting a second data object from the set of data objects of the at least one type to be combined with the first data object to generate a plurality of data object combinations; and selecting the data object combination which accords with the matching rule from the plurality of data object combinations as a clothing matching example which can be referred by the user. The technical scheme provided by the embodiment of the application has strong universality, and the produced collocation has richer property on the premise of ensuring the collocation quality and has better collocation satisfaction.

Description

Processing method of clothing matching information, data object processing method, system and equipment

Technical Field

The present application relates to the field of computer technologies, and in particular, to a method, a system, and a device for processing clothing matching information.

Background

For a long time, people are always used to select the commodities suitable for themselves by repeatedly selecting the commodities of shoes, clothes and bags which are satisfied by Linglan and have various styles and colors. Most of the ordinary users do not know how to match the goods, such as the dress and the shoes, and the style of the bag.

At present, the common logic of recommending commodities by an e-commerce platform is based on a recommendation technology of user behaviors, and most of the results of recommending commodities to users belong to the same category as the commodities purchased, collected or browsed. For example, a user may be interested in a dress, frequently browsing and clicking on the dress at the platform, and may present a variety of similar dresses in the recommended traffic window. The recommendation mode brings repeated recommendation results, the user obviously purchases commodities, the system also repeatedly recommends similar commodities, and the system is easy to cause trouble of the user and poor in user experience.

Disclosure of Invention

Embodiments of the present application provide a method, a system, and an apparatus for processing clothing matching information, a method, a system, and an apparatus for processing data objects, which solve or partially solve the above problems.

In one embodiment of the application, a clothing matching information processing method is provided. The method comprises the following steps:

acquiring at least one type of data object set different from the type of the first data object;

extracting a second data object from the set of data objects of the at least one type to be combined with the first data object to generate a plurality of data object combinations;

and selecting the data object combination which accords with the matching rule from the plurality of data object combinations as a clothing matching example which can be referred by the user.

In another embodiment of the present application, a data object processing method is provided. The method comprises the following steps:

when a specified event related to a first data object occurs, acquiring at least one type of second data object different from the type of the first data object;

determining whether the at least one second data object is collocated with the first data object according to the attribute of the first data object and the attribute of the at least one second data object;

when the at least one second data object is matched with the first data object, the at least one second data object is provided for the user.

In yet another embodiment of the present application, a method for processing clothing matching information is provided. The method comprises the following steps:

responding to the operation of a user on the first data object, and sending request information to a server;

receiving a clothing matching example fed back by the server aiming at the first data object;

providing the apparel collocation instance to the user;

the clothing matching example is a data object combination which is selected from a plurality of data object combinations and accords with matching rules; the plurality of data object combinations are generated by combining a second data object extracted from a set of data objects of at least one type different from the type to which the first data object belongs with the first data object.

acquiring request information sent by a client aiming at the first data object;

extracting a second data object to be combined with the first data object from a set of data objects of at least one type different from a type to which the first data object belongs to generate a plurality of data object combinations;

and feeding back the data object combination which accords with the collocation rule in the plurality of data object combinations to the client as a clothing collocation example.

In one embodiment of the present application, a system for processing clothing matching information is provided. The system comprises:

the client is used for responding to the appointed operation of a user for the first data object and sending request information corresponding to the appointed operation to the server; receiving a clothing matching example fed back by the server aiming at the first data object; providing the apparel collocation instance to the user;

the server is used for acquiring request information sent by the client aiming at the first data object; extracting a second data object to be combined with the first data object from a set of data objects of at least one type different from a type to which the first data object belongs to generate a plurality of data object combinations; and feeding back the data object combination which accords with the collocation rule in the plurality of data object combinations to the client as a clothing collocation example.

In one embodiment of the present application, an electronic device is provided. The electronic device comprises a memory and a processor; wherein,

the memory is used for storing programs;

the processor, coupled with the memory, to execute the program stored in the memory to:

In another embodiment of the present application, an electronic device is provided. The electronic device comprises a memory and a processor; wherein,

the memory is used for storing programs;

acquiring at least one type of second data object different from the type of the first data object;

In one embodiment of the present application, a client device is provided. The client device comprises a memory and a processor; wherein,

the memory is used for storing programs;

providing the apparel collocation instance to the user;

In one embodiment of the present application, a server device is provided. The server-side equipment comprises a memory and a processor; wherein,

the memory is used for storing programs;

acquiring request information sent by a client aiming at the first data object;

According to the technical scheme provided by the embodiment of the application, the second data object used for being combined with the first data object is extracted from the data object set of at least one type different from the type of the first data object to generate a plurality of data object combinations, and then a clothing matching example which can be referred by a user is selected from the plurality of data object combinations by using a matching rule instead of a strict matching template, so that the universality is high, and the matching produced on the premise of ensuring the matching quality is richer; in addition, through actual effect measurement, the technical scheme provided by the embodiment of the application can output multiple sets of matching suggestions for 99% of the commodities in the target commodity pool, and the satisfaction rate of manual evaluation reaches 80%.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a schematic flow chart of a processing method of clothing matching information according to an embodiment of the present application;

fig. 2 is a data object flow diagram schematically illustrating a method for processing clothing matching information by using a one-piece dress as an example according to an embodiment of the present application;

FIG. 3 is a schematic diagram of an embodiment of the present application;

FIG. 4 is a table representing a style collocation relationship according to an embodiment of the present application;

fig. 5 is a schematic flowchart of a data object processing method according to another embodiment of the present application;

FIG. 6 is a schematic structural diagram of a system for processing clothing matching information according to an embodiment of the present application;

fig. 7 is a schematic flowchart of a processing method of clothing matching information according to another embodiment of the present application;

fig. 8 is a schematic flowchart of a processing method of clothing matching information according to another embodiment of the present application;

fig. 9 is a schematic structural diagram of a device for processing clothing matching information according to an embodiment of the present application;

FIG. 10 is a schematic structural diagram of a data object processing apparatus according to another embodiment of the present application;

fig. 11 is a schematic structural diagram of a device for processing clothing matching information according to another embodiment of the present application;

fig. 12 is a schematic structural diagram of a device for processing clothing matching information according to another embodiment of the present application;

fig. 13 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

In some of the flows described in the specification, claims, and above-described figures of the present application, a number of operations are included that occur in a particular order, which operations may be performed out of order or in parallel as they occur herein. The sequence numbers of the operations, e.g., 101, 102, etc., are used merely to distinguish between the various operations, and do not represent any order of execution per se. Additionally, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first", "second", etc. in this document are used for distinguishing different messages, devices, modules, etc., and do not represent a sequential order, nor limit the types of "first" and "second" to be different. In addition, the following embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

The prior art generally adopts the following three schemes to realize the clothing matching recommendation. The following will be explained separately:

first, based on manual editing

Based on the manual editing scheme, the quality is high, but the collocation quantity is very small. The quality is high by manually making and matching groups such as starting experts and dawners, but the daily output is usually hundreds of thousands of orders and the quantity is very small due to the manual efficiency. Platform goods are in the order of hundreds of millions, so that a large number of goods cannot be suggested by manual matching. The matching irrelevant to the purchased goods of the user cannot play a role for the user.

Second, purchase based on associations

The quality is general and the quantity is large based on the scheme of the associated purchase. The collocation samples are formed by mining the commodities which are jointly purchased by the users in a period of time window, and the models are trained to automatically produce the collocation. The method has the problem that the noise of the training sample is too large, because the clothing commodity purchased by a user in a period of time is not guaranteed to be purchased for the same person, the quality of the training sample is not high (or the cost for processing the sample to meet the matching quality requirement is too high), and the model effect is limited; the method is only suitable for making related commodity recommendations.

Third, based on templates

The scheme based on the matching template has the advantages of high quality, large quantity under an open commodity set and small quantity under a limited commodity set. Aiming at the problem that the commodities in the first scheme are few, manual collocation is used as a template, a single product is used for finding similar functions, and a set of collocation is expanded into a plurality of sets of similar collocation. However, platform-capable (manual) collocation templates are typically in the order of millions or less, and the number of goods that can be met and the collocation formats that can be covered is still limited compared to tens of millions or even billions of platform goods (the number of collocations that can be combined in an array is nearly infinite). The richness requirement of a matching template on a commodity is very high. For example, a 4-piece matching template means that a match qualified under the template is formed only when commodities of 4 parts of a specific upper garment, a specific lower garment, a specific shoe and a specific bag are required to meet similar conditions. Therefore, especially when a match is desired to be produced under a limited commodity set (for example, a certain brand and a certain shop), due to limited candidate commodities, the match production fails because the requirements that all parts of a specific template are similar are not satisfied enough. In summary, the template method still has limited production matching and cannot cover the matching requirements of all the commodities on the platform.

In order to provide richer matching results with good satisfaction for users, the technical challenges to be overcome are as follows:

A. computational complexity

One set of matching usually consists of 3-5 part commodities. For e-commerce platforms, apparel, shoes, and bags typically cover more than 40 types, with up to ten million goods candidates for each type. Therefore, a matched commodity candidate search space is huge, and the commodity permutation and combination which can be used as combination candidates is not exhaustive.

B. Coverage of goods

On the premise that there is a sufficient candidate commodity set, if the method cannot produce collocation for a certain commodity, the chance of showing collocation and commodity conversion by a user is missed once. Therefore, for the clothing goods with huge platforms, the coverage rate of the method should reach more than 95%.

C. Quality of collocation

The user likewise does not continue to browse and click if the collocation advice surrounding the user's request for merchandise display does not meet basic apparel fit aesthetic requirements. Therefore, methods are required to ensure coverage and richness of output collocation, and at the same time, higher qualification rate is required.

D. Efficiency of implementation

For each piece of mass clothing goods of the e-commerce platform to provide matching service, real-time computing capability must be provided, and computing results are not cached in advance (the enumeration number is too large and the cost is too high). Therefore, the calculation time required for returning 10 sets and above for a single commodity is less than 200 ms.

Fig. 1 is a schematic flow chart illustrating a method for processing clothing matching information according to an embodiment of the present application. The execution main body of the solution provided in this embodiment may be a data object processing device, and the device may be a hardware integrated on a client or a server and having an embedded program, may also be an application software installed in the client or the server, and may also be a tool software embedded in an operating system of the client or the server, which is not limited in this embodiment of the present application. The client may be any terminal device including a mobile phone, a tablet computer, a PDA (Personal Digital Assistant), a POS (Point of Sales), a vehicle-mounted computer, and the like. The server can be a server, a cloud and the like; this embodiment is not particularly limited thereto. As shown in fig. 1, the method includes:

101. a set of data objects of at least one type different from the type to which the first data object belongs is obtained.

102. A second data object to be combined with the first data object is extracted from the set of data objects of the at least one type to generate a plurality of data object combinations.

103. And selecting the data object combination which accords with the matching rule from the plurality of data object combinations as a clothing matching example which can be referred by the user.

Under the scene of the costume type electric market, the first data object and the second data object are specific costume type commodities such as one-piece dress, coats, shoes, bags and the like. The first data object, the second data object may be a commodity number or a commodity name or the like for uniquely identifying the commodity. For the client, the above steps may be triggered to be executed after the user performs any one of the following actions:

the user clicks a control requesting to view the clothing matching example of the commodity A;

the user sends out a voice requesting to recommend a clothing matching example for the commodity A;

a user adds the commodity A into the shopping cart;

the user pays attention to the commodity A;

the user purchased commodity a; and so on.

The commodity a, which is the object of the user behavior, is the first data object in this embodiment.

For the server, the server may perform the above steps after receiving a collocation request sent by the client for the first data object.

The embodiment can be applied to a plurality of application scenes, such as costume type electric market scenes. The type to which the first data object belongs is the type of the article, such as footwear, clothing, bags, accessories (such as necklaces, headwear, earrings, rings, etc.), and so on. The clothes are divided into the following parts according to different wearing positions: top-loading type, bottom-loading type, conjoined type and the like. The conjoined class can be divided into: dresses, jumpsuits, and the like. Similarly, the ornaments can be classified into the following items according to wearing positions: necklace type, headwear type, earring type, ring type, etc.

Assuming that in the clothing type e-market scene, the first data object is clothing information, the type of the clothing information is one-piece dress, and accordingly, at least one of the following types of data object sets can be obtained:

a set of footwear data objects, a set of jewelry type data objects, a set of top-loading type data objects, a set of bag type data objects, a set of bottom-loading type data objects, and the like.

For example, a set of data objects for footwear and a set of data objects for a bag class are obtained. In specific implementation, the data objects (i.e., commodities) included in the various data object sets may be hot or newly shelved commodities. In an implementation technical solution, the technical solution provided in this embodiment may further include the following steps:

collecting data objects of the same type;

sequencing the collected data objects according to a sequencing rule;

and selecting a plurality of data objects arranged in front to obtain the data object set of the type.

For example, the ordering rule may specifically be: and (4) reversing the arrangement according to the popularity and the commodity shelf-loading time. The number of data objects to be selected may be set manually, for example, 100, 150, 200 or more, which is not limited in this embodiment.

The type of data object (e.g., footwear, clothing, etc.), and the sort parameters (e.g., popularity, time to put on shelf, etc.) may be recorded in the attributes of the data object. That is, taking a data object set of a first type different from the type to which the first data object belongs as an example, a process of acquiring the data object set of the first type is described, which specifically includes:

acquiring the attribute of each data object in the first type data object pool;

sorting the data objects in the first type data object pool according to the attribute of each data object in the first type data object pool;

and selecting a plurality of data objects which are ranked at the front to form the data object set of the first type.

Wherein the attributes of the data object may include: type, style, color information, texture information, sales records, purchase rate, click rate, time on shelf, season, occasion, style, and the like. The popularity score can be calculated through sales records, purchase rate, click rate and the like.

In 102, when there is a type of data object set, the second data object can be directly extracted from the data object set and combined with the first data object to obtain a plurality of data object combinations. For example, the first data object is a one-piece dress, step 101 obtains a data object set of a piece of footwear, the data object set of the piece of footwear contains 200 data objects, and 200 data objects contained in the data object set can be combined with the one-piece dress respectively to obtain 200 data object combinations.

When there are two or more sets of data objects, a second data object may be extracted from some or all of the sets of data objects and combined with the first data object to obtain a plurality of data object combinations. For example, where the first data object is a dress, step 101 obtains a set of data objects for the footwear, a set of bag-type data objects, and a set of coat-type data objects. In particular, a data object a may be extracted as a second data object from the footwear data object set, a data object b may be extracted as a second data object from the package data object set, and then the data object a, the data object b, and the first data object may be combined into a data object combination; alternatively, a data object c may be extracted as a second data object in the footwear data object set, a data object d may be extracted as a second data object from the bag data object set, a data object e may be extracted as a second data object from the coat data object set, and then data object c, data object d, and data object e may be combined with the first data object into a data object combination; and so on.

Here, it should be noted that: the way how to extract the data object from the data object set of at least one type to generate the data object combination is not specifically limited in this embodiment, as long as it is ensured that the types of the data objects in all the data objects included in the data object combination are different, and the data object combinations are all different.

In the above 103, the collocation rule may include one or more rules, which is not specifically limited in this embodiment. The setting of the collocation rules may be different in different application scenarios. For example, in the shoe and clothing package shopping mall, the matching rules may include: rules for style, rules for color, rules for highlights, etc. The collocation rules may be based on expert knowledge. Specifically, the rule for style is used to filter out the data object combinations with different composition structures, seasonal temperatures, and the like, such as: data object combinations of "T-shirt" + "Down pants". Rules for color and rules for bright spots may be used to filter out data object combinations that do not meet aesthetic requirements; wherein aesthetic requirements may be expert or up to the default dressing rules, such as no more than 3 dominant colors, no more than 1 bright spot, etc. throughout the body.

In addition, under the condition that a plurality of obtained clothing matching examples which can be referred by the user are assumed, the method provided by the embodiment further comprises the following steps:

sequencing the multiple clothing matching examples;

and providing a plurality of clothing matching examples for the user according to the sequencing result.

The steps are added to further optimize the click rate of the collocation. In specific implementation, the popularity score (which may be sales volume, click rate, purchase rate, etc., and may be adjusted according to different business objectives) of the goods contained in each clothing matching example may be obtained, and the score of each clothing matching example is calculated. For example, the average value of the popularity scores of the data objects other than the first data object in the included commodities is used as the score obtained by the final calculation. Then, according to the size of the score, the clothing matching examples are ranked in a mode that the score is large and the score is small and the clothing matching examples are ranked in the front. Finally, the user may be presented with the scores sorted from high to low.

According to the technical scheme provided by the embodiment, the second data object used for being combined with the first data object is extracted from the data object set of at least one type different from the type of the first data object to generate a plurality of data object combinations, and then a clothing matching example which can be referred by a user is selected from the plurality of data object combinations by using a matching rule instead of a strict matching template, so that the universality is high, and the matching produced on the premise of ensuring the matching quality is richer; the method can meet the actual matching requirements of users and is beneficial to improving the commodity conversion rate of the e-commerce platform; in addition, through actual effect measurement, the technical scheme provided by the embodiment of the application can output multiple sets of matching suggestions for 99% of the commodities in the target commodity pool, and the satisfaction rate of manual evaluation reaches 80%.

In an implementation solution, this embodiment provides that step 102 in the method is: "extracting a second data object from said set of at least one type of data objects for combination with said first data object to generate a plurality of data object combinations" may be performed by:

1021. acquiring at least one combination rule;

1022. selecting one data object from a part of or all of the data object sets of the at least one type of data object set as a data object combination with the first data object, respectively, using the at least one combination rule; until the number of different generated data object combinations meets the preset number requirement.

At 1021, at least one combination rule is obtained based on the type of the first data object. As shown in FIGS. 2 and 3, assuming that the first data object is of the dress type, the at least one obtained composition rule may compriseComprises the following steps: combination rule 1 combined with shoes and bags, combination rule 2 combined with coats, shoes and bags, combination rule 3 combined with trousers, shoes and bags, combination rule 4 combined with coats, trousers, shoes and bags, and the like. Wherein, the symbols in FIG. 3

The characterization may not be selected. Referring to fig. 3, assuming that the type of the first data object is a jacket class, the at least one combination rule obtained may include: combination rule 1 ' in combination with pants, shoes and bags, combination rule 2 ' in combination with a bust, shoes and bags, combination rule 3 ' in combination with a coat, bust, shoes and bags, and the like.

In practical application, the combination rule corresponding to each type may be set in advance. Therefore, when a user purchases, checks or applies for a collocation request for a certain type of data object, the user can directly call the combination rule corresponding to the type of the data object.

The following includes with the at least one combination rule: the above step 1022 is explained by taking the first combination rule and the second combination rule as examples. Typically, two or more types of data object sets or more than two types of data object sets require two or more combination rules. That is, when at least two types of data object sets are acquired in step 101, step 1022 may specifically be:

1022', using the first combination rule and the second combination rule, selecting one data object from a part of data object sets or all data object sets of at least two types of data object sets as a second data object to be combined with the first data object into a data object combination; until the number of different generated data object combinations meets the preset number requirement.

Specifically, the step 1022' can be implemented by the following steps:

s1, acquiring current sequencing serial numbers Ni corresponding to the data object sets of the types in the data object sets of the at least two types, wherein the values of i of the data object sets of different types are different;

s2, respectively acquiring data objects ordered as Ni from all data object sets specified by the first combination rule as a second data object to be combined with the first data object to form a data object combination, and updating Ni corresponding to all the data object sets specified by the first combination rule to Ni + 1;

s3, respectively acquiring data objects ordered as Ni from all data object sets specified by the second combination rule as a second data object to be combined with the first data object to form a data object combination, and updating Ni corresponding to all the data object sets specified by the second combination rule to Ni + 1;

and S4, repeating the steps until the number of the generated different combination examples meets the preset number requirement.

The above steps S1 to S4 can be simply understood as the following processes:

referring to the example shown in fig. 2, data object set 3 for footwear corresponds to N3 of 1, data object set 4 for bags corresponds to N4 of 1, and data object set 1 for outer bags corresponds to N1 of 1;

firstly, acquiring Ni of each data object set; n3 of the data object set 3 is 1, N4 of the data object set 4 is 1, and N1 of the data object set 1 is 1.

Then, acquiring commodities with the ranking of N3 being 1 from the data object set 3 specified by the first rule; acquiring commodities ordered with N4 being 1 from a data object set 4 specified by a first rule, and forming a data object combination (31, 41); and updating N3 to 1+ 1-2; n4 is updated to 1+1 ═ 2;

subsequently, the commodity with the ranking of N3 ═ 2 is acquired from the data object set 3 specified by the second rule, and the commodity with the ranking of N4 ═ 2 is acquired from the data object set 4 specified by the second rule; acquiring commodities ordered with N1 being 1 from the data object set 1 specified by the second rule, and forming a data object combination (11, 32 and 42); and N3, N4 and N1 are all added with +1 on the existing basis, namely N3 is updated to 3, N4 is updated to 3, and N1 is updated to 2.

When only two combination rules are described herein, the process of generating different combination instances. Principle types when there are more combination rules, such as third combination rule, fourth combination rule, … …, etc. An example including 4 combination rules is shown in fig. 2 below. The process of generating a plurality of combined instances may be embodied as:

using a combination rule 1, extracting a commodity 31 with a sequence number N of 1 from a data object set 3, extracting a commodity 41 with a sequence number N of 1 from a data object set 4, and combining the dress 00, the commodity 31 and the commodity 41 which are purchased, browsed or requested to be matched by a user into a 3-piece data object combination [00, 31,41 ];

using a combination rule 2, extracting a commodity 32 with a sequence number N + 1-2 from a data object set 3, extracting a commodity 42 with a sequence number N + 1-2 from a data object set 4, and extracting a commodity 11 with a sequence number N-1 from a data object set 1, and combining the one-piece dress 00, the commodity 32, the commodity 42, and the commodity 11 into one 4-piece data object set [00, 32, 42, 11 ];

using a combination rule 3, extracting a commodity 33 with a sequence number N + 2-3 from the data object set 3, extracting a commodity 43 with a sequence number N + 2-3 from the data object set 4, and extracting a commodity 21 with a sequence number N-1 from the data object set 2, and combining the one-piece dress 00, the commodity 33, the commodity 43, and the commodity 21 into one 4-piece data object set [00, 33, 43, 21 ];

using the combination rule 4, the item 34 with the ranking number N +3 of 4 is extracted from the data object set 3, the item 44 with the ranking number N +3 of 4 is extracted from the data object set 4, the item 12 with the ranking number N +1 of 2 is extracted from the data object set 1, and the item 22 with the ranking number N +1 of 2 is extracted from the data object set 2, and the one-piece dress 00, the item 34, the item 44, the item 12, and the item 22 are combined into one 4-piece data object set [00, 34, 44, 12, 22 ].

And updating the sequencing serial number N to be N +1, and repeating the steps until the number of the generated different combination examples meets the requirement of the preset number.

In a specific implementation, the preset number may be 100, 200, … or more, and this embodiment is not limited in this respect.

In the above example, the set of data objects specified by the composition rule 1 includes: a set of data objects 3 and a set of data objects 4; the set of data objects specified by the composition rule 2 includes: a set of data objects 1, a set of data objects 3 and a set of data objects 4; the set of data objects specified by the composition rule 3 includes: a set of data objects 2, a set of data objects 3 and a set of data objects 4; the set of data objects specified by the composition rule 4 includes: a set of data objects 1, a set of data objects 2, a set of data objects 3 and a set of data objects 4. Here, it should be noted that: in particular implementations, the set of data objects specified by the composition rule may be one or more; the data object sets designated by each of the plurality of combination rules used simultaneously are different, but there may be partially identical data object sets. In addition, the above only shows a generation process of a combination example, and in fact, a manner of extracting data objects from each data object set in what order to combine with the first data object is adopted, and this embodiment is not particularly limited as long as the finally obtained data object combinations meeting the quantity requirement are different, and the data object combinations are all different.

Further, in the method provided by this embodiment, step 103: "select the data object combination which accords with the matching rule from the plurality of data object combinations as the clothing matching example which can be referred by the user", which can be realized by adopting the following method:

1031. and acquiring the attribute of the data object contained in each data object combination in the plurality of data object combinations.

1032. And using a filtering rule aiming at least one attribute item in the attributes to judge whether the attribute combination of the data object contained in each data object combination in the plurality of data object combinations conforms to the filtering rule.

1033. And deleting the data object combinations of which the attribute combinations of the data objects in the plurality of data object combinations meet the filtering rule.

In a specific application scenario, the attribute of the data object includes at least one of the following attribute items: type, style, color information, texture information, sales records, purchase rate, click rate, time on shelf. Here, it should be noted that: the type, style and the like in the attribute of the data object can be obtained from information filled by a merchant; the color information and the texture information can be obtained through corresponding algorithm identification, and the sales record, the purchase rate, the click rate and the shelf-loading time can be obtained from the server. The attribute obtaining method of the data object is not specifically limited in this embodiment, and can be obtained by using a corresponding technique in the prior art.

The filtering rules may include, but are not limited to: rules for style, rules for color, rules for highlights, etc.

When the filtering rule includes: for the rules of style, the step 1032 may include: using a style-specific rule, it is determined whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the style-specific rule. Further, the step of "determining whether the attribute combination of the data object included in the first data object combination of the plurality of data object combinations conforms to the rule for style" using the rule for style may specifically be:

obtaining the style of the first data object contained in the first data object combination and the style of at least one second data object combined with the first data object;

determining whether the style combination of the first data object and the style combination of each of the at least one second data object meets the style-specific rule.

Correspondingly, step 1033, "delete the data object combination whose attribute combination of the data object included in the plurality of data object combinations meets the filtering rule" includes:

and deleting the first data object combination when the style of the first data object and the style combination of any second data object in the at least one second data object accord with the rule aiming at the style.

When the filtering rule includes: for the rule of color, the step 1032 may include: using the rules for color, determining whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the rules for style. Further, the step of determining whether the attribute combination of the data object included in the first data object combination of the plurality of data object combinations conforms to the style-specific rule using the color-specific rule may specifically include:

acquiring color information of the first data object contained in the first data object combination and color information of at least one second data object combined with the first data object;

determining the number of dominant colors satisfying a dominant color condition based on the color information of the first data object and the color information of the at least one second data object;

determining whether the dominant color number complies with the color-specific rule.

and when the number of the dominant colors is larger than a first threshold value, deleting the first data object combination.

When the filtering rule includes: when the bright point is specified, the step 1032 may include: using a rule for a highlight, determining whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the rule for a highlight. Further, the step of determining whether the attribute combination of the data object included in the first data object combination of the plurality of data object combinations conforms to the rule for the bright point using the rule for the bright point may specifically include:

acquiring color information and texture information of the first data object contained in the first data object combination and color information and texture information of at least one second data object combined with the first data object;

performing a bright spot analysis on the color information of the first data object and the color information of the at least one second data object to obtain a first analysis result related to a bright spot;

performing a highlight analysis on the texture information of the first data object and the texture information of the at least one collocated data object to obtain a second analysis result related to a highlight;

determining whether the first analysis result and the second analysis result meet the rule for the bright spot.

and deleting the first data object combination when the bright spot number obtained according to the first analysis result and the second analysis result is larger than a second threshold value.

Based on the above, the technical solution provided by this embodiment can be simply summarized as follows: data object combination construction-filtering-sorting process. For example, in the shoe and clothing package electronic market scene, the clothing expert or the clothing wearing knowledge of the person who wears the shoes can be modeled into the execution logic in the 3 links, so that the universal collocation automatic calculation and output are realized. The combination rule is introduced into the data object combination construction part, potential clothing matching example candidates are quickly constructed in a huge commodity arrangement combination space, and the problem of computational complexity is solved. In the filtering part, a filtering rule is adopted to accurately filter out dress collocation examples meeting the requirements of beauty and dignity in daily wearing and taking in a plurality of data object combinations, and the problem of collocation quality is solved. Finally, a search framework of online recall and filtering sequencing can be used to support rapid operation and return results, so that the real-time requirement is met; through experimental tests, the matched commodity coverage rate exceeds 99% and the matching satisfaction rate exceeds 80% on a commodity set with the quantity exceeding 200 and the categories being balanced.

Data object composition construction part

In the data object combination construction part, the combination rule precipitated by clothing experts or acquaintances is used for guiding the commodity requested by the user to be combined with the commodity of which types to form the data object combination. As an example shown in fig. 2, the user requests a dress. As shown in figure 3, the dress can be selected to be a hollow coat at the upper dress part and hollow trousers at the lower dress part, and the shoe and bag parts are necessary. Then a total of 2X2 ═ 4 rules can be chosen, namely dress + shoe + bag combination rule 1, dress + coat + shoe + bag combination rule 2, dress + trousers + shoe + bag combination rule 3, dress + coat + trousers + shoe + bag combination rule 4. Through index control, the 4 commodities (coats, trousers, shoes and bags) in the candidate commodity pool can be called out from the storage module, for example, 200 commodities of various types can be used as a data object set of each type of commodity to generate collocation candidates.

When 200 commodities of a certain type are called out from the candidate commodity pool, for example, shoes can be arranged reversely according to the baby score plus the shelf-on time to form a data object set 3 of the shoes, and hot commodities and newer commodities can be preferentially participated in target commodities (such as one-piece dress).

Next, the 4 combination rules shown in FIG. 2 may be used in turn to yield a data object combination. For example, firstly, a rule of '+ shoes + bags' (namely a combination rule 1) is used, the first commodity in each shoe and bag queue is used, and the existing one-piece dress commodities of the user are combined to form a (one-piece dress + shoes + bags) 3-piece suit candidate; then, using a second "+ coat + shoe + bag" rule (i.e., combination rule 2), the first item is taken from the coat queue, and the second items are each taken from the shoe and bag queue, to form (the same dress but with a coat, new shoe and bag) 4-piece suit candidates, which are passed on to subsequent modules. And so on, stop when the number requirement (say 200) is satisfied by the candidate produced.

Filtration section

The filtering portion may use filtering rules to filter out objectionable, aesthetically unappealing data object combinations in color texture. In particular implementations, the filtering rules may include, but are not limited to, at least one of: filtering rules for styles, filtering rules for colors, filtering rules for highlights, etc.

In specific implementation, the filtering rule for the style can be simply understood as the style collocation relationship represented in the table shown in fig. 4. Referring to the tabular form shown in FIG. 4, which is composed of a combination of styles deposited by experts or darts that cannot be worn together at the same time, 2000 or more boxes can be generatedPoint; the collocation of some versions is shown in fig. 4 by way of example only. In the lattice point

And

shows a style and a lattice point which can be worn together at the same time

Indicating a style that is not suitable for wearing at the same time, such as "sweater" + "suit", "T-shirt" + "down pants", etc. In essence, the filtering rules for style serve to filter out data object combinations that do not meet the graceful requirements in composition structure and seasonal temperature.

The data object combination is usually 3-5 pieces, every two items of commodities contained in the data object combination are subjected to table lookup, and if the combination does not meet the filtering rule aiming at the style, the data object combination needs to be deleted. For example, a certain combination of data objects includes: coats, tops and trousers; correspondingly, whether the coat and the upper garment accord with the filtering rules for the style, whether the upper garment and the trousers accord with the filtering rules for the style, and whether the coat and the trousers accord with the filtering rules for the style are inquired. If a matched combination of the coat and the jacket, the coat and the pants and the coat and the pants belongs to a combination which is not suitable for being worn at the same time through inquiry, the data object combination needs to be filtered, namely, the data object combination is deleted from a plurality of data object combinations.

Generally, the color information of each commodity has 1-3 colors, and the texture information has 1-3 texture labels, as follows:

product A:

color information: (Color1, AreaC1), (Color2, AreaC2) …

Texture information: (Texture1, area 1), (Texture2, area 2) …

Wherein Color is a Color tag, such as carmine, and Area is the Area percentage of the Color tag in the product, such as 40%; similarly, Texture is a Texture tag, such as a bump, and the corresponding Area is the Area fraction of the Texture tag in the product, such as 70%.

The filtering rules for the colors may specifically be: more than 3 dominant colors in a data object combination need to be filtered out. The condition that a color is defined as a dominant color is that the Area of the color label > is 30%, and the S > of the color in the HSV space is 0.14 (otherwise, black, white and gray are in the black-white-gray color system, and black, white and gray do not occupy the dominant color number). The rule is executed by sequentially traversing each commodity in a data object combination, counting the number of dominant colors, and when the number of dominant colors of different types is greater than 3, the data object combination needs to be filtered, namely, the data object combination is deleted from a plurality of data object combinations.

The filtering rule for bright spots may specifically be: more than 1 light in a data object combination needs to be filtered out. The condition that a color is defined as a bright spot is that S > of the color in HSV space is 0.72 and V > is 0.87 (very vivid); the condition that a texture is defined as a highlight is that the Area of the texture is 80% (large Area texture). The rule is executed by sequentially traversing each commodity in a data object combination, counting the number of bright spots, and when the total number of the bright spots of different types is greater than 1, the data object combination needs to be filtered, namely, the data object combination is deleted from the data object combinations.

After the above process, the remaining data object combinations which are not filtered out are the clothing matching examples which can be referred by the user.

Sorting section

Generally, after the filtering process, about half of the remaining area can be provided to the user as a clothing matching example. These apparel matching instances may be ranked prior to being provided to the user to determine the final user-oriented presentation order. This step is typically done to further optimize the click-through rate of the collocation. Specifically, the matching points can be calculated according to the popularity points (which can be sales volume, click rate, purchase rate and the like, and can be adjusted according to different business targets) of each commodity in each clothing matching example, and then the matching points are sorted according to the matching points. And then presented to the user in a high-to-low ranking of the match score.

Fig. 5 is a schematic flowchart illustrating a data object processing method according to another embodiment of the present application. As shown in the figure, the data object processing method includes:

201. and acquiring at least one type of second data object different from the type of the first data object.

202. And judging whether the at least one second data object is collocated with the first data object or not according to the attribute of the first data object and the attribute of the at least one second data object.

203. When the at least one second data object is matched with the first data object, the at least one second data object is provided for the user.

In the foregoing 201, the second data objects of each type in the at least one type of second data object may be obtained in the data object sets of corresponding types. In the costume type e-market scene, the first data object and the second data object may be costume type commodities such as dress, coat, shoe, bag and the like. The first data object, the second data object may be a commodity number, a commodity name, or the like for uniquely identifying the commodity.

In 202, if in the service class application scenario, the attribute may include at least one of the following: part, style, color information, texture information, sales record, purchase rate, click rate, and shelf time. Accordingly, the step 202 may include at least one of the following steps:

2021. the style of the first data object and the style of any one second data object in at least one second data object are used for judging whether the first data object and the at least one second data object are matched or not;

2022. determining the number of dominant colors according to the color information of the first data object and the color information of at least one second data object; judging whether the first data object and the at least one second data object are matched or not according to the number of the dominant colors;

2023. determining whether there is a color highlight as a first determination result according to the color information of the first data object and the color information of the at least one second data object; and determining whether the texture bright point exists as a second determination result according to the texture information of the first data object and the texture information of the at least one second data object, and determining whether the first data object and the at least one second data object are collocated according to the first determination result and the second determination result.

The condition that one color is defined as a light point is: s > of this color in HSV space is 0.72 and V > is 0.87 (very vivid); the condition that a texture is defined as a highlight is that the Area of the texture is 80% (large Area texture). And if the bright point of the combination of the first data object and the at least one second data object is not more than 1 according to the first determination result and the second determination result, the first data object is collocated with the at least one second data object.

Here, it should be noted that: the specific implementation of the foregoing 2021-2023 can refer to the corresponding content in the foregoing embodiments, and is not described herein again.

According to the technical scheme provided by the embodiment, the second data object used for being combined with the first data object is extracted from the data object set of at least one type different from the type of the first data object to generate a plurality of data object combinations, and then a clothing matching example which can be referred by a user is selected from the plurality of data object combinations by using a matching rule instead of a strict matching template, so that the universality is high, and the matching produced on the premise of ensuring the matching quality is richer; in addition, through actual effect measurement, the technical scheme provided by the embodiment of the application can output multiple sets of matching suggestions for 99% of the commodities in the target commodity pool, and the satisfaction rate of manual evaluation reaches 80%.

The technical scheme provided by each embodiment of the application can be applied to a clothing matching recommendation scene. For example, a user selects a dress via a client Application (APP) and adds the selected dress to a shopping cart. After monitoring the user behavior, the client recommends a coat, a shoe, a bag and the like which can be matched with the one-piece dress for the user; and the page information corresponding to the coats, shoes and bags which can be matched with the dress is displayed on the user interface as recommendation information. For example, in one possible scenario, the client sends a collocation request to the server for the dress. The server can feed back a coat, a shoe, a bag and the like which can be recommended for the user to the client by adopting the method provided by each embodiment. That is, the technical solution provided in this embodiment can also be implemented by using the following hardware system architecture. As shown in fig. 6, the system for processing clothing matching information includes:

the client 301 is used for responding to the operation of the user on the first data object and sending request information to the server; receiving a clothing matching example fed back by the server aiming at the first data object; providing the apparel collocation instance to the user;

the server 302 is configured to obtain request information sent by the client for the first data object; extracting a second data object to be combined with the first data object from a set of data objects of at least one type different from a type to which the first data object belongs to generate a plurality of data object combinations; and feeding back the data object combination which accords with the collocation rule in the plurality of data object combinations to the client as a clothing collocation example.

In the system for processing clothing matching information, the specific implementation methods of the components such as the client and the server are described in detail in the following embodiments.

Fig. 7 is a schematic flowchart illustrating a processing method of the clothing matching information according to an embodiment of the present application. The execution main body of the method provided by this embodiment may be a client, and the client may be any terminal device such as a mobile phone, a tablet computer, a PDA (Personal Digital Assistant), a POS (Point of Sales), and a vehicle-mounted computer. As shown in fig. 7, the method includes:

401. and responding to the operation of the user on the first data object, and sending request information to the server.

402. And receiving a clothing matching example fed back by the server aiming at the first data object.

403. Providing the clothing collocation instance to the user.

In 401 above, the specifying operation may include: the operation of clicking the corresponding control key by the user, the control voice sent by the user for the first data object, or the user making a specified action, etc., which is not specifically limited in this embodiment.

Here, it should be noted that: for the generation process of the clothing matching example, reference may be made to the relevant contents in the above embodiments, and details are not repeated here.

Further, the step 403 "providing the clothing match example to the user" may include at least one of:

displaying the clothing matching example;

displaying a virtual fitting image on a virtual model based on the clothing matching example;

and acquiring a photo of the user, and generating an image of the virtual try-on of the user according to the photo.

Fig. 8 is a flowchart illustrating a method for processing clothing matching information according to an embodiment of the present application. The execution main body of the method provided in this embodiment may be a server, and the server may be a common server, a virtual server, a cloud, or the like, which is not particularly limited in this embodiment. As shown in fig. 8, the method includes:

501. and acquiring request information sent by a client aiming at the first data object.

502. From a set of data objects of at least one type different from the type to which the first data object belongs, a second data object to be combined with the first data object is extracted to generate a plurality of data object combinations.

503. And feeding back the data object combination which accords with the collocation rule in the plurality of data object combinations to the client as a clothing collocation example.

The specific implementation of the steps 501 to 502 can refer to the related contents in the above embodiments, and the details are not repeated herein.

Further, the method provided by this embodiment may also implement other steps described in the above embodiments in addition to the above steps, and specifically refer to the description in the foregoing embodiments.

Fig. 9 is a schematic structural diagram illustrating a device for processing clothing matching information according to an embodiment of the present application. As shown in fig. 9, the data object processing apparatus includes: an acquisition module 11, a combination module 12 and a selection module 13. The obtaining module 11 is configured to obtain a set of data objects of at least one type different from a type to which the first data object belongs; the combining module 12 is configured to extract a second data object from the set of data objects of the at least one type to be combined with the first data object to generate a plurality of data object combinations; the selection module 13 is configured to select a data object combination meeting the matching rule from the plurality of data object combinations as a clothing matching example for the user to refer to.

Further, the combination module 12 may be specifically configured to:

acquiring at least one combination rule;

selecting one data object from a part of or all of the data object sets of the at least one type of data object set as a data object combination with the first data object, respectively, using the at least one combination rule; until the number of different generated data object combinations meets the preset number requirement.

Further, the at least one combination rule includes: a first combination rule and a second combination rule;

and the combination module 12 is further configured to:

acquiring current sequencing serial numbers Ni corresponding to the data object sets of the at least two types, wherein the i values of the data object sets of different types are different;

respectively acquiring data objects which are sequenced into Ni from all data object sets specified by the first combination rule as second data objects to be combined with the first data objects into data object combinations, and updating Ni corresponding to all the data object sets specified by the first combination rule into Ni + 1;

respectively acquiring data objects which are ranked as Ni from all data object sets specified by the second combination rule to serve as second data objects to be combined with the first data objects to form data object combinations, and updating Ni corresponding to all the data object sets specified by the second combination rule to Ni + 1;

and repeating the steps until the number of the generated different combination examples meets the preset number requirement.

Further, the selection module 13 is further configured to:

acquiring the attribute of the data object contained in each data object combination in the plurality of data object combinations;

determining whether a combination of attributes of data objects included in each of the plurality of data object combinations conforms to a filtering rule using the filtering rule for at least one of the attribute items;

and deleting the data object combinations of which the attribute combinations of the data objects in the plurality of data object combinations meet the filtering rule.

Further, the attribute of the data object comprises at least one attribute item of: type, style, color information, texture information, sales records, purchase rate, click rate, time on shelf.

Further, the filtering rules include: rules for style; correspondingly, the selection module 13 is further configured to:

Still further, the selecting module 13 is further configured to: and deleting the first data object combination when the style of the first data object and the style combination of any second data object in the at least one second data object accord with the rule aiming at the style.

Further, the filtering rules include: rules for color; and the selection module 13 is further configured to:

Still further, the selection module is further configured to: and when the number of the dominant colors is larger than a first threshold value, deleting the first data object combination.

Further, the filtering rules include: rules for bright spots; and the selection module 13 is further configured to:

Still further, the selecting module 13 is further configured to:

Further, the obtaining module is further configured to:

acquiring the attribute of each data object in the first type data object pool;

Further, the method provided by this embodiment further includes: and a sorting module. The sequencing module is used for sequencing a plurality of clothes matching examples; and providing a plurality of clothing matching examples for the user according to the sequencing result.

Here, it should be noted that: the processing device for clothing matching information provided in the above embodiments may implement the technical solutions described in the above method embodiments, and the specific implementation principle of each module or unit may refer to the corresponding content in the above method embodiments, and will not be described herein again.

Fig. 10 is a schematic structural diagram illustrating a data object processing apparatus according to an embodiment of the present application. As shown in fig. 10, the data object processing apparatus includes: an acquisition module 21, a determination module 22 and a providing module 23. The obtaining module 21 is configured to obtain a second data object of at least one type different from the type to which the first data object belongs; the determining module 22 is configured to determine whether the at least one second data object is collocated with the first data object according to the attribute of the first data object and the attribute of the at least one second data object; the providing module 23 is configured to provide the at least one second data object to the user when the at least one second data object is matched with the first data object.

Further, in the clothing application scenario, the first data object and the second data object are specific clothing goods, and accordingly, the attributes of the first data object and the second data object may specifically include at least one of the following: part, style, color information, texture information, sales record, purchase rate, click rate and shelf time; correspondingly, the determining module 22 is further configured to:

the style of the first data object and the style of any one second data object in at least one second data object are used for judging whether the first data object and the at least one second data object are matched or not; and/or

Determining the number of dominant colors according to the color information of the first data object and the color information of at least one second data object; and/or

Judging whether the first data object and the at least one second data object are matched or not according to the number of the dominant colors; and/or

Determining whether there is a color highlight as a first determination result according to the color information of the first data object and the color information of the at least one second data object; and determining whether the texture bright point exists as a second determination result according to the texture information of the first data object and the texture information of the at least one second data object, and determining whether the first data object and the at least one second data object are collocated according to the first determination result and the second determination result.

Here, it should be noted that: the data object processing apparatus provided in the foregoing embodiments may implement the technical solutions described in the foregoing method embodiments, and the specific implementation principle of each module or unit may refer to the corresponding content in the foregoing method embodiments, and is not described herein again.

Fig. 11 shows a schematic structural diagram of a device for processing clothing matching information according to an embodiment of the present application. As shown in fig. 11, the data object processing apparatus includes: a transmitting module 31, a receiving module 32 and an output module 33. The sending module 31 is configured to send request information to the server in response to an operation performed by the user on the first data object; the receiving module 32 is configured to receive a clothing matching example fed back by the server for the first data object; the output module 33 is used for providing the clothing match example to the user. The clothing matching example is a data object combination which is selected from a plurality of data object combinations and accords with matching rules; the plurality of data object combinations are generated by combining a second data object extracted from a set of data objects of at least one type different from the type to which the first data object belongs with the first data object.

Further, the output module is further configured to:

displaying the clothing matching example; and/or

Displaying a virtual fitting image on a virtual model based on the clothing matching example; and/or

Fig. 12 is a schematic structural diagram illustrating a device for processing clothing matching information according to an embodiment of the present application. As shown in fig. 12, the data object processing apparatus includes: an acquisition module 41, an extraction module 42, and a feedback module 43. The obtaining module 41 is configured to obtain request information sent by a client for the first data object; the extracting module 42 is configured to extract a second data object to be combined with the first data object from a set of data objects of at least one type different from a type to which the first data object belongs, so as to generate a plurality of data object combinations; the feedback module 43 is configured to feed back, to the client, a data object combination that meets the matching rule in the plurality of data object combinations as a clothing matching example.

Fig. 13 shows a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device includes a memory 51 and a processor 52. The memory 51 may be configured to store other various data objects to support operations on the electronic device. Examples of such data objects include instructions for any application or method operating on the electronic device. The memory 51 may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.

The processor 52, coupled to the memory 51, is configured to execute the program stored in the memory 51, so as to:

When the processor 52 executes the program in the memory 51, in addition to the above functions, other functions may be implemented, and reference may be specifically made to the description of the foregoing embodiments.

Further, as shown in fig. 13, the electronic device further includes: display 54, communications component 53, power component 55, audio component 56, and the like. Only some of the components are schematically shown in fig. 13, and the electronic device is not meant to include only the components shown in fig. 13.

An embodiment of the application further provides the electronic equipment. The structure of the electronic device provided by this embodiment is similar to the structure of the electronic device provided by the above embodiment, and is shown in fig. 13. The electronic device includes a memory and a processor. The processor, coupled with the memory, to execute the program stored in the memory to:

When the processor executes the program in the memory, the processor may implement other functions in addition to the above functions, which may be specifically referred to the description of the foregoing embodiments.

An embodiment of the application further provides the client device. The structure of the client device provided in this embodiment is similar to that of the electronic device embodiment described above, and is shown in fig. 13. The client device includes: a memory and a processor. Wherein the processor, coupled with the memory, is configured to execute the program stored in the memory to:

providing the apparel collocation instance to the user;

An embodiment of the application further provides the server side equipment. The structure of the server device provided in this embodiment is similar to that of the electronic device embodiment, and is shown in fig. 13. The server device includes: a memory and a processor. Wherein the processor, coupled with the memory, is configured to execute the program stored in the memory to:

acquiring request information sent by a client aiming at the first data object;

Accordingly, embodiments of the present application further provide a computer-readable storage medium storing a computer program, where the computer program, when executed by a computer, can implement the steps or functions of the method for processing clothing matching information provided in the foregoing embodiments.

In addition, embodiments of the present application also provide a computer-readable storage medium storing a computer program, where the computer program can implement the steps or functions of the data object processing method provided in each of the above embodiments when executed by a computer.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims

1. A processing method of clothes matching information is characterized by comprising the following steps:

and selecting the data object combination which accords with the matching rule from the plurality of data object combinations as a clothing matching example for the user to refer to.

2. The method of claim 1, wherein extracting a second data object from the set of data objects of the at least one type to be combined with the first data object to generate a plurality of data object combinations comprises:

acquiring at least one combination rule;

3. The method of claim 2, wherein the at least one combination rule comprises: a first combination rule and a second combination rule; and

selecting one data object from part of or all of the data object sets of at least two types of data object sets as a second data object to be combined with the first data object into a data object combination by using the first combination rule and the second combination rule; until the number of the generated different data object combinations meets the preset number requirement, the method comprises the following steps:

4. The method according to any one of claims 1 to 3, wherein selecting a data object combination that meets a matching rule from the plurality of data object combinations as a clothing matching instance that can be referred to by a user comprises:

5. The method of claim 4, wherein the attributes of the data object include at least one of the following attribute items: type, style, color information, texture information, sales records, purchase rate, click rate, time on shelf.

6. The method of claim 5, wherein filtering rules comprises: rules for style; and

using a style-specific rule to determine whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the style-specific rule, comprising:

7. The method according to claim 6, wherein deleting the data object combinations whose attribute combinations of the data objects included in the plurality of data object combinations meet the filtering rule comprises:

8. The method of claim 5, wherein filtering rules comprises: rules for color; and

using a color-directed rule to determine whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the style-directed rule, comprising:

9. The method according to claim 8, wherein deleting the data object combinations whose attribute combinations of the data objects included in the plurality of data object combinations meet the filtering rule comprises:

10. The method of claim 5, wherein filtering rules comprises: rules for bright spots; and

using a rule for a highlight, determining whether a combination of attributes of data objects contained in a first data object combination of the plurality of data object combinations complies with the rule for a highlight, comprising:

11. The method according to claim 8, wherein deleting the data object combinations whose attribute combinations of the data objects included in the plurality of data object combinations meet the filtering rule comprises:

12. The method of any of claims 1 to 3, wherein obtaining a set of data objects of a first type different from the type to which the first data object belongs comprises:

acquiring the attribute of each data object in the first type data object pool;

13. The method according to any one of claims 1 to 3, wherein in the case that there are a plurality of obtained examples of dress matches that can be referred to by the user, the method further comprises:

sequencing the multiple clothing matching examples;

14. A method for processing a data object, comprising:

15. The method of claim 14, wherein the attribute comprises at least one of: part, style, color information, texture information, sales record, purchase rate, click rate and shelf time; and

determining whether the at least one second data object is collocated with the first data object according to the attribute of the first data object and the attribute of the at least one second data object, wherein the determining includes at least one of the following:

the style of the first data object and the style of any one second data object in at least one second data object are used for judging whether the first data object and the at least one second data object are matched or not;

determining the number of dominant colors according to the color information of the first data object and the color information of at least one second data object; judging whether the first data object and the at least one second data object are matched or not according to the number of the dominant colors;

16. A processing method of clothes matching information is characterized by comprising the following steps:

providing the apparel collocation instance to the user;

17. The method of claim 16, wherein providing the apparel collocation instance to the user comprises at least one of:

displaying the clothing matching example;

18. A processing method of clothes matching information is characterized by comprising the following steps:

acquiring request information sent by a client aiming at the first data object;

19. A system for processing clothes matching information is characterized by comprising:

the client is used for responding to the operation of the user on the first data object and sending request information to the server; receiving a clothing matching example fed back by the server aiming at the first data object; providing the apparel collocation instance to the user;

20. An electronic device comprising a memory and a processor; wherein,

the memory is used for storing programs;

21. An electronic device comprising a memory and a processor; wherein,

the memory is used for storing programs;

22. A client device comprising a memory and a processor; wherein,

the memory is used for storing programs;

providing the apparel collocation instance to the user;

23. The server-side equipment is characterized by comprising a memory and a processor; wherein,

the memory is used for storing programs;

acquiring request information sent by a client aiming at the first data object;