CN112581230A - Commodity recommendation method and device - Google Patents
Commodity recommendation method and device Download PDFInfo
- Publication number
- CN112581230A CN112581230A CN202011546923.9A CN202011546923A CN112581230A CN 112581230 A CN112581230 A CN 112581230A CN 202011546923 A CN202011546923 A CN 202011546923A CN 112581230 A CN112581230 A CN 112581230A
- Authority
- CN
- China
- Prior art keywords
- commodity
- user
- determining
- recommendation
- commodities
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000008451 emotion Effects 0.000 claims abstract description 40
- 238000004091 panning Methods 0.000 claims description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 7
- 230000001815 facial effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 210000003128 head Anatomy 0.000 description 15
- 238000000605 extraction Methods 0.000 description 7
- 238000001514 detection method Methods 0.000 description 6
- 210000004709 eyebrow Anatomy 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000001914 filtration Methods 0.000 description 4
- 230000010355 oscillation Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000007635 classification algorithm Methods 0.000 description 2
- 230000008909 emotion recognition Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/174—Facial expression recognition
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a commodity recommendation method and a commodity recommendation device, wherein the method comprises the following steps: acquiring a first face image of a user; determining identity information of the user according to the first face image, and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities; pushing the commodities in the commodity recommendation combination to the user one by one, and acquiring a second face image when the user browses each pushed commodity; extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index; and optimizing the commodities in the commodity recommendation combination according to the grading result. The technical scheme of the invention can improve the accuracy of commodity recommendation and the use experience of the user.
Description
Technical Field
The invention relates to the technical field of commodity recommendation, in particular to a commodity recommendation method and device.
Background
With the rapid development of electronic commerce, the shopping experience of users is improved, more and more trading platforms recommend commodities to the users in a shopping interface, the current commodity recommendation often adopts a recommendation method based on collaborative filtering, a recommendation method based on content filtering and a recommendation method based on association rule filtering, the recommendation method carries out commodity recommendation according to the identity characteristics or shopping habits of the users, in order to judge whether the recommended commodities meet the mind of the users, manual operation of the users is needed to feed back whether the commodities are interested, the operation is complex, and the accuracy is not high.
Disclosure of Invention
The invention solves the problem of how to reduce user operation and improve the accuracy of commodity recommendation so as to improve the use experience of users.
In order to solve the above problems, the present invention provides a method and an apparatus for recommending a commodity.
In a first aspect, the present invention provides a commodity recommendation method, including:
acquiring a first face image of a user;
determining identity information of the user according to the first face image, and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities;
pushing the commodities in the commodity recommendation combination to the user one by one, and acquiring a second face image when the user browses each pushed commodity;
extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index;
and optimizing the commodities in the commodity recommendation combination according to the grading result.
Optionally, the determining, according to the first face image, the identity information of the user, and the determining, according to the identity information, a combination of goods recommended to the user includes:
extracting first face feature data in the first face image;
comparing the first face feature data with face registration information in a database, wherein the face registration information corresponds to the identity information one by one;
determining the identity information corresponding to the first face feature data according to the face registration information, wherein the identity information corresponds to historical commodity recommendation combinations one to one;
and determining the corresponding historical commodity recommendation combination according to the identity information, and determining the current commodity recommendation combination according to the historical commodity recommendation combination.
Optionally, the determining a current product recommendation combination according to the historical product recommendation combination includes:
when the historical commodity recommendation combination is empty, determining the historical consumption record of the user in a database according to the identity information;
when the historical consumption record of the user is found in the database, determining the current commodity recommendation combination according to the historical consumption record;
and when the historical consumption record of the user is not found in the database, determining a user portrait corresponding to the user according to the facial feature data, and determining the current commodity recommendation combination according to the user portrait.
Optionally, the determining a current product recommendation combination according to the historical product recommendation combination includes:
when the historical commodity recommendation combination is not empty, determining commodities liked by the user according to the historical commodity recommendation combination;
determining the similarity between each commodity in the database and each favorite commodity of the user;
determining a similar commodity set of commodities liked by each user according to the similarity, wherein the similar commodity set comprises a plurality of similar commodities of the commodities liked by the user;
calculating the interest degree of the user for each similar commodity according to the predetermined interest degree of the user for the favorite commodities of the user and the similarity;
and determining the current commodity recommendation combination in the similar commodity set according to the interestingness.
Optionally, the similarity between each product in the database and each product liked by the user is determined by using a first formula, where the first formula includes:
wherein, WijRepresenting the similarity between a commodity i and a commodity j, assuming that the commodity i is a commodity liked by any user, the commodity j is any commodity in a database, N (i) represents the number of commodities i purchased by the user, N (j) represents the number of commodities j purchased by the user, and N (i) n (j) represents the number of commodities i and j purchased by the user;
for any user u, calculating the interest degree of the user u in each similar commodity by adopting a second formula, wherein the second formula comprises the following steps:
Puk＝∑i∈N(u)Wik*ruk，
wherein, PujRepresenting the interest degree of the user u in a commodity k, wherein the commodity k is any similar commodity in the similar commodity set, N (u) represents the historical commodity recommendation set of the user u, and WijIndicates the similarity between the product i and the product k, ruiAnd representing the interest degree of the user u in the commodity i.
Optionally, the determining, according to the second face image, an emotion index when the user browses each pushed commodity includes:
extracting second face feature data in the second face image;
carrying out normalization processing on the second face feature data to obtain normalized face feature data;
and determining the emotion index according to the normalized human face feature data based on a facial animation parameter algorithm.
Optionally, the emotion index includes a smiling index and an oscillation index, the head oscillation angle includes a head raising angle, a plane rotation angle, and an oscillation angle, the emotion index when the user browses each pushed commodity is determined according to the smile and the head oscillation angle, and scoring each pushed commodity according to the emotion index includes:
determining the smiling index and the panning index respectively by adopting a third formula according to the smile and the head swing angle, wherein the third formula comprises:
wherein smile represents the smile index, shakeidedex represents the panning index, V1 represents the smile, V2 represents the head-up angle, V3 represents the plane rotation angle, and V4 represents the panning angle;
and scoring each pushed commodity according to the emotion index by adopting a fourth formula, wherein the fourth formula comprises the following steps:
X＝[5/(1+e-(0.25*smile-1.2*shakeindex)/10)]，
wherein X represents the score of the product, smile represents the smile index, and shakeidedex represents the shake-head index.
Optionally, the optimizing the commodities in the commodity recommendation combination according to the scoring result includes:
and updating the interest degree of the user for each commodity in the current commodity recommendation combination according to the grade of the commodity.
In a second aspect, the present invention provides a commodity recommending apparatus, including:
the acquisition module is used for acquiring a first face image of a user;
the identity recognition module is used for determining identity information of the user according to the first face image and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities;
the pushing module is used for pushing the commodities in the commodity recommendation combination to the user one by one and acquiring a second face image when the user browses each pushed commodity;
the grading module is used for extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index;
and the optimization module is used for optimizing the commodities in the commodity recommendation combination according to the grading result.
In a third aspect, the present invention provides a merchandise recommendation device, including a memory and a processor;
the memory for storing a computer program;
the processor is configured to implement the product recommendation method as described above when executing the computer program.
The commodity recommendation method and the commodity recommendation device have the beneficial effects that: firstly, a first face image of a user is obtained, face recognition is carried out according to the first face image, and identity information of the user is determined. The commodity recommendation combination pushed to the user is determined according to the identity information of the user, and the commodity recommendation is specifically carried out on different identities, so that the use experience of the user can be improved, and the consumption is promoted. The commodities in the commodity recommendation combination are pushed to a display interface one by one to be displayed, second face images when the user browses the pushed commodities are obtained, emotion indexes of the user on the commodities are determined according to the second face images, the commodities are scored according to the emotion indexes, whether the user is interested in the pushed commodities and the interest degree of the user can be determined, the commodities in the commodity recommendation combination are optimized according to scoring results, namely, the recommendation results are optimized according to feedback of the user, and the accuracy of next recommendation is improved. According to the technical scheme, emotion recognition is carried out on the face image when the user browses the pushed commodity, the recommended commodity is optimized according to the recognized emotion index, manual operation of the user is not needed for feedback, the use experience of the user is greatly improved, and the accuracy of the recommendation result can be gradually improved.
Drawings
Fig. 1 is a schematic flow chart of a commodity recommendation method according to an embodiment of the present invention;
fig. 2 is a flowchart illustrating a commodity recommendation method according to another embodiment of the present invention;
FIG. 3 is a schematic diagram of a process of extracting face feature data according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a scoring level according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a merchandise recommendation interface according to an embodiment of the invention;
fig. 6 is a schematic structural diagram of a commodity recommending apparatus according to an embodiment of the present invention.
Description of reference numerals:
10-a goods recommendation interface; 20-a face image area; 30-a goods presentation area; 40-information display area.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein.
As shown in fig. 1 and fig. 2, a commodity recommendation method according to an embodiment of the present invention includes:
step S110, a first face image when a user browses a commodity interface is obtained.
Optionally, when the user browses the goods interface, the camera may be called to implement obtaining of the first face image of the user, and the Web client receives the first face image transmitted by the camera and transmits the first face image to the server for processing.
Step S120, determining the identity information of the user according to the first face image, and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities.
Optionally, the server performs face recognition according to the first face image, and first performs preprocessing on the acquired first face image, where the preprocessing may include filtering, smoothing and sharpening the image, binarization, extracting feature points, and the like, and the preprocessed first face image determines identity information of the user in a database.
Optionally, the determining, according to the first face image, the identity information of the user, and the determining, according to the identity information, a combination of goods recommended to the user includes:
and step S121, extracting first face feature data in the first face image.
Specifically, the first face image may be input to a face recognition platform, such as a face + + face recognition platform, and the first face feature data obtained by the face recognition platform is obtained. The first face feature data includes feature data of eyebrows, eyes, nose, and mouth.
As shown in fig. 3, face detection is performed on an image captured by a camera, a function candidate region is divided into a face image, the function candidate region includes a mouth corner region, an eye region, a nose region, and an eyebrow region, and mouth boundary extraction, eye boundary extraction, nose detection, and eyebrow detection are performed on the corresponding function candidate region, wherein the eyebrow detection and nose detection may only employ a single algorithm to perform feature extraction, so as to obtain features of a nose and features of eyebrows. Because the detection of the mouth and the eyes is relatively complex, a single algorithm is adopted for extraction, and the result may have a large deviation, in this embodiment, a plurality of algorithms are adopted for extracting the eyes and the mouth respectively, a neural network classification algorithm, a classification algorithm based on a region growing technology, a canny operator edge extraction algorithm and the like can be adopted for extracting features, and then all the extracted eye features or mouth features are fused, for example, the features extracted by each algorithm are weighted and summed to obtain more accurate features.
In order to improve the accuracy of feature extraction, the feature extraction result may be verified, for example, the face actual data of a tester is measured, an index is established according to the face actual data, the data extracted by the feature is checked using the index, and whether the extracted data is an effective feature is determined.
And step S122, comparing the first face feature data with face registration information in a database, wherein the face registration information corresponds to the identity information one by one.
Specifically, the first face feature data is normalized, the processed first face feature data is compared with face registration information in a database, the face registration information is face information input during user registration, and each face registration information includes corresponding feature data, such as mouth corner information, eye information, nose information and eyebrow information.
Step S123, determining the identity information corresponding to the first face feature data according to the face registration information, wherein the identity information corresponds to historical commodity recommendation combinations one to one;
and step S124, determining the corresponding historical commodity recommendation combination according to the identity information, and determining the current commodity recommendation combination according to the historical commodity recommendation combination.
Optionally, the determining a current product recommendation combination according to the historical product recommendation combination includes:
and step S1241, when the historical commodity recommendation combination is not empty, determining the favorite commodities of the user according to the historical commodity recommendation combination.
In this optional embodiment, when the historical commodity recommendation combination is not empty, that is, when a commodity is included in the historical commodity recommendation combination, the score of each commodity in the historical commodity recommendation combination may be compared with a preset threshold, and the commodity with the score higher than the preset threshold is a commodity that the user likes, that is, the score is determined according to the interest level of each commodity.
In step S1242, the similarity between each product in the database and each product liked by the user is determined.
Optionally, the similarity between each product in the database and each product liked by the user is determined by using a first formula, where the first formula includes:
wherein, WijRepresenting the similarity between a commodity i and a commodity j, assuming that the commodity i is a commodity liked by any user, the commodity j is any commodity in a database, N (i) representing the number of commodities i purchased by the user, N (j) representing the number of commodities j purchased by the user, and N (i) # N (j) representing the number of commodities i and j purchased by the user at the same time.
Step S1243, determining a similar commodity set of each favorite commodity of the user according to the similarity, where the similar commodity set includes a plurality of similar commodities of the favorite commodity of the user.
In this optional embodiment, for a commodity liked by any user, the similarity between each commodity in the database and the commodity is determined, and the commodities with the calibrated number can be sequentially selected from high to low according to the similarity to form a similar commodity set, wherein each commodity in the similar commodity set is a similar commodity of the commodity liked by the user.
Step S1244, calculating the interest level of the user in each similar product according to the predetermined interest level and the similarity level of the user in the product that the user likes.
Optionally, for any user u, a second formula is adopted to calculate the interest degree of the user u in each similar product, where the second formula includes:
Puk＝∑i∈N(u)Wik*ruk，
wherein, PujRepresenting the interest degree of the user u in a commodity k, wherein the commodity k is any similar commodity in the similar commodity set, N (u) represents the historical commodity recommendation set of the user u, and WijIndicates the similarity between the product i and the product k, ruiRepresents the interest degree r of the user u in the commodity iuiMay be a ratio of the number of times the user u purchases the item i to the total number of purchases of all items.
Optionally, the determining a current product recommendation combination according to the historical product recommendation combination includes:
when the historical commodity recommendation combination is empty, determining the historical consumption record of the user in a database according to the identity information;
and when the historical consumption record of the user is found in the database, determining the current commodity recommendation combination according to the historical consumption record.
In this optional embodiment, when the historical commodity recommendation combination is empty, and there is no commodity in the historical commodity recommendation combination, it is determined whether the historical consumption record of the user can be found in the database according to the identity information, and if the historical consumption record can be found, the commodity that the user likes is determined according to the commodity in the historical consumption record, for example, the score of each commodity in the historical consumption record may be compared with a preset threshold, and the commodity with the score higher than the preset threshold is the commodity that the user likes.
And when the historical consumption record of the user is not found in the database, determining a user portrait corresponding to the user according to the facial feature data, and determining the current commodity recommendation combination according to the user portrait.
Specifically, the user portrait is an effective tool for delineating a target user and contacting the user with complaints and design directions, including gender, age, and the like of the user, and corresponding commodities are recommended according to the user portrait corresponding to the user, for example: fashion trend commodities can be recommended for young people, home life commodities can be recommended for middle-aged people, and the like.
In the optional embodiment, when the commodity recommendation combination is empty, that is, no commodity exists in the commodity recommendation combination, the commodity in the commodity recommendation combination is determined according to the historical consumption record or the user image of the user, and the commodity can be recommended for the user in cold start, so that the use experience of the user is improved.
Step S1245, determining the current commodity recommendation combination in the similar commodity set according to the interestingness.
In this optional embodiment, a calibrated number of commodities may be selected from the similar commodity set from high to low according to the interestingness, all the selected commodities constitute a current commodity recommendation combination, and the current commodity recommendation combination may further include commodities that the user likes.
Step S130, pushing the commodities in the commodity recommendation combination to the user one by one, and acquiring a second face image when the user browses each pushed commodity.
Specifically, the commodities in the current commodity recommendation combination are pushed to a commodity interface one by one, and a second face image of the user when the user browses each commodity can be obtained by calling a camera.
Step S140, extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index.
Optionally, the determining, according to the second face image, an emotion index when the user browses each pushed commodity includes:
extracting second face feature data in the second face image;
and carrying out normalization processing on the second face feature data to obtain normalized face feature data.
Specifically, after the second face feature data is normalized, 4 components, namely smile (V1) (smile), pitchangle (V2) (head raising angle), rolangle (V3) (plane rotation angle) and yawangle (V1) (head shaking angle), are adopted to represent the face feature data, and the face + + face recognition platform can be adopted to recognize the face feature data.
And determining the emotion index according to the normalized human face feature data based on a facial animation parameter algorithm.
Optionally, based on the four components, determining the smiling index and the panning index respectively using a third formula, the third formula including:
wherein smile represents the smile index and shakeidedex represents the panning index.
Optionally, the emotion indexes include a smiling index and an shaking head index, and scoring the pushed commodities according to the emotion indexes includes:
and scoring each pushed commodity according to the emotion index by adopting a fourth formula, wherein the fourth formula comprises the following steps:
X＝[5/(1+e-(0.25*smile-1.2*shakeindex)/10)]，
wherein X represents the score of the product, smile represents the smile index, and shakeidedex represents the shake-head index.
In this optional embodiment, the commodities in the current commodity recommendation combination are pushed to the user one by one, and the attitude of the user to each pushed commodity can be determined according to the score, for example: as shown in fig. 4, the total score can be set to 5, and the score is divided into 5 levels, wherein the score is between (0, 1) and represents that the user is not satisfied with the commodity, the score is between (1, 2) and represents that the user is not satisfied with the commodity, the score is between (2, 3) and represents that the user maintains a moderate attitude with respect to the commodity, the score is between (3, 4) and represents that the user is satisfied with the commodity, and the score is between (4, 5) and represents that the user is satisfied with the commodity.
As shown in the commodity recommendation interface 10 in fig. 5, the commodities in the current commodity recommendation combination are pushed to the commodity display area 30 one by one for display, the face images of the user browsing the recommended commodities are captured by the camera, the face images are processed, the emotion index and score of the user on the commodities are determined, and the identity information, the emotion index, the score and other information of the user are displayed in the information display area 40.
And S150, optimizing the commodities in the commodity recommendation combination according to the grading result.
Optionally, the optimizing the commodities in the commodity recommendation combination according to the scoring result includes:
and updating the interest degree of the user for each commodity in the current commodity recommendation combination according to the grade of the commodity.
In this optional embodiment, for any commodity in the current commodity recommendation combination, the interest degree of the commodity is updated according to the score determined by the emotion index corresponding to the commodity, and the current interest degree of the commodity is updated to the ratio of the score of the commodity to the total score of the score, for example: and if the total score is 5 points, taking the ratio of the score of the commodity to the 5 points as the new interest level of the commodity.
In the commodity recommendation method of this embodiment, a first face image of a user is first obtained, face recognition is performed according to the first face image, and identity information of the user is determined. The commodity recommendation combination pushed to the user is determined according to the identity information of the user, and the commodity recommendation is specifically carried out on different identities, so that the use experience of the user can be improved, and the consumption is promoted. The commodities in the commodity recommendation combination are pushed to a display interface one by one to be displayed, second face images when the user browses the pushed commodities are obtained, emotion indexes of the user on the commodities are determined according to the second face images, the commodities are scored according to the emotion indexes, whether the user is interested in the pushed commodities and the interest degree of the user can be determined, the commodities in the commodity recommendation combination are optimized according to scoring results, namely, the recommendation results are optimized according to feedback of the user, and the accuracy of next recommendation is improved. According to the technical scheme, emotion recognition is carried out on the face image when the user browses the pushed commodity, the recommended commodity is optimized according to the recognized emotion index, manual operation of the user is not needed for feedback, the use experience of the user is greatly improved, and the accuracy of the recommendation result can be gradually improved.
As shown in fig. 6, a commodity recommendation device according to an embodiment of the present invention includes:
the acquisition module is used for acquiring a first face image when a user browses a commodity interface;
the identity recognition module is used for determining identity information of the user according to the first face image and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities;
the pushing module is used for pushing the commodities in the commodity recommendation combination to the user one by one and acquiring a second face image when the user browses each pushed commodity;
the grading module is used for extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index;
and the optimization module is used for optimizing the commodities in the commodity recommendation combination according to the grading result.
Another embodiment of the present invention provides a merchandise recommendation device including a memory and a processor; the memory for storing a computer program; the processor is configured to implement the product recommendation method as described above when executing the computer program. The device may be a computer or a server, etc.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like. In this application, 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 units can be selected according to actual needs to achieve the purpose of the solution of the embodiment of the present invention. In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
Although the present disclosure has been described above, the scope of the present disclosure is not limited thereto. Various changes and modifications may be effected therein by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present disclosure, and these changes and modifications are intended to be within the scope of the present disclosure.
Claims (10)
1. A method for recommending an article, comprising:
acquiring a first face image of a user;
determining identity information of the user according to the first face image, and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities;
pushing the commodities in the commodity recommendation combination to the user one by one, and acquiring a second face image when the user browses each pushed commodity;
extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index;
and optimizing the commodities in the commodity recommendation combination according to the grading result.
2. The item recommendation method according to claim 1, wherein the determining of the identity information of the user from the first face image, and the determining of the item combination recommended to the user from the identity information includes:
extracting first face feature data in the first face image;
comparing the first face feature data with face registration information in a database, wherein the face registration information corresponds to the identity information one by one;
determining the identity information corresponding to the first face feature data according to the face registration information, wherein the identity information corresponds to historical commodity recommendation combinations one to one;
and determining the corresponding historical commodity recommendation combination according to the identity information, and determining the current commodity recommendation combination according to the historical commodity recommendation combination.
3. The item recommendation method of claim 2, wherein said determining a current item recommendation combination from said historical item recommendation combinations comprises:
when the historical commodity recommendation combination is empty, determining the historical consumption record of the user in a database according to the identity information;
when the historical consumption record of the user is found in the database, determining the current commodity recommendation combination according to the historical consumption record;
and when the historical consumption record of the user is not found in the database, determining a user portrait corresponding to the user according to the facial feature data, and determining the current commodity recommendation combination according to the user portrait.
4. The item recommendation method of claim 2, wherein said determining a current item recommendation combination from said historical item recommendation combinations comprises:
when the historical commodity recommendation combination is not empty, determining commodities liked by the user according to the historical commodity recommendation combination;
determining the similarity between each commodity in the database and each favorite commodity of the user;
determining a similar commodity set of commodities liked by each user according to the similarity, wherein the similar commodity set comprises a plurality of similar commodities of the commodities liked by the user;
calculating the interest degree of the user for each similar commodity according to the predetermined interest degree of the user for the favorite commodities of the user and the similarity;
and determining the current commodity recommendation combination in the similar commodity set according to the interestingness.
5. The item recommendation method according to claim 4, wherein the similarity between each item in the database and each item liked by the user is determined using a first formula, the first formula comprising:
wherein, WijRepresenting the similarity between a commodity i and a commodity j, assuming that the commodity i is a commodity liked by any user, the commodity j is any commodity in a database, N (i) represents the number of commodities i purchased by the user, N (j) represents the number of commodities j purchased by the user, and N (i) n (j) represents the number of commodities i and j purchased by the user;
for any user u, calculating the interest degree of the user u in each similar commodity by adopting a second formula, wherein the second formula comprises the following steps:
Puk＝∑i∈N(u)Wik*ruk，
wherein, PujRepresenting the interest degree of the user u in a commodity k, wherein the commodity k is any similar commodity in the similar commodity set, N (u) represents the historical commodity recommendation set of the user u, and WijIndicates the similarity between the product i and the product k, ruiAnd representing the interest degree of the user u in the commodity i.
6. The item recommendation method according to claim 4, wherein the determining, from the second face image, an emotion index when the user browses each pushed item comprises:
extracting second face feature data in the second face image;
carrying out normalization processing on the second face feature data to obtain normalized face feature data;
and determining the emotion index according to the normalized human face feature data based on a facial animation parameter algorithm.
7. The commodity recommendation method according to claim 5, wherein the emotion indexes include a smile index and a panning index, the head swing angle includes a head raising angle, a plane rotation angle, and a panning angle, the emotion indexes when the user browses each pushed commodity are determined according to the smile and the head swing angle, and scoring each pushed commodity according to the emotion indexes includes:
determining the smiling index and the panning index respectively by adopting a third formula according to the smile and the head swing angle, wherein the third formula comprises:
wherein smile represents the smile index, shakeidedex represents the panning index, V1 represents the smile, V2 represents the head-up angle, V3 represents the plane rotation angle, and V4 represents the panning angle;
and scoring each pushed commodity according to the emotion index by adopting a fourth formula, wherein the fourth formula comprises the following steps:
X＝[5/(1+e-(0.25*smile-1.2*shakeindex)/10)]，
wherein X represents the score of the product, smile represents the smile index, and shakeidedex represents the shake-head index.
8. The item recommendation method according to claim 7, wherein said optimizing the items in the item recommendation group according to the scoring result comprises:
and updating the interest degree of the user for each commodity in the current commodity recommendation combination according to the grade of the commodity.
9. An article recommendation device, comprising:
the acquisition module is used for acquiring a first face image of a user;
the identity recognition module is used for determining identity information of the user according to the first face image and determining a commodity recommendation combination pushed to the user according to the identity information, wherein the commodity recommendation combination comprises a plurality of commodities;
the pushing module is used for pushing the commodities in the commodity recommendation combination to the user one by one and acquiring a second face image when the user browses each pushed commodity;
the grading module is used for extracting the smile and the head swing angle of the user in the second face image, determining the emotion index of the user when the user browses each pushed commodity according to the smile and the head swing angle, and grading each pushed commodity according to the emotion index;
and the optimization module is used for optimizing the commodities in the commodity recommendation combination according to the grading result.
10. A merchandise recommendation device comprising a memory and a processor;
the memory for storing a computer program;
the processor, when executing the computer program, is configured to implement the item recommendation method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011546923.9A CN112581230A (en) | 2020-12-24 | 2020-12-24 | Commodity recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011546923.9A CN112581230A (en) | 2020-12-24 | 2020-12-24 | Commodity recommendation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112581230A true CN112581230A (en) | 2021-03-30 |
Family
ID=75139354
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011546923.9A Pending CN112581230A (en) | 2020-12-24 | 2020-12-24 | Commodity recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112581230A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113254491A (en) * | 2021-06-01 | 2021-08-13 | 平安科技（深圳）有限公司 | Information recommendation method and device, computer equipment and storage medium |
| CN113435970A (en) * | 2021-06-23 | 2021-09-24 | 平安银行股份有限公司 | Product recommendation method and device based on biological information, electronic equipment and medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341688A (en) * | 2017-06-14 | 2017-11-10 | 北京万相融通科技股份有限公司 | The acquisition method and system of a kind of customer experience |
| CN108182624A (en) * | 2017-12-26 | 2018-06-19 | 努比亚技术有限公司 | Method of Commodity Recommendation, server and computer readable storage medium |
| CN108830633A (en) * | 2018-04-26 | 2018-11-16 | 华慧视科技（天津）有限公司 | A kind of friendly service evaluation method based on smiling face's detection |
| CN110147822A (en) * | 2019-04-16 | 2019-08-20 | 北京师范大学 | A kind of moos index calculation method based on the detection of human face action unit |
| CN110363154A (en) * | 2019-07-17 | 2019-10-22 | 安徽航天信息有限公司 | A kind of service quality examining method and system based on Emotion identification |
| CN111080398A (en) * | 2019-11-19 | 2020-04-28 | 浙江大搜车软件技术有限公司 | Commodity recommendation method and device, computer equipment and storage medium |
| CN111104590A (en) * | 2019-11-19 | 2020-05-05 | 泰康保险集团股份有限公司 | Information recommendation method, device, medium and electronic equipment |
| CN111415216A (en) * | 2020-02-11 | 2020-07-14 | 广州探途网络技术有限公司 | Commodity recommendation method and device, server and storage medium |
| CN111724183A (en) * | 2019-03-20 | 2020-09-29 | 中国银联股份有限公司 | Merchant recommendation method and merchant recommendation system |
-
2020
- 2020-12-24 CN CN202011546923.9A patent/CN112581230A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341688A (en) * | 2017-06-14 | 2017-11-10 | 北京万相融通科技股份有限公司 | The acquisition method and system of a kind of customer experience |
| CN108182624A (en) * | 2017-12-26 | 2018-06-19 | 努比亚技术有限公司 | Method of Commodity Recommendation, server and computer readable storage medium |
| CN108830633A (en) * | 2018-04-26 | 2018-11-16 | 华慧视科技（天津）有限公司 | A kind of friendly service evaluation method based on smiling face's detection |
| CN111724183A (en) * | 2019-03-20 | 2020-09-29 | 中国银联股份有限公司 | Merchant recommendation method and merchant recommendation system |
| CN110147822A (en) * | 2019-04-16 | 2019-08-20 | 北京师范大学 | A kind of moos index calculation method based on the detection of human face action unit |
| CN110363154A (en) * | 2019-07-17 | 2019-10-22 | 安徽航天信息有限公司 | A kind of service quality examining method and system based on Emotion identification |
| CN111080398A (en) * | 2019-11-19 | 2020-04-28 | 浙江大搜车软件技术有限公司 | Commodity recommendation method and device, computer equipment and storage medium |
| CN111104590A (en) * | 2019-11-19 | 2020-05-05 | 泰康保险集团股份有限公司 | Information recommendation method, device, medium and electronic equipment |
| CN111415216A (en) * | 2020-02-11 | 2020-07-14 | 广州探途网络技术有限公司 | Commodity recommendation method and device, server and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| 吴梅梅: "《机器学习算法及其应用》", 31 May 2020, 机械工业出版社, pages: 111 - 112 * |
| 陈果，等: ""基于人脸识别的商品推荐系统的设计与实现"", 《计算机时代》, no. 11, pages 52 - 55 * |
| 龙飞: "《基于云计算的电子商务个性化推荐研究》", 30 September 2020, 湖南师范大学出版社, pages: 38 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113254491A (en) * | 2021-06-01 | 2021-08-13 | 平安科技（深圳）有限公司 | Information recommendation method and device, computer equipment and storage medium |
| CN113435970A (en) * | 2021-06-23 | 2021-09-24 | 平安银行股份有限公司 | Product recommendation method and device based on biological information, electronic equipment and medium |
| CN113435970B (en) * | 2021-06-23 | 2023-02-28 | 平安银行股份有限公司 | Product recommendation method and device based on biological information, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11250487B2 (en) | Computer vision and image characteristic search | |
| US20180268458A1 (en) | Automated recommendation and virtualization systems and methods for e-commerce | |
| CN107861972B (en) | Method and equipment for displaying full commodity result after user inputs commodity information | |
| EP3267362B1 (en) | Machine learning image processing | |
| US11423076B2 (en) | Image similarity-based group browsing | |
| US10043109B1 (en) | Attribute similarity-based search | |
| US9607010B1 (en) | Techniques for shape-based search of content | |
| US9348898B2 (en) | Recommendation system with dual collaborative filter usage matrix | |
| US20140379617A1 (en) | Method and system for recommending information | |
| US10467507B1 (en) | Image quality scoring | |
| WO2016029796A1 (en) | Method, device and system for identifying commodity in video image and presenting information thereof | |
| US20220405607A1 (en) | Method for obtaining user portrait and related apparatus | |
| CN106202316A (en) | Merchandise news acquisition methods based on video and device | |
| KR20120085707A (en) | System and method for learning user genres and styles and matching products to user preferences | |
| US10672055B2 (en) | Method and system for presenting personalized products based on digital signage for electronic commerce | |
| CN112581230A (en) | Commodity recommendation method and device | |
| KR20180077959A (en) | Method and apparatus of recommending contents | |
| CN111967924A (en) | Commodity recommendation method, commodity recommendation device, computer device, and medium | |
| CN110648170A (en) | Article recommendation method and related device | |
| CN114154013A (en) | Video recommendation method, device, equipment and storage medium | |
| US20150278910A1 (en) | Directed Recommendations | |
| CN114997956B (en) | Mother and infant product intelligent recommendation system based on big data | |
| KR102323861B1 (en) | System for selling clothing online | |
| KR20200140588A (en) | System and method for providing image-based service to sell and buy product | |
| US11295118B2 (en) | Online user verification without prior knowledge of the user |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |