CN102254028A - Personalized commodity recommendation method and system integrating attributes and structural similarity - Google Patents
Personalized commodity recommendation method and system integrating attributes and structural similarity Download PDFInfo
- Publication number
- CN102254028A CN102254028A CN201110216438XA CN201110216438A CN102254028A CN 102254028 A CN102254028 A CN 102254028A CN 201110216438X A CN201110216438X A CN 201110216438XA CN 201110216438 A CN201110216438 A CN 201110216438A CN 102254028 A CN102254028 A CN 102254028A
- Authority
- CN
- China
- Prior art keywords
- user
- commodity
- information
- similarity
- attribute
- 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 50
- 238000004364 calculation method Methods 0.000 claims abstract description 28
- 238000005516 engineering process Methods 0.000 claims abstract description 11
- 238000005259 measurement Methods 0.000 claims abstract description 9
- 230000010354 integration Effects 0.000 claims abstract description 5
- 238000010586 diagram Methods 0.000 claims description 14
- 238000000691 measurement method Methods 0.000 claims description 8
- 238000013506 data mapping Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000013507 mapping Methods 0.000 claims description 5
- 230000007547 defect Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 claims description 2
- 230000006399 behavior Effects 0.000 claims 6
- 241000531116 Blitum bonus-henricus Species 0.000 claims 1
- 235000008645 Chenopodium bonus henricus Nutrition 0.000 claims 1
- 238000006243 chemical reaction Methods 0.000 claims 1
- 238000010187 selection method Methods 0.000 claims 1
- 238000012163 sequencing technique Methods 0.000 claims 1
- 238000005303 weighing Methods 0.000 claims 1
- 238000000605 extraction Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- BQCADISMDOOEFD-UHFFFAOYSA-N Silver Chemical compound [Ag] BQCADISMDOOEFD-UHFFFAOYSA-N 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 229910052709 silver Inorganic materials 0.000 description 2
- 239000004332 silver Substances 0.000 description 2
- 238000011524 similarity measure Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 229910003460 diamond Inorganic materials 0.000 description 1
- 239000010432 diamond Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a personalized commodity recommendation method integrating attributes and structural similarity. The method integrates attribute and structural similarity information, maps users and commodities as nodes with characteristic information to a network, establishes an information network graph according to the purchasing relation between customers and commodities, measures interest preference between user node pairs by utilizing the integrated attribute and the structural similarity in the information network graph, and selects nearest neighbors according to the interest preference so as to improve recommendation accuracy. On the basis of the recommendation method, the invention also discloses a personalized commodity recommendation system integrating the attribute and the structural similarity measurement. The system accurately measures the interest preference of the user by using a calculation method with similar integration attributes and similar node structure backgrounds in an information network graph, and improves the generation efficiency of nearest neighbors by using a clustering technology in the recommendation process. The method and the system can be used in electronic commerce application and provide personalized commodity recommendation for users.
Description
Technical Field
The invention relates to the technical field of computer internet, in particular to the field of electronic commerce, and particularly relates to a personalized commodity recommendation method and system integrating attributes and structural similarity.
Background
With the continuous development of electronic commerce, the number and the types of commodities are rapidly increased, and in order to find needed commodities as soon as possible, a user hopes a function similar to a shopping guide to help the user to select and purchase proper commodities or services, and a personalized recommendation system is developed. The personalized recommendation is based on mass data analysis and mining technology, and according to the behavior habits and interest characteristics of the user, the interested commodities and information are recommended to the user. At present, almost all large-scale electronic commerce websites integrate personalized recommendation technology into a system to different degrees, wherein the application of collaborative filtering technology is wider.
Recommending based on collaborative filtering, firstly collecting information representing user interests, then selecting neighbor users, and finally predicting the interests of target users to generate a recommendation result. The core problem is to find a group of users with similar interests to the target user, and the similarity between the group of users and the target user is obtained by collecting and comparing behavior selection vectors representing the interests of the two users. At present, a similarity calculation method for comparing behavior selection vectors mainly comprises a poisson correlation coefficient and cosine similarity, which are generally based on a user item scoring matrix and are characterized by simple calculation and easy understanding. However, this scoring matrix based approach makes it difficult for the system to find users of the same interest when performing similarity measurements on users or items because the matrix is sparse. In addition, modeling user interests through scoring does not fully and truly depict user interests, and partial information may be lost. In order to retain more potential information reflecting user interests in data, some methods map users and items to a network as nodes, and perform collaborative recommendation by utilizing structural similarity of the nodes. In addition, some content filtering-based methods measure the similarity between users or items by using attribute description information of the users and the items, and then recommend commodities. However, both the above two calculation methods based on the structural similarity and the attribute similarity lose part of information when recommendation is performed, and according to the similarity enhancement assumption: the similarity between two objects depends not only on their properties but also on the similarity between other objects with which they are related. Therefore, how to combine attribute description information and node structure information when recommending is carried out, a comprehensive and effective similarity measurement method for describing interests among users is provided, and therefore the problem of improving recommendation accuracy by selecting nearest neighbors is needed to be solved.
Disclosure of Invention
In order to solve the above problems, the present invention provides a personalized merchandise recommendation method integrating attributes and structural similarities, which further includes the following steps: collecting basic information of users and commodities of an e-commerce platform and historical purchase transaction records of the users; preprocessing data to obtain basic characteristics, statistical characteristics and behavior characteristics of a user; mapping the user and the commodity as nodes with attribute information to a network, and establishing an information network graph according to the purchasing relationship between the user and the commodity; measuring interest preference among user node pairs by using a measurement method combining attributes and structural similarity in an information network graph; clustering the users by taking the similarity between the node pairs as input so as to narrow the selection range of nearest neighbors and improve the recommendation speed; selecting M nearest neighbor users from a cluster where the active users are located as nearest neighbor users of the active users and generating a near neighbor set; performing predictive scoring on goods that have not been purchased by the active user in the candidate goods database; and performing Top-N recommendation on the active users, and returning the Top-N recommendation as a recommended candidate commodity set.
As an embodiment of the invention, the collection of data comprises basic information of the user, basic information of the commodity, and transaction data of the commodity purchased by the user, namely comment data. The basic information of the user refers to registration information submitted by the user when shopping is carried out on the network, and the registration information comprises a name, registration time and current level when the user registers and comes from a region. The basic information of the commodity refers to description information of the commodity on the shelf, such as a commodity number, a commodity name, a commodity brand, a commodity belonging field, a commodity belonging type and commodity on-shelf time. The transaction data of the commodities purchased by the users comprise commodity numbers, user names, purchasing time, comment advantages, comment defects, comment titles, comment main contents and scores.
As an embodiment of the invention, the preprocessing of the data comprises removing noise data, filling vacancy number items, data normalization and user feature extraction. The user characteristic extraction refers to the fact that statistical characteristics and behavior characteristics are obtained after historical purchase transaction data of a user are counted, and the characteristics comprise historical purchase frequency of the user, enthusiastic brand number, average time interval of purchase time and comment time, average consumption amount, time interval of registration time and new commodity purchase, time interval of commodity shelf time and purchase time, useful frequency and useless frequency of comment publication, the comment is a poor comment proportion, the proportion of commodities purchased in categories, types and fields to which the current purchased commodities belong, the average length of advantages and disadvantages in the comment, the average length of the whole comment, and the proportion of the insufficient comments in the comment as default comments.
As one embodiment of the invention, the construction of the user commodity information network graph is to map data to the network. And representing the user and the commodity as nodes with attributes in the network graph according to the preprocessed data and information, wherein if the user purchases a certain commodity, a directed connecting edge is arranged between the user and the commodity node, and the direction of the edge is from the commodity to the user.
As an embodiment of the invention, the similarity measurement method among users is based on the attribute information and the structure information of the nodes in the graph. The closer the users are if their attribute values are more similar and their historical purchases are more similar, thereby indicating the more similar interests and preferences among the users. The calculation formula of the similarity among users in the recommendation method is as follows:
wherein, <math>
<mrow>
<msub>
<mi>S</mi>
<mi>attribute</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>,</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msup>
<mi>e</mi>
<mrow>
<mrow>
<mrow>
<mo>-</mo>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>p</mi>
</munderover>
</mrow>
<mo>|</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ij</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ik</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>|</mo>
</mrow>
<mo>+</mo>
<mi>μ</mi>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mi>p</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mi>δ</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ij</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ik</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>,</mo>
</mrow>
</math> <math>
<mrow>
<msub>
<mi>S</mi>
<mi>link</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>,</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mi>C</mi>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</munderover>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</munderover>
<msub>
<mi>S</mi>
<mi>link</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>I</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mo>,</mo>
<msub>
<mi>I</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
</math>
at SASimRank(a, b) in the formula Sattribute(a, b) representing the attribute similarity scores of nodes a and b, SlinkAnd (a, b) calculating a structural similarity score according to the connection relation of the a and the b in the information network diagram. At SattributeIn the calculation of (2) xij.ArAnd xik.ArRepresenting an object xijAnd xikThe r-th attribute value of (1); n is the total number of attributes, the 1 st to the p th attributes are numerical attributes or ordered category attributes, and the p +1 st to the N th attributes are unordered category attributes. Function delta (x)ij.Ar，xik.Ar) The function value is 0 when the two attribute values are the same, or 1 when the two attribute values are the same. The parameter mu in the formula is used for adjusting the importance degree of the unordered class attribute relative to the numerical attribute in the whole attribute similarity calculation, and the value range of the parameter mu is [0, 1 ]]In the examples of the present inventionMu is 0.5. Wherein SlinkThe parameter C in the formula (a, b) is between [0, 1]And (c) constant therebetween, which represents the decay rate in the similarity calculation. In the embodiment of the invention, the value of C is 0.8. I (a) and i (b) represent the set of in-degree neighbor nodes for node a and node b, respectively. I isi(a) And Ij(b) Respectively representing the ith in-degree neighbor node of a and the jth in-degree neighbor node of b. | i (a) | and | i (b) | represent the degree of entry of node a and node b, respectively. If the in degree of a or b is 0, SlinkThe value of (a, b) is 0, if a and b represent the same node object, SlinkThe value of (a, b) is 1. From SlinkAs can be seen from the calculation formulas of (a, b), in general, the similarity between a and b is the average value of the similarity between the in-degree neighbor node of a and the in-degree neighbor node of b. SlinkAnd (a, b) calculating a similarity score between the node pairs, wherein the similarity score is symmetrical, namely S (a, b) is S (b, a). By using SlinkAnd (a, b) calculating the structural similarity score between any user node pair a and b in the graph through multiple iterative operations until convergence. Final similarity score S for nodes a and bASimRankAnd (a, b) is determined by the similarity scores. SattributeAnd SlinkThe relative importance of the two is regulated by a parameter lambda in a formula, and the value of lambda is [0, 1%]In the present example, λ is 0.5.
As an embodiment of the present invention, in order to increase the recommendation speed, all users are clustered by using a clustering technique according to the similarity between the users. And narrowing the nearest neighbor search range of the active user from the global user to a certain clustered cluster.
As an embodiment of the invention, the nearest neighbor of the active user is generated by using the clustering result, and M users most similar to the active user are selected from the cluster of the active user as the nearest neighbor users.
As an embodiment of the present invention, the predictive scoring refers to performing predictive scoring on commodities that the active user has not purchased in the candidate commodity database, and weighting the scores of all users in the neighbor set on the target commodities by using a weighted addition method and taking the weighted sum as the score of the active user on the target commodities.
In one embodiment of the present invention, Top-N recommendation refers to ranking the prediction scores of all target commodities of an active user, recommending N commodities with the Top scores to the active user, and returning the recommended candidate commodities as a recommended commodity set.
On the basis of the recommendation method, the invention also provides a personalized commodity recommendation system integrating attributes and structural similarity. The system comprises at least the following components and modules: the system comprises a user terminal, a shared information server, a user and commodity information collector, a user basic information database, a commodity basic information database, a user historical transaction database, a user preference model processor, a data mapping converter, a user preference measure, a user matching degree database, a recommendation accelerator and an individualized recommendation processor.
The user terminal is used for submitting a user recommendation request and returning to a commodity list recommended by the terminal display system. The shared information server is used for storing the information shared by the system. The user and commodity information collector is used for collecting data, and the collected data are respectively stored in the user basic information database, the commodity basic information database and the user historical transaction database. The user preference model processor is used for processing data and establishing a user preference model. The data mapping converter maps the user and the commodity to the network graph to form nodes with attribute information, and constructs the information network graph according to the purchasing relation between the user and the commodity. The user preference measuring device measures the preference among users in the information network diagram by using the measuring method of the integrated attribute and the structural similarity provided by the invention, and the measuring result is stored in a user matching degree database. The recommendation accelerator takes the similarity between user node pairs as input, and utilizes a clustering technology to narrow the search range of nearest neighbors of an active user so as to improve the recommendation efficiency. Based on the set of nearest neighbors, the personalized recommendation processor is operable to predict a predicted score for the active user for the target item.
The invention adopts various characteristics and attributes of the user to establish a user preference model; accurately measuring the interest preference of the user by utilizing a similarity calculation method of similar integration attributes and similar node structure backgrounds in an information network graph; in the recommendation process, the clustering technology is utilized to improve the generation efficiency of nearest neighbors, and the system can finally and quickly respond to the recommendation request of the user in real time and provide personalized commodity recommendation for the client.
Advantages and features of the invention will be set forth in part in the detailed description which follows, or may be learned by practice of the system. According to the invention, the interests and preferences of the users can be deeply mined by combining attributes and structural similarity, and the user interest model can be more accurately constructed, so that accurate recommended contents are generated.
Drawings
FIG. 1 is a flow chart of measuring user interests in a recommendation method of the present invention.
Fig. 2 is an exemplary diagram of user commodity relation information.
Fig. 3 is a schematic diagram of the construction of a user commodity information network diagram.
Fig. 4 is a general flowchart of a personalized goods recommendation method according to the present invention.
FIG. 5 is a flow chart of the operation of a recommended accelerator in an embodiment of the invention.
Fig. 6 is a structural diagram of a personalized goods recommendation system integrating attributes and structural similarities according to the present invention.
Detailed Description
In order to clearly explain the details of the present invention, the present invention will be explained in detail by way of examples with reference to the accompanying drawings.
The invention is mainly characterized in that an effective recommendation method is provided and an efficient and practical personalized recommendation system is designed. The method is characterized in that the interest and the preference of the users in the information network graph are accurately measured by combining attributes and structural similarity, and the searching range of nearest neighbors is narrowed by utilizing a clustering technology. The system responds to the recommendation request of the user in real time and returns a commodity list really interested by the user to the client in time.
Fig. 1 shows a flow of calculating inter-user similarity by combining the measurement method of attribute and structural similarity. The recommendation method disclosed by the invention takes a user interest preference model and basic commodity information as input s 101; mapping user and commodity data to a network to represent nodes, wherein the nodes have attribute information, and the directional connecting edges represent the purchasing relation between the user and the commodity, and if the user purchases a certain commodity, the direction of the connecting edges points to the user from the commodity, so that an information network graph s102 is formed; measuring interest preference similarity between any user node pair in the network graph by using a similarity calculation method combining attribute similarity and structural background similarity s 103; and returning the similarity between all the user node pairs and storing the similarity in the user matching degree database s 104.
As an embodiment of the invention, the calculation process of the measurement method combining the attributes and the structural similarity is described in detail in a small example by constructing an information network diagram by using the data of the commodities purchased by the users in the recommendation system.
The similarity calculation in the invention is based on a directed graph model in an information network, such as G ═ V, E, wherein V represents a node set in the graph, and E represents an edge in the graph<u，v>The set of (a) and (b),<u，v>e, (u, V E V). In the user and commodity information network diagram G, the user and the commodity correspond to a certain node v in the G, and I (v) represents the degree-of-entry neighbor node set of the node v in the inventioni(v) The ith in-degree node representing v, where 1 ≦ i (v) |, where each node object in G has multiple attribute features. The generation of the connection edge in the user commodity information network diagram is realized by the purchase relation between the user and the commodity, and the relation is embodied in a comment table, as shown in FIG. 2As shown. When the user makes a comment on a certain commodity, the user and the commodity are connected by one side. Meanwhile, the user and the commodity node have attribute lists of the user and the commodity node.
As shown in FIG. 3, we take two types of objects, namely user and commodity, in the recommendation system as an example, where P represents commodity, C1, C2 and C3 represent different users respectively, and the diagram shows that commodity P is purchased by users C1, C2 and C3 at the same time. In order to clearly understand the calculation of the similarity between the user nodes in the network diagram, the calculation process of the similarity between the users is described by using the example. For the sake of convenience of calculation, only 4 attributes are considered here, such as the user level, the area where the user is located, the time interval between the time when the user makes a comment and the time when the user purchases a commodity, and the score of the user on the commodity. The attribute values for the 3-bit users are shown in the table below.
| User' s | Rank of | Region(s) | Time interval (sky) | Scoring |
| C1 | D | SC | 10 | 5 |
| C2 | T | EC | 16 | 4 |
| C3 | T | NC | 15 | 4 |
From the example of fig. 3, if only the structural similarity between nodes is considered, the similarity between pairs of { C1, C2}, { C1, C3} and { C2, C3} nodes cannot be distinguished, because they are all pointed to by the same commodity P, and C1, C2, and C3 have the same in-degree neighbor node as commodity P. According to the formula <math>
<mrow>
<msub>
<mi>S</mi>
<mi>link</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>,</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mi>C</mi>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</munderover>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>|</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</munderover>
<msub>
<mi>S</mi>
<mi>link</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>I</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>)</mo>
</mrow>
<mo>,</mo>
<msub>
<mi>I</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
</math> Then when C takes 0.8, the structural similarity values between the { C1, C2}, { C1, C3} and { C2, C3} node pairs are all 0.8. And when the similarity calculation is carried out, the similarity of the { C1, C2}, { C1, C3} and { C2, C3} node pairs is better distinguished by simultaneously considering the attribute characteristics. The calculation process of the similarity score between them will be described in detail below. Before the attribute similarity calculation is carried out by using the formula, the data needs to be processed. If the user current level is quantified by applying ordered numerical values, the user level is totally divided into 6 levels, namely T (diamond member), A (gold member), B (silver member), C (copper member), D (iron member) and E (registered member), and the quantified numerical values corresponding to the levels are 5, 4, 3, 2, 1 and 0. The region attribute belongs to the unordered class attribute value, and only the difference and the identity of the region attribute are judged. According to the formula <math>
<mrow>
<msub>
<mi>S</mi>
<mi>attribute</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>a</mi>
<mo>,</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mrow>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>p</mi>
</munderover>
<mo>|</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ij</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ik</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>|</mo>
</mrow>
<mo>+</mo>
<mi>μ</mi>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mi>p</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mi>δ</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ij</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>ik</mi>
<mo>.</mo>
</mrow>
</msub>
<msub>
<mi>A</mi>
<mi>r</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>,</mo>
</mrow>
</math> The attribute similarity value between pairs of user nodes can be calculated as follows:
Sattribute(C1，C2)＝e-(|1-5|+|10-16|+|5-4|+0.5*1)＝1.013*10-5，
Sattribute(C1，C3)＝e-(|1-5|+|10-15|+|5-4|+0.5*1)＝2.754*10-5，
Sattribute(C2，C3)＝e-(|5-5|+|16-15|+|4-4|+0.5*1)＝0.2231。
using the formula SASimRank(a，b)＝(1-λ)*Sattribute(a，b)+λ*Slink(a, b) when λ is 0.5, the following value is calculated:
SASimRank(C1，C2)＝(1-0.5)*1.013*10-5+0.5*0.8＝0.4+5.065*10-6，
SASimRank(C1，C3)＝(1-0.5)*2.754*10-5+0.5*0.8＝0.4+1.377*10-5，
SASimRank(C2，C3)＝(1-0.5)*0.2231+0.5*0.8＝0.51155。
from the final similarity scores, although C1, C2 and C3 are all pointed to by the same P and have the same in-degree neighbor nodes, the difference between the attribute values of { C1 and C2} is large, so that the similarity score of the two nodes is 0.400005065 at the lowest, and the similarity score of { C2 and C3} is 0.51155 at the highest because the attribute values of the two nodes are basically the same. From this example, it can be illustrated that more real information can be retained in combination with attributes and structural similarity, so that the obtained similarity score can better distinguish the interest preference matching between users.
Fig. 4 is a general flowchart of the personalized goods recommendation method according to the present invention. The whole process is divided into 8 steps, and the brief description of each step is as follows: collecting data s401, preprocessing the data and building a user model s402, converting the data and building an information network graph s403, measuring interest similarity of user nodes s404, generating nearest neighbors s405, selecting the nearest neighbors s406, predicting scores s407, and recommending Top-N commodities s 408.
In more detail, in s401, collecting data refers to sorting user information registered on the e-commerce platform, information of goods on shelves, and a purchase transaction record of the user on a website. All data is stored in a uniform format. If the basic information format of the commodity is as follows: a commodity (No. 143076, name: kingston DDR 313332G, brand: kingston, field: computer product, type: core accessory, shelf life: 2008-12-1117: 18: 45); the format of the basic information of the user is as follows: a user [ user name ═ lihui581203, from shanghai, current level ═ silver membership, registration time ═ 2009-11-13 ]; the purchase transaction of the user is embodied in the comment information, and the format of the comment information is as follows: the product number is 143076, the user name is lihui581203, the purchase time is 2010-05-12, the comment time is 2010-07-0214: 11, the advantages are compatibility and good quality are guaranteed, and the defects are not found temporarily! The title is true and full, and the main content is always full of the product, and the score is 5.
In s402, the data preprocessing refers to filling some vacant data items with default values or average values, and removing some noise data. And storing the user, commodity and comment data into a corresponding database. And carrying out statistics and connection operation among tables on the data by using SQL statements in a database so as to obtain final user attribute characteristics and establish a user model.
In s403, the data is converted and the information network graph is built, goods and users are mapped to the network graph as nodes by using the user model built in s402 and the processed data, and the connection edges between the nodes are generated according to the purchasing relation between the users and the goods. Since the in-degree information of the nodes is required when the structural similarity calculation metric is calculated, the direction of the connecting edge is directed to the user by the commodity. According to the method, an information relation network diagram of the commodity and the user can be constructed. The nodes in the graph are provided with attribute information to provide input for the next calculation of the similarity of the nodes in the network graph.
In s404, the recommendation method in the present invention is based on the user interests, so the user node interest similarity measure is a key step in the process of the method. The invention can completely and accurately depict the interest preference condition among users by adopting a measurement method combining attributes and structural similarity. This similarity measure method is based on the similarity enhancement assumption: the similarity between two data objects depends not only on their properties but also on the similarity between other objects with which they are related. The similarity measurement method provided by the invention considers the structural background information of the nodes in the network graph converted from the data and the attribute information of the nodes, thereby greatly retaining the potential interest information of the user. By utilizing the similarity calculation formula disclosed by the invention, the similarity between any user node pair in the network graph can be finally obtained by carrying out iterative operation on the similarity score between the nodes in the information network graph.
In s405, in order to improve the recommendation speed and greatly reduce the search range of nearest neighbors, the method clusters the users by using a K-modes clustering technology. And positioning the searching range of the nearest neighbor from all the users to a certain user cluster so as to achieve the effect of improving the recommending speed.
The execution of this step s405 is described in detail in conjunction with fig. 5. First, K users are randomly selected from all user nodes as cluster centers, other user objects are distributed to a cluster most similar to a certain cluster center by using the user similarity result stored in s404, and then the cluster center is reselected. In each cluster, a user object is sequentially selected, and the consumption and cost E after the selected object is used for replacing the original cluster center are calculated. The user object with the smallest E is selected to replace the original cluster center as the new cluster center. The circulation is repeated until the convergence condition is met, namely that the centers of the K clusters are not changed. And then, according to the result after clustering, selecting M most similar users from the cluster where the user needing to be recommended is located as nearest neighbors, thereby generating an active user nearest neighbor set.
In s406, the selected M neighbor users are returned using the nearest neighbor set generated in step s 405.
In s407, the prediction score is obtained by weighting the scores of the target product by all the users in the neighborhood by using a weighted addition method, and the weighted sum is used as the score of the target product by the active user. Assumptions are based on active users uaIs equal to { U ═ U-1，u2，...，unThen user uaFor non-scoring commodities tiIs defined as all users in the neighbor set U to the commodity tiThe weighted sum of the score values is,the formula is as follows:
wherein, s (u)a，uk) For active users uaAnd neighbor ukSimilarity of (c); r (u)k，ti) Is ukFor goods tiScoring of (4);
is ukAverage scores for all evaluated items;
for the currently active user uaA priori average score; λ is the normalization coefficient.
In s408, the prediction scores of all the target commodities of the active user are ranked by using the prediction scores calculated in s407, and the Top-N recommendation, i.e., the Top-N recommendation, of N commodities with the Top scores higher than that of the active user is recommended.
Fig. 6 is a structural diagram of a personalized goods recommendation system designed according to the recommendation method disclosed by the invention. The system mainly comprises the following components: the system comprises a user terminal, a shared information server, a user and commodity information collector 600, a user basic information database 610, a commodity basic information database 620, a user historical transaction database 630, a user preference model processor 640, a data mapping converter 650, a user preference measure 660, a user matching degree database 670, a recommendation accelerator 680 and a personalized recommendation processor 690.
As an embodiment of the present invention, the terminal mainly refers to a PC, but is not limited thereto, and may refer to any electronic device having a network communication function, such as a mobile handheld device.
As an embodiment of the invention, the shared information server is used for storing all public information and knowledge of the electronic commerce platform.
The user and goods information collector 600 is used for collecting original registration information of a user on an e-commerce platform, and processing the registration information and storing the registration information into the user basic information database 610 as an embodiment of the invention. The collector also collects the basic information of all the goods on the shelves on the website, preprocesses the basic information of the goods and stores the basic information into the basic information database 620 of the goods. In addition, the method also comprises the comment behavior of the user on the e-commerce website to embody the purchase record and the transaction information of the user, and the information collector 600 monitors the comment behavior of the user and stores the processed data into the user historical transaction database 630.
As an embodiment of the present invention, the user basic information database 610 is used to store the user basic information output by the user and goods information collector 600 as one of the inputs of the processor 640 for establishing the user preference model, where the stored content refers to the registration information submitted by the user when shopping online, and includes the user level, the user from the area, the user name when the user registers, and the registration time.
The basic information database 620 is used for storing basic information of the commodities output by the user and the commodity information collector 600, as an embodiment of the present invention. The stored content refers to description information of goods on the e-commerce website, such as a goods number, a goods name, time for putting the goods on the shelf, a goods price, a goods category, a goods type, a goods belonging field and a goods belonging brand.
As an embodiment of the present invention, the user historical transaction database 630 is used for storing comment information published after the user purchases a commodity, where the comment information includes a user name, a commodity name, a purchase time, a comment publication time, comment content, and the like, and the information includes purchasing behavior characteristics of the user.
As an embodiment of the present invention, the user preference model processor 640 models the user using data stored in the user basic information database 610 and the user historical transaction database 630. The processor comprises the steps of data preprocessing, user characteristic extraction and integration and the like when being executed. Finally, all the characteristics of the user are utilized to build a model for the user so as to accurately describe the user. The establishment of the user preference model is carried out in real time when a user proposes a commodity recommendation request, if the user browses a certain type of commodity, the establishment of the user model is triggered, the user is established by utilizing the user basic information and the user historical purchase transaction record, the establishment of the user model comprises statistical and behavior information besides the user basic information, including the historical commodity purchasing times of the user, the enthusiastic brand numbers, the purchasing time and the average time interval for making comments, the average consumption amount per time, the time interval between the registration time and the purchasing of a new commodity, the time interval between the commodity shelf time and the purchasing time, the useful times and the useless times for making comments by the user, the ratio of bad comments, the ratio of purchasing commodities in the category, the type and the field of the current purchased commodity, the average length of advantages and disadvantages in the comments and the average length of the whole comments, the insufficient comments in the comments are the proportion of the default comments.
As an embodiment of the present invention, the data mapping converter 650 is configured to map the user and the goods with the attribute information as nodes to the network by using the established user preference model, and connect the user nodes and the goods nodes with directed edges according to the purchasing relationship between the user and the goods, thereby constructing the information network graph. Therefore, the commodity user information is converted into the network, and further expressed into the nodes and the connecting edges to form the network graph.
As an embodiment of the present invention, the user preference measure 660 measures interest preference of a user node pair in an information network graph by using a method combining attributes and structural similarities. The metric value depends on the attribute description information and the structure background information of the node, where the structure background information refers to the degree-of-entry neighbor node of the user node pair, and the metric result is used as the input of the user matching degree database 670. Specifically, the user preference measurement comprises the following steps:
aiming at the attribute information of the user nodes, obtaining the attribute similarity between the user node pairs by using an attribute similarity calculation method;
aiming at the structural background information of the user nodes, obtaining the structural similarity between the user node pairs by using a structural similarity calculation method;
and combining the two similarity measurement values and adjusting the weight factor to measure the preference similarity between the end user pairs aiming at the user node pairs with similar integration attributes and structures.
As an embodiment of the present invention, the user matching degree database 670 is used to store similarity values between pairs of user nodes, which is used as a basis for the active user to select the nearest neighbor.
As an embodiment of the present invention, the recommendation accelerator 680 uses data in the user matching degree database 670 as input, and utilizes a clustering technique to greatly narrow the search range of nearest neighbors of a user, thereby improving the real-time performance of recommendation. Specifically, clustering user nodes in a network graph mainly comprises the following steps.
Firstly, randomly selecting M cluster center users;
secondly, distributing the non-cluster-center users to the clusters nearest to the non-cluster-center users;
thirdly, adjusting the cluster center until the clustering result is not changed any more;
and finally, forming user clusters, wherein each user belongs to the cluster in which the user is located.
As an embodiment of the present invention, the personalized recommendation processor 690 scores, based on the user preference model, the commodities that the active user has not purchased in the candidate commodity database by using a weight and sum method, the score of the active user depends on the score of the target commodity by the nearest neighbor user generated in the recommendation accelerator 680, then sorts the commodity prediction scores, and recommends N commodities with top scores to the active user. And feeding back a commodity recommendation request provided by a user in real time, and returning a recommendation result to the user side.
Claims (6)
1. A personalized commodity recommendation method integrating attributes and structural similarity is characterized by comprising the following steps:
step A, collecting user basic information, commodity basic information and user historical purchase record information of an electronic commerce platform;
b, after data are collected, preprocessing the data and extracting attribute features of the user, wherein the preprocessing of the data refers to processing incomplete, noisy and inconsistent data in the data by using a corresponding technology, the attribute features of the user comprise basic attributes, behavior features and statistical features, and a model is built for the user after all features of the user are obtained;
c, mapping the user and the commodity to a network to form nodes by utilizing the attribute characteristics of the user and the commodity and the purchasing relationship between the user and the commodity, and constructing a user and commodity information network graph;
d, performing interest preference measurement on the user model constructed in the step C by using an integrated attribute and structure similarity measurement method;
step E, clustering the user nodes by using the similarity between the user node pairs as input and utilizing a clustering technology, so as to narrow the search range of nearest neighbors and improve the recommendation speed;
step F, generating a nearest neighbor set according to a clustering result, performing descending ordering on the similarity between node pairs, and returning M neighbor users with the top ordering;
g, predicting the scores of the active users on the unevaluated commodities by using the scores of the M neighbor users on the target commodities returned in the step F through the weight addition method;
and H, sequencing the prediction scores of all the target commodities, and recommending N commodities with the grades higher than the previous grades to the active user.
2. The method according to claim 1, wherein said step D comprises in particular the steps of:
step D1, inputting the user commodity information network graph constructed in the step C, wherein the network graph comprises user nodes with attribute characteristic information, commodity nodes and the connection relation among the user nodes and the commodity nodes;
d2, calculating the attribute similarity between user node pairs in the network graph, and selecting different attribute value matching methods according to different attribute types;
d3, calculating the structural similarity among the user node pairs in the network graph, wherein the calculation of the structural similarity depends on the average value of the similarity among the in-degree neighbor nodes of the user node pairs;
d4, selecting a proper weighing factor, and weighting and summing the attribute similarity and the structure similarity by using the results of the steps D2 and D3 to obtain a final similarity score among the user node pairs in the network diagram;
3. the method according to claim 1, wherein step E comprises in particular the steps of:
e1, inputting similarity scores between all user node pairs in the user and commodity information network graph;
step E2, randomly selecting K user nodes in the information network graph as clustering centers;
e3, based on the similarity of the user node pairs in the step E1, allocating other non-cluster-center user objects to the cluster most similar to a certain cluster center;
e4, reselecting the cluster center, the selection method is to select a user object in each cluster in sequence, calculate the consumption and cost E after replacing the original cluster center with the selected object, select the user object with the minimum E to replace the original cluster center as a new cluster center, repeat the step until all the user cluster centers are not changed;
and E5, taking the clustering result in the step E4 as input, sorting the similarity of all users in the user cluster and the current active user in a descending order, and selecting M users in the top order as a nearest neighbor user set.
4. The method according to claim 1, wherein step F comprises in particular the steps of:
step F1, inputting the active user's neighbor set and the user's score for the goods generated in step E;
step F2, weighting the scores of all the users in the neighbor set on the unscored commodities of the active users by using a weight addition method and predicting the scores of the active users on the target commodities;
and F3, returning the prediction scores of the active users for all the target commodities.
5. A personalized goods recommendation system integrating attributes and structural similarities, characterized in that said components and modules comprise: the system comprises a user terminal, a shared information server, a user and commodity information collector, a user basic information database, a commodity basic information database, a user historical transaction database, a user preference model processor, a data mapping converter, a user preference measure, a user matching degree database, a recommendation accelerator and a personalized recommendation processor, and is characterized in that:
the terminal mainly refers to a PC, and can also be any electronic equipment with a network communication function, such as a mobile handheld device and the like;
the shared information server is a computer for storing all public information and knowledge of the electronic commerce platform;
the user and commodity information collector collects original registration information of a user on an electronic commerce platform, processes the registration information and stores the registration information into a user basic information database, the collector also collects basic information of all goods on shelves on a website, pre-processes the goods basic information and stores the goods basic information into a commodity basic information database, in addition, the collector also comprises comment behaviors of the user on the electronic commerce website, and the information collector monitors the comment behaviors of the user and stores the processed data into a user historical transaction database;
the user basic information database is used for storing user basic information such as user names, areas, current levels and registration time output by the user and commodity information collector, and the user basic information is used as one of the inputs of the user preference model;
the commodity basic information database is used for storing commodity basic information output by a user and the commodity information collector, such as commodity numbers, commodity names, commodity brands, commodity fields, commodity types and commodity shelf-loading time;
the user historical transaction database is used for storing transaction data of the user purchasing the commodities, which is output by the user and the commodity information collector, wherein the transaction data includes information such as behavior characteristics of the user, purchasing relations between the user and the commodities and the like, and the information includes: commodity number, user name, purchase time, comment advantages, comment defects, comment titles, comment main contents and scores, if a user makes a comment on a commodity, the user purchases the commodity, and the purchase relation between the user and the commodity and the behavior characteristics expressed by the user purchasing the commodity are respectively used as the input of a data mapping converter and a user preference model processor;
the user preference model processor is used for extracting various types of characteristics of users, such as basic attribute characteristics, statistical characteristics, behavior characteristics and the like, by utilizing data of a user basic information database, a commodity basic information database and a user historical transaction database to build a model for the interest and preference of the users, describing the users by the model, and establishing the user model;
the data mapping converter is used for mapping the user and the commodity with the attribute information as nodes to a network by utilizing the established user preference model, and connecting the user nodes and the commodity nodes by using directed edges according to the purchase relation between the user and the commodity so as to construct an information network graph, thereby realizing the conversion and mapping of the original data to the nodes and the connecting edges in the network graph;
the user preference measurement device measures interest preference of user node pairs in an information network graph by using an integrated attribute and structure similarity method, wherein the measurement value depends on attribute description information and structure background information of the nodes, the structure background information refers to the different conditions of the approach neighbor nodes of the user nodes, and the measurement result is used as the input of a user matching degree database;
the user matching degree database is used for storing similarity values between user node pairs, the data can reflect the similarity condition of interest preference among users, and the data in the user matching degree database can be used as a basis for selecting nearest neighbors by active users;
the recommendation accelerator takes data in a user matching degree database as input, and greatly reduces the searching range of nearest neighbors of a user by utilizing a clustering technology, so that the recommendation efficiency is improved;
the personalized recommendation processor is used for scoring commodities which are not purchased by an active user in a candidate commodity database by using a weight and addition method based on data in a user preference model and a user matching degree database, the score of the active user depends on the score of a nearest neighbor user generated in a recommendation accelerator for a target item, then the predicted scores of the commodities are sorted, N commodities with the scores higher than that of the active user are recommended, a system can feed back a commodity recommendation request provided by the user in real time, and a recommendation result is returned to a user side.
6. The personalized goods recommendation system of claim 5, wherein the user preference similarity calculation method in the user preference measure comprises the steps of:
step I, aiming at attribute information of user nodes, obtaining attribute similarity between user node pairs by using an attribute similarity calculation method;
step J, aiming at the structural background information of the user nodes, obtaining the structural similarity between the user node pairs by using a structural similarity calculation method;
and step K, aiming at the user node pairs with similar integration attributes and structures, combining the two similarity measurement values and adjusting the weight factor to calculate the preference similarity between the final user pairs.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110216438XA CN102254028A (en) | 2011-07-22 | 2011-07-22 | Personalized commodity recommendation method and system integrating attributes and structural similarity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110216438XA CN102254028A (en) | 2011-07-22 | 2011-07-22 | Personalized commodity recommendation method and system integrating attributes and structural similarity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102254028A true CN102254028A (en) | 2011-11-23 |
Family
ID=44981292
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110216438XA Pending CN102254028A (en) | 2011-07-22 | 2011-07-22 | Personalized commodity recommendation method and system integrating attributes and structural similarity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102254028A (en) |
Cited By (104)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102411754A (en) * | 2011-11-29 | 2012-04-11 | 南京大学 | Personalized recommendation method based on commodity property entropy |
| CN102567900A (en) * | 2011-12-28 | 2012-07-11 | 尚明生 | Method for recommending commodities to customers |
| CN102722524A (en) * | 2012-05-07 | 2012-10-10 | 北京邮电大学 | Website recommendation result displaying method and device and terminal with the device |
| CN102750895A (en) * | 2012-06-28 | 2012-10-24 | 宇龙计算机通信科技(深圳)有限公司 | Intelligent building information terminal and system as well as communication method |
| CN103136683A (en) * | 2011-11-24 | 2013-06-05 | 阿里巴巴集团控股有限公司 | Method and device for calculating product reference price and method and system for searching products |
| WO2014036892A1 (en) * | 2012-09-10 | 2014-03-13 | 腾讯科技（深圳）有限公司 | Method and system for generating list of associated recommended items |
| CN103744966A (en) * | 2014-01-07 | 2014-04-23 | Tcl集团股份有限公司 | Item recommendation method and device |
| CN103778214A (en) * | 2014-01-16 | 2014-05-07 | 北京理工大学 | Commodity property clustering method based on user comments |
| CN103793837A (en) * | 2012-10-31 | 2014-05-14 | Ncr公司 | Techniques for recommending a retailer, retail product, or retail services |
| CN103886001A (en) * | 2013-09-22 | 2014-06-25 | 天津思博科科技发展有限公司 | Personalized commodity recommendation system |
| CN104035983A (en) * | 2014-05-29 | 2014-09-10 | 西安理工大学 | Classified variable clustering method based on attribute weight similarity |
| CN104239020A (en) * | 2013-06-21 | 2014-12-24 | Sap欧洲公司 | Decision-making standard driven recommendation |
| CN104361507A (en) * | 2014-11-20 | 2015-02-18 | 携程计算机技术（上海）有限公司 | Commodity recommending method and system |
| CN104392003A (en) * | 2014-12-15 | 2015-03-04 | 北京国双科技有限公司 | Displaying method and device for network search keywords |
| CN104394231A (en) * | 2014-12-10 | 2015-03-04 | 合肥城市云数据中心有限公司 | Data interaction processing method based on intelligent terminal and cloud data technique |
| WO2015027828A1 (en) * | 2013-08-26 | 2015-03-05 | 腾讯科技（深圳）有限公司 | Method and apparatus for implementing clustering of user information |
| CN104463637A (en) * | 2014-12-23 | 2015-03-25 | 北京石油化工学院 | Commodity recommendation method and device based on electronic business platform and server |
| CN104615721A (en) * | 2015-02-06 | 2015-05-13 | 北京京东尚科信息技术有限公司 | Method and system for recommending communities based on returned goods related information |
| CN104615658A (en) * | 2014-12-31 | 2015-05-13 | 中国科学院深圳先进技术研究院 | Method for confirming user identity |
| CN104636419A (en) * | 2013-11-11 | 2015-05-20 | 国际商业机器公司 | Information processing device, information processing method |
| CN104778607A (en) * | 2015-04-13 | 2015-07-15 | 南京八天在线网络科技有限公司 | Data mining system based on campus cloud supermarket |
| CN104778237A (en) * | 2015-04-03 | 2015-07-15 | 电子科技大学 | Individual recommending method and system based on key users |
| CN104809626A (en) * | 2015-03-17 | 2015-07-29 | 徐邑江 | Customized commodity recommending method based on user credit assessment |
| CN104969224A (en) * | 2013-03-13 | 2015-10-07 | 谷歌公司 | Tailoring user experience for unrecognized and new users |
| CN105023175A (en) * | 2015-07-24 | 2015-11-04 | 金鹃传媒科技股份有限公司 | Online advertisement classified pushing method and system based on consumer behavior data analysis and classification technology |
| CN105022614A (en) * | 2014-04-21 | 2015-11-04 | 上海墨盾电脑科技有限公司 | Generic architecture of electronic commerce |
| CN105205702A (en) * | 2015-09-28 | 2015-12-30 | 魔线科技(深圳)有限公司 | Method and system for pushing targeted advertisement based on consumption pattern |
| CN105247552A (en) * | 2013-06-24 | 2016-01-13 | 英特尔公司 | Digital receipt economy |
| CN105260922A (en) * | 2015-10-15 | 2016-01-20 | 广东欧珀移动通信有限公司 | Commodity information matching method and commodity information matching device |
| CN105320719A (en) * | 2015-01-16 | 2016-02-10 | 焦点科技股份有限公司 | Crowdfunding website project recommendation method based on project tag and graphical relationship |
| CN105354729A (en) * | 2015-12-14 | 2016-02-24 | 电子科技大学 | Commodity recommendation method in electronic commerce system |
| CN105528716A (en) * | 2015-12-03 | 2016-04-27 | 山东烟草研究院有限公司 | Tobacco brand remote intelligent recommendation method facing retailer individual need |
| CN105590234A (en) * | 2014-12-30 | 2016-05-18 | 中国银联股份有限公司 | Method and system for recommending commodities to target users |
| CN105653543A (en) * | 2014-11-11 | 2016-06-08 | 阿里巴巴集团控股有限公司 | Method and device for setting user label in information system |
| CN105869024A (en) * | 2016-04-20 | 2016-08-17 | 北京小米移动软件有限公司 | Commodity recommending method and device |
| CN105956048A (en) * | 2016-04-27 | 2016-09-21 | 上海遥薇(集团)有限公司 | Community service big data algorithm mining system |
| CN105989037A (en) * | 2015-02-03 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information display method and device |
| CN106021269A (en) * | 2015-03-27 | 2016-10-12 | 国际商业机器公司 | Transforming social media re-shares to track referrer history and identify influencers |
| WO2016183898A1 (en) * | 2015-05-18 | 2016-11-24 | 向莉妮 | Personalized commodity matching and recommendation system and method, and electronic device |
| CN106227835A (en) * | 2016-07-25 | 2016-12-14 | 中南大学 | Team's research direction method for digging based on two subnetwork figure hierarchical clusterings |
| CN106446943A (en) * | 2016-09-19 | 2017-02-22 | 大连海事大学 | Commodity correlation big data sparse network quick clustering method |
| WO2017028687A1 (en) * | 2015-08-14 | 2017-02-23 | 阿里巴巴集团控股有限公司 | Unused commodity object information processing method and device |
| WO2017050188A1 (en) * | 2015-09-24 | 2017-03-30 | 中国银联股份有限公司 | Transaction data processing method and device |
| CN106570573A (en) * | 2015-10-13 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Parcel attribute information prediction method and device |
| CN106651432A (en) * | 2016-10-31 | 2017-05-10 | 南京魔格信息科技有限公司 | Building advertisement accurate putting system and method |
| CN106779985A (en) * | 2017-02-24 | 2017-05-31 | 武汉奇米网络科技有限公司 | A kind of method and system of personalized commercial sequence |
| WO2017114276A1 (en) * | 2015-12-31 | 2017-07-06 | 中国银联股份有限公司 | User analysis method and system based on image |
| CN106951436A (en) * | 2017-02-09 | 2017-07-14 | 华南理工大学 | A kind of extensive online recommendation method based on mobile contextual |
| CN107016589A (en) * | 2016-08-10 | 2017-08-04 | 阿里巴巴集团控股有限公司 | The determination method and device of recommended products |
| CN107103499A (en) * | 2017-04-26 | 2017-08-29 | 中国人民大学 | Method is recommended in a kind of cross-site cold start-up based on deep learning |
| CN107103028A (en) * | 2017-03-03 | 2017-08-29 | 北京小度信息科技有限公司 | A kind of information processing method and device |
| CN107123016A (en) * | 2017-03-22 | 2017-09-01 | 重庆允升科技有限公司 | A kind of industrial material Method of Commodity Recommendation |
| CN107154109A (en) * | 2017-06-02 | 2017-09-12 | 深圳正品创想科技有限公司 | A kind of commodity rendering method, device and self-service cabinet |
| CN107274247A (en) * | 2017-05-03 | 2017-10-20 | 浙江工商大学 | Wisdom based on cloud computing perceives recommendation method |
| CN107424012A (en) * | 2017-07-31 | 2017-12-01 | 京东方科技集团股份有限公司 | A kind of intelligent shopping guide method, intelligent shopping guide equipment |
| CN107515943A (en) * | 2017-09-01 | 2017-12-26 | 安徽教育网络出版有限公司 | User preferences modeling construction method based on collaborative filtering |
| CN107633073A (en) * | 2017-09-22 | 2018-01-26 | 广州阿里巴巴文学信息技术有限公司 | Resource determining method, apparatus and system |
| CN107665451A (en) * | 2016-07-29 | 2018-02-06 | 百度在线网络技术（北京）有限公司 | A kind of method and device that product push is carried out based on client characteristics data |
| CN107730337A (en) * | 2016-08-12 | 2018-02-23 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
| CN107862566A (en) * | 2017-10-17 | 2018-03-30 | 杨明 | A kind of Method of Commodity Recommendation and system |
| CN107909421A (en) * | 2017-09-29 | 2018-04-13 | 中国船舶重工集团公司第七0九研究所 | A kind of implicit feedback of more GRU layers of neutral net based on user's space recommends method and system |
| CN108140203A (en) * | 2015-08-18 | 2018-06-08 | 万事达卡国际股份有限公司 | For passing through the system and method for property graphical model production Methods |
| CN108154425A (en) * | 2018-01-19 | 2018-06-12 | 广州天源信息科技股份有限公司 | Method is recommended by the Xian Xia trade companies of a kind of combination community network and position |
| CN108230029A (en) * | 2017-12-29 | 2018-06-29 | 西南大学 | Client trading behavior analysis method |
| CN108256018A (en) * | 2018-01-09 | 2018-07-06 | 北京奇艺世纪科技有限公司 | A kind of item recommendation method and device |
| CN105023175B (en) * | 2015-07-24 | 2018-08-31 | 金鹃传媒科技股份有限公司 | A kind of online advertisement classification method for pushing and its system based on consumer behaviour data analysis and classification technology |
| CN108696418A (en) * | 2017-04-06 | 2018-10-23 | 腾讯科技（深圳）有限公司 | Method for secret protection and device in a kind of social networks |
| CN108763496A (en) * | 2018-05-30 | 2018-11-06 | 南京讯高科技有限公司 | A kind of sound state data fusion client segmentation algorithm based on grid and density |
| CN108876537A (en) * | 2018-06-15 | 2018-11-23 | 重庆知遨科技有限公司 | A kind of mixed recommendation method for on-line mall system |
| CN109034960A (en) * | 2018-07-12 | 2018-12-18 | 电子科技大学 | A method of more inferred from attributes based on user node insertion |
| CN109118330A (en) * | 2018-08-09 | 2019-01-01 | 珠海格力电器股份有限公司 | Household appliance recommendation method and device, storage medium and server |
| CN109242656A (en) * | 2018-11-27 | 2019-01-18 | 广东珍三叶电子商务有限公司 | A kind of commodity supplying system based on cloud computing sales data and user's evaluation |
| CN109345348A (en) * | 2018-09-30 | 2019-02-15 | 重庆誉存大数据科技有限公司 | The recommended method of multidimensional information portrait based on travel agency user |
| CN109446402A (en) * | 2017-08-29 | 2019-03-08 | 阿里巴巴集团控股有限公司 | A kind of searching method and device |
| CN109474703A (en) * | 2018-12-24 | 2019-03-15 | 杭州云纪网络科技有限公司 | Personalized product combines method for pushing, apparatus and system |
| CN109615460A (en) * | 2018-11-09 | 2019-04-12 | 北京京东尚科信息技术有限公司 | Gather the selection method and selection system of single commodity |
| CN109670107A (en) * | 2018-12-07 | 2019-04-23 | 杭州飞弛网络科技有限公司 | A kind of stranger's social activity recommended method and system based on interest big data |
| CN109740054A (en) * | 2018-12-27 | 2019-05-10 | 丹翰智能科技(上海)有限公司 | It is a kind of for determining the method and apparatus of the association financial information of target user |
| CN109858985A (en) * | 2017-11-30 | 2019-06-07 | 阿里巴巴集团控股有限公司 | Merchandise news processing, the method shown and device |
| CN109934646A (en) * | 2017-12-15 | 2019-06-25 | 北京京东尚科信息技术有限公司 | Predict the method and device of new commodity complementary buying behavior |
| CN110348876A (en) * | 2019-05-28 | 2019-10-18 | 成都美美臣科技有限公司 | A kind of customer personalized data method for building up of e-commerce website |
| CN110557422A (en) * | 2018-06-01 | 2019-12-10 | 上海沛香信息科技有限公司 | Consumption information pushing method and system and big data platform |
| CN111291904A (en) * | 2020-02-06 | 2020-06-16 | 腾讯科技（深圳）有限公司 | Preference prediction method and device and computer equipment |
| CN111340601A (en) * | 2020-03-30 | 2020-06-26 | 贝壳技术有限公司 | Commodity information recommendation method and device, electronic equipment and storage medium |
| CN111353862A (en) * | 2020-03-30 | 2020-06-30 | 贝壳技术有限公司 | Commodity recommendation method and device, electronic equipment and storage medium |
| WO2020135144A1 (en) * | 2018-12-29 | 2020-07-02 | 北京世纪好未来教育科技有限公司 | Method and apparatus for predicting object preference, and computer-readable medium |
| CN111428143A (en) * | 2020-06-11 | 2020-07-17 | 北京每日优鲜电子商务有限公司 | Commodity recommendation method and system, server and storage medium |
| CN111429161A (en) * | 2020-04-10 | 2020-07-17 | 杭州网易再顾科技有限公司 | Feature extraction method, feature extraction device, storage medium, and electronic apparatus |
| CN111523918A (en) * | 2019-02-02 | 2020-08-11 | 北京极智嘉科技有限公司 | Commodity clustering method, commodity clustering device, commodity clustering equipment and storage medium |
| CN111611496A (en) * | 2020-04-09 | 2020-09-01 | 浙江口碑网络技术有限公司 | Product recommendation method and device |
| CN111951035A (en) * | 2019-05-17 | 2020-11-17 | 上海树融数据科技有限公司 | Consumption analysis method, system, device and consumption analysis platform |
| US20210073732A1 (en) * | 2019-09-11 | 2021-03-11 | Ila Design Group, Llc | Automatically determining inventory items that meet selection criteria in a high-dimensionality inventory dataset |
| CN112487297A (en) * | 2020-12-09 | 2021-03-12 | 贝壳技术有限公司 | Information pushing method and device, electronic equipment and storage medium |
| CN112767085A (en) * | 2021-01-22 | 2021-05-07 | 武汉蔚来能源有限公司 | Commodity similarity analysis and commodity recommendation method, commodity similarity analysis and commodity recommendation device, and computer storage medium |
| CN112847376A (en) * | 2021-02-02 | 2021-05-28 | 广东技术师范大学 | Safe medicine purchasing robot and control method thereof |
| CN113674043A (en) * | 2020-05-15 | 2021-11-19 | 北京沃东天骏信息技术有限公司 | Commodity recommendation method and device, computer-readable storage medium and electronic equipment |
| CN114049180A (en) * | 2021-12-13 | 2022-02-15 | 浙江创邻科技有限公司 | Commodity recommendation method and system based on graph model |
| CN114202395A (en) * | 2022-01-21 | 2022-03-18 | 福州大学 | Online commodity recommendation method and system based on key commodity identification |
| CN114780861A (en) * | 2022-06-20 | 2022-07-22 | 上海二三四五网络科技有限公司 | Clustering technology-based user multi-interest recommendation method, device, equipment and medium |
| CN114912015A (en) * | 2022-03-30 | 2022-08-16 | 阿里巴巴（中国）有限公司 | Object recommendation method, model training method, device, equipment and medium |
| TWI787196B (en) * | 2016-11-29 | 2022-12-21 | 香港商阿里巴巴集團服務有限公司 | Method, device and system for generating business object attribute identification |
| CN117934113A (en) * | 2024-01-19 | 2024-04-26 | 广州千分位科技有限公司 | Commodity recommendation method and device based on big data |
| CN118134610A (en) * | 2024-05-08 | 2024-06-04 | 东莞巨正源科技有限公司 | Commodity recommendation method and system based on user information processing |
| US12051038B2 (en) | 2018-12-30 | 2024-07-30 | Beijing Geekplus Technology Co., Ltd. | Order processing method, apparatus and device, and storage medium |
-
2011
- 2011-07-22 CN CN201110216438XA patent/CN102254028A/en active Pending
Cited By (141)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103136683A (en) * | 2011-11-24 | 2013-06-05 | 阿里巴巴集团控股有限公司 | Method and device for calculating product reference price and method and system for searching products |
| CN103136683B (en) * | 2011-11-24 | 2017-03-01 | 阿里巴巴集团控股有限公司 | Calculate method, device and product search method, the system of product reference price |
| CN102411754A (en) * | 2011-11-29 | 2012-04-11 | 南京大学 | Personalized recommendation method based on commodity property entropy |
| CN102567900A (en) * | 2011-12-28 | 2012-07-11 | 尚明生 | Method for recommending commodities to customers |
| CN102722524A (en) * | 2012-05-07 | 2012-10-10 | 北京邮电大学 | Website recommendation result displaying method and device and terminal with the device |
| CN102750895A (en) * | 2012-06-28 | 2012-10-24 | 宇龙计算机通信科技(深圳)有限公司 | Intelligent building information terminal and system as well as communication method |
| CN102750895B (en) * | 2012-06-28 | 2016-03-09 | 宇龙计算机通信科技(深圳)有限公司 | A kind of intelligent building information terminal and system, exchange method |
| CN103679489A (en) * | 2012-09-10 | 2014-03-26 | 腾讯科技（深圳）有限公司 | Item-based collaborative recommendation method and system |
| CN104685522A (en) * | 2012-09-10 | 2015-06-03 | 上海阅潮网络科技有限公司 | Method and system for generating list of associated recommended items |
| WO2014036892A1 (en) * | 2012-09-10 | 2014-03-13 | 腾讯科技（深圳）有限公司 | Method and system for generating list of associated recommended items |
| CN103793837A (en) * | 2012-10-31 | 2014-05-14 | Ncr公司 | Techniques for recommending a retailer, retail product, or retail services |
| CN104969224A (en) * | 2013-03-13 | 2015-10-07 | 谷歌公司 | Tailoring user experience for unrecognized and new users |
| CN104969224B (en) * | 2013-03-13 | 2020-02-14 | 谷歌有限责任公司 | Improved user experience for unacknowledged and new users |
| CN104239020A (en) * | 2013-06-21 | 2014-12-24 | Sap欧洲公司 | Decision-making standard driven recommendation |
| CN105247552B (en) * | 2013-06-24 | 2020-09-08 | 英特尔公司 | Digital receipt economy |
| CN105247552A (en) * | 2013-06-24 | 2016-01-13 | 英特尔公司 | Digital receipt economy |
| WO2015027828A1 (en) * | 2013-08-26 | 2015-03-05 | 腾讯科技（深圳）有限公司 | Method and apparatus for implementing clustering of user information |
| CN103886001A (en) * | 2013-09-22 | 2014-06-25 | 天津思博科科技发展有限公司 | Personalized commodity recommendation system |
| CN104636419B (en) * | 2013-11-11 | 2018-08-17 | 国际商业机器公司 | Information processing unit, information processing method |
| CN104636419A (en) * | 2013-11-11 | 2015-05-20 | 国际商业机器公司 | Information processing device, information processing method |
| CN103744966B (en) * | 2014-01-07 | 2018-06-22 | Tcl集团股份有限公司 | A kind of item recommendation method, device |
| CN103744966A (en) * | 2014-01-07 | 2014-04-23 | Tcl集团股份有限公司 | Item recommendation method and device |
| CN103778214B (en) * | 2014-01-16 | 2017-08-01 | 北京理工大学 | A kind of item property clustering method based on user comment |
| CN103778214A (en) * | 2014-01-16 | 2014-05-07 | 北京理工大学 | Commodity property clustering method based on user comments |
| CN105022614A (en) * | 2014-04-21 | 2015-11-04 | 上海墨盾电脑科技有限公司 | Generic architecture of electronic commerce |
| CN104035983A (en) * | 2014-05-29 | 2014-09-10 | 西安理工大学 | Classified variable clustering method based on attribute weight similarity |
| CN105653543A (en) * | 2014-11-11 | 2016-06-08 | 阿里巴巴集团控股有限公司 | Method and device for setting user label in information system |
| CN104361507A (en) * | 2014-11-20 | 2015-02-18 | 携程计算机技术（上海）有限公司 | Commodity recommending method and system |
| CN104394231B (en) * | 2014-12-10 | 2018-03-20 | 合肥城市云数据中心有限公司 | A kind of data interactive processing method based on intelligent terminal Yu cloud data technique |
| CN104394231A (en) * | 2014-12-10 | 2015-03-04 | 合肥城市云数据中心有限公司 | Data interaction processing method based on intelligent terminal and cloud data technique |
| CN104392003B (en) * | 2014-12-15 | 2018-09-18 | 北京国双科技有限公司 | Web search keyword shows method and device |
| CN104392003A (en) * | 2014-12-15 | 2015-03-04 | 北京国双科技有限公司 | Displaying method and device for network search keywords |
| CN104463637A (en) * | 2014-12-23 | 2015-03-25 | 北京石油化工学院 | Commodity recommendation method and device based on electronic business platform and server |
| CN105590234A (en) * | 2014-12-30 | 2016-05-18 | 中国银联股份有限公司 | Method and system for recommending commodities to target users |
| CN104615658B (en) * | 2014-12-31 | 2018-01-16 | 中国科学院深圳先进技术研究院 | A kind of method for determining user identity |
| CN104615658A (en) * | 2014-12-31 | 2015-05-13 | 中国科学院深圳先进技术研究院 | Method for confirming user identity |
| CN105320719A (en) * | 2015-01-16 | 2016-02-10 | 焦点科技股份有限公司 | Crowdfunding website project recommendation method based on project tag and graphical relationship |
| CN105320719B (en) * | 2015-01-16 | 2019-02-05 | 焦点科技股份有限公司 | A kind of crowd based on item label and graphics relationship raises website item recommended method |
| CN105989037A (en) * | 2015-02-03 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information display method and device |
| CN104615721A (en) * | 2015-02-06 | 2015-05-13 | 北京京东尚科信息技术有限公司 | Method and system for recommending communities based on returned goods related information |
| CN104615721B (en) * | 2015-02-06 | 2018-02-16 | 北京京东尚科信息技术有限公司 | For the method and system based on return of goods related information Recommendations |
| CN104809626A (en) * | 2015-03-17 | 2015-07-29 | 徐邑江 | Customized commodity recommending method based on user credit assessment |
| US10614471B2 (en) | 2015-03-27 | 2020-04-07 | International Business Machines Corporation | Transforming social media re-shares to track referrer history and identify influencers |
| CN106021269A (en) * | 2015-03-27 | 2016-10-12 | 国际商业机器公司 | Transforming social media re-shares to track referrer history and identify influencers |
| CN106021269B (en) * | 2015-03-27 | 2019-06-14 | 国际商业机器公司 | The method and system shared again for converting social media |
| CN104778237A (en) * | 2015-04-03 | 2015-07-15 | 电子科技大学 | Individual recommending method and system based on key users |
| CN104778607A (en) * | 2015-04-13 | 2015-07-15 | 南京八天在线网络科技有限公司 | Data mining system based on campus cloud supermarket |
| WO2016183898A1 (en) * | 2015-05-18 | 2016-11-24 | 向莉妮 | Personalized commodity matching and recommendation system and method, and electronic device |
| CN105023175B (en) * | 2015-07-24 | 2018-08-31 | 金鹃传媒科技股份有限公司 | A kind of online advertisement classification method for pushing and its system based on consumer behaviour data analysis and classification technology |
| CN105023175A (en) * | 2015-07-24 | 2015-11-04 | 金鹃传媒科技股份有限公司 | Online advertisement classified pushing method and system based on consumer behavior data analysis and classification technology |
| WO2017028687A1 (en) * | 2015-08-14 | 2017-02-23 | 阿里巴巴集团控股有限公司 | Unused commodity object information processing method and device |
| CN108140203A (en) * | 2015-08-18 | 2018-06-08 | 万事达卡国际股份有限公司 | For passing through the system and method for property graphical model production Methods |
| CN108140203B (en) * | 2015-08-18 | 2022-06-03 | 万事达卡国际股份有限公司 | System and method for generating relationships through a property graph model |
| WO2017050188A1 (en) * | 2015-09-24 | 2017-03-30 | 中国银联股份有限公司 | Transaction data processing method and device |
| CN105205702A (en) * | 2015-09-28 | 2015-12-30 | 魔线科技(深圳)有限公司 | Method and system for pushing targeted advertisement based on consumption pattern |
| CN106570573B (en) * | 2015-10-13 | 2022-05-27 | 菜鸟智能物流控股有限公司 | Method and device for predicting package attribute information |
| CN106570573A (en) * | 2015-10-13 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Parcel attribute information prediction method and device |
| CN105260922A (en) * | 2015-10-15 | 2016-01-20 | 广东欧珀移动通信有限公司 | Commodity information matching method and commodity information matching device |
| CN105528716A (en) * | 2015-12-03 | 2016-04-27 | 山东烟草研究院有限公司 | Tobacco brand remote intelligent recommendation method facing retailer individual need |
| CN105354729A (en) * | 2015-12-14 | 2016-02-24 | 电子科技大学 | Commodity recommendation method in electronic commerce system |
| WO2017114276A1 (en) * | 2015-12-31 | 2017-07-06 | 中国银联股份有限公司 | User analysis method and system based on image |
| CN105869024A (en) * | 2016-04-20 | 2016-08-17 | 北京小米移动软件有限公司 | Commodity recommending method and device |
| CN105956048A (en) * | 2016-04-27 | 2016-09-21 | 上海遥薇(集团)有限公司 | Community service big data algorithm mining system |
| CN106227835A (en) * | 2016-07-25 | 2016-12-14 | 中南大学 | Team's research direction method for digging based on two subnetwork figure hierarchical clusterings |
| CN106227835B (en) * | 2016-07-25 | 2018-01-19 | 中南大学 | Team's research direction method for digging based on two subnetwork figure hierarchical clusterings |
| CN107665451B (en) * | 2016-07-29 | 2021-09-21 | 百度在线网络技术（北京）有限公司 | Method and device for pushing product based on customer characteristic data |
| CN107665451A (en) * | 2016-07-29 | 2018-02-06 | 百度在线网络技术（北京）有限公司 | A kind of method and device that product push is carried out based on client characteristics data |
| CN107016589A (en) * | 2016-08-10 | 2017-08-04 | 阿里巴巴集团控股有限公司 | The determination method and device of recommended products |
| CN107730337A (en) * | 2016-08-12 | 2018-02-23 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
| CN106446943A (en) * | 2016-09-19 | 2017-02-22 | 大连海事大学 | Commodity correlation big data sparse network quick clustering method |
| CN106651432A (en) * | 2016-10-31 | 2017-05-10 | 南京魔格信息科技有限公司 | Building advertisement accurate putting system and method |
| TWI787196B (en) * | 2016-11-29 | 2022-12-21 | 香港商阿里巴巴集團服務有限公司 | Method, device and system for generating business object attribute identification |
| CN106951436B (en) * | 2017-02-09 | 2020-06-19 | 华南理工大学 | Large-scale online recommendation method based on mobile situation |
| CN106951436A (en) * | 2017-02-09 | 2017-07-14 | 华南理工大学 | A kind of extensive online recommendation method based on mobile contextual |
| CN106779985A (en) * | 2017-02-24 | 2017-05-31 | 武汉奇米网络科技有限公司 | A kind of method and system of personalized commercial sequence |
| CN107103028A (en) * | 2017-03-03 | 2017-08-29 | 北京小度信息科技有限公司 | A kind of information processing method and device |
| CN107123016A (en) * | 2017-03-22 | 2017-09-01 | 重庆允升科技有限公司 | A kind of industrial material Method of Commodity Recommendation |
| CN108696418A (en) * | 2017-04-06 | 2018-10-23 | 腾讯科技（深圳）有限公司 | Method for secret protection and device in a kind of social networks |
| CN107103499A (en) * | 2017-04-26 | 2017-08-29 | 中国人民大学 | Method is recommended in a kind of cross-site cold start-up based on deep learning |
| CN107274247A (en) * | 2017-05-03 | 2017-10-20 | 浙江工商大学 | Wisdom based on cloud computing perceives recommendation method |
| CN107154109A (en) * | 2017-06-02 | 2017-09-12 | 深圳正品创想科技有限公司 | A kind of commodity rendering method, device and self-service cabinet |
| CN107424012A (en) * | 2017-07-31 | 2017-12-01 | 京东方科技集团股份有限公司 | A kind of intelligent shopping guide method, intelligent shopping guide equipment |
| CN109446402B (en) * | 2017-08-29 | 2022-04-01 | 阿里巴巴集团控股有限公司 | Searching method and device |
| CN109446402A (en) * | 2017-08-29 | 2019-03-08 | 阿里巴巴集团控股有限公司 | A kind of searching method and device |
| CN107515943A (en) * | 2017-09-01 | 2017-12-26 | 安徽教育网络出版有限公司 | User preferences modeling construction method based on collaborative filtering |
| CN107633073A (en) * | 2017-09-22 | 2018-01-26 | 广州阿里巴巴文学信息技术有限公司 | Resource determining method, apparatus and system |
| CN107909421A (en) * | 2017-09-29 | 2018-04-13 | 中国船舶重工集团公司第七0九研究所 | A kind of implicit feedback of more GRU layers of neutral net based on user's space recommends method and system |
| CN107862566A (en) * | 2017-10-17 | 2018-03-30 | 杨明 | A kind of Method of Commodity Recommendation and system |
| CN109858985A (en) * | 2017-11-30 | 2019-06-07 | 阿里巴巴集团控股有限公司 | Merchandise news processing, the method shown and device |
| CN109934646B (en) * | 2017-12-15 | 2021-09-17 | 北京京东尚科信息技术有限公司 | Method and device for predicting associated purchasing behavior of new commodity |
| CN109934646A (en) * | 2017-12-15 | 2019-06-25 | 北京京东尚科信息技术有限公司 | Predict the method and device of new commodity complementary buying behavior |
| CN108230029A (en) * | 2017-12-29 | 2018-06-29 | 西南大学 | Client trading behavior analysis method |
| CN108256018A (en) * | 2018-01-09 | 2018-07-06 | 北京奇艺世纪科技有限公司 | A kind of item recommendation method and device |
| CN108154425B (en) * | 2018-01-19 | 2021-03-26 | 广州天源信息科技股份有限公司 | Offline merchant recommendation method combining social network and location |
| CN108154425A (en) * | 2018-01-19 | 2018-06-12 | 广州天源信息科技股份有限公司 | Method is recommended by the Xian Xia trade companies of a kind of combination community network and position |
| CN108763496A (en) * | 2018-05-30 | 2018-11-06 | 南京讯高科技有限公司 | A kind of sound state data fusion client segmentation algorithm based on grid and density |
| CN110557422A (en) * | 2018-06-01 | 2019-12-10 | 上海沛香信息科技有限公司 | Consumption information pushing method and system and big data platform |
| CN108876537B (en) * | 2018-06-15 | 2021-08-17 | 上海百秋电子商务有限公司 | Mixed recommendation method for online marketplace system |
| CN108876537A (en) * | 2018-06-15 | 2018-11-23 | 重庆知遨科技有限公司 | A kind of mixed recommendation method for on-line mall system |
| CN109034960A (en) * | 2018-07-12 | 2018-12-18 | 电子科技大学 | A method of more inferred from attributes based on user node insertion |
| CN109034960B (en) * | 2018-07-12 | 2022-09-06 | 电子科技大学 | Multi-attribute inference method based on user node embedding |
| CN109118330A (en) * | 2018-08-09 | 2019-01-01 | 珠海格力电器股份有限公司 | Household appliance recommendation method and device, storage medium and server |
| CN109118330B (en) * | 2018-08-09 | 2020-09-22 | 珠海格力电器股份有限公司 | Household appliance recommendation method and device, storage medium and server |
| CN109345348A (en) * | 2018-09-30 | 2019-02-15 | 重庆誉存大数据科技有限公司 | The recommended method of multidimensional information portrait based on travel agency user |
| CN109615460A (en) * | 2018-11-09 | 2019-04-12 | 北京京东尚科信息技术有限公司 | Gather the selection method and selection system of single commodity |
| CN109242656A (en) * | 2018-11-27 | 2019-01-18 | 广东珍三叶电子商务有限公司 | A kind of commodity supplying system based on cloud computing sales data and user's evaluation |
| CN109670107A (en) * | 2018-12-07 | 2019-04-23 | 杭州飞弛网络科技有限公司 | A kind of stranger's social activity recommended method and system based on interest big data |
| CN109474703A (en) * | 2018-12-24 | 2019-03-15 | 杭州云纪网络科技有限公司 | Personalized product combines method for pushing, apparatus and system |
| CN109740054A (en) * | 2018-12-27 | 2019-05-10 | 丹翰智能科技(上海)有限公司 | It is a kind of for determining the method and apparatus of the association financial information of target user |
| WO2020135144A1 (en) * | 2018-12-29 | 2020-07-02 | 北京世纪好未来教育科技有限公司 | Method and apparatus for predicting object preference, and computer-readable medium |
| US11651255B2 (en) | 2018-12-29 | 2023-05-16 | Beijing Century Tal Education Technology Co., Ltd. | Method and apparatus for object preference prediction, and computer readable medium |
| US12051038B2 (en) | 2018-12-30 | 2024-07-30 | Beijing Geekplus Technology Co., Ltd. | Order processing method, apparatus and device, and storage medium |
| CN111523918B (en) * | 2019-02-02 | 2023-09-19 | 北京极智嘉科技股份有限公司 | Commodity clustering method, device, equipment and storage medium |
| CN111523918A (en) * | 2019-02-02 | 2020-08-11 | 北京极智嘉科技有限公司 | Commodity clustering method, commodity clustering device, commodity clustering equipment and storage medium |
| CN111951035B (en) * | 2019-05-17 | 2024-06-11 | 嘉兴树融数据科技有限公司 | Consumption analysis method, system, device and platform |
| CN111951035A (en) * | 2019-05-17 | 2020-11-17 | 上海树融数据科技有限公司 | Consumption analysis method, system, device and consumption analysis platform |
| CN110348876A (en) * | 2019-05-28 | 2019-10-18 | 成都美美臣科技有限公司 | A kind of customer personalized data method for building up of e-commerce website |
| US11494734B2 (en) * | 2019-09-11 | 2022-11-08 | Ila Design Group Llc | Automatically determining inventory items that meet selection criteria in a high-dimensionality inventory dataset |
| US20210073732A1 (en) * | 2019-09-11 | 2021-03-11 | Ila Design Group, Llc | Automatically determining inventory items that meet selection criteria in a high-dimensionality inventory dataset |
| CN111291904B (en) * | 2020-02-06 | 2023-04-18 | 腾讯科技（深圳）有限公司 | Preference prediction method and device and computer equipment |
| CN111291904A (en) * | 2020-02-06 | 2020-06-16 | 腾讯科技（深圳）有限公司 | Preference prediction method and device and computer equipment |
| CN111340601A (en) * | 2020-03-30 | 2020-06-26 | 贝壳技术有限公司 | Commodity information recommendation method and device, electronic equipment and storage medium |
| CN111353862B (en) * | 2020-03-30 | 2024-03-26 | 贝壳技术有限公司 | Commodity recommendation method and device, electronic equipment and storage medium |
| CN111353862A (en) * | 2020-03-30 | 2020-06-30 | 贝壳技术有限公司 | Commodity recommendation method and device, electronic equipment and storage medium |
| CN111611496A (en) * | 2020-04-09 | 2020-09-01 | 浙江口碑网络技术有限公司 | Product recommendation method and device |
| CN111429161A (en) * | 2020-04-10 | 2020-07-17 | 杭州网易再顾科技有限公司 | Feature extraction method, feature extraction device, storage medium, and electronic apparatus |
| CN111429161B (en) * | 2020-04-10 | 2023-10-10 | 杭州网易再顾科技有限公司 | Feature extraction method, feature extraction device, storage medium and electronic equipment |
| CN113674043A (en) * | 2020-05-15 | 2021-11-19 | 北京沃东天骏信息技术有限公司 | Commodity recommendation method and device, computer-readable storage medium and electronic equipment |
| CN111428143A (en) * | 2020-06-11 | 2020-07-17 | 北京每日优鲜电子商务有限公司 | Commodity recommendation method and system, server and storage medium |
| CN112487297A (en) * | 2020-12-09 | 2021-03-12 | 贝壳技术有限公司 | Information pushing method and device, electronic equipment and storage medium |
| CN112767085A (en) * | 2021-01-22 | 2021-05-07 | 武汉蔚来能源有限公司 | Commodity similarity analysis and commodity recommendation method, commodity similarity analysis and commodity recommendation device, and computer storage medium |
| CN112767085B (en) * | 2021-01-22 | 2024-05-24 | 武汉蔚来能源有限公司 | Commodity similarity analysis and commodity recommendation method and device and computer storage medium |
| CN112847376A (en) * | 2021-02-02 | 2021-05-28 | 广东技术师范大学 | Safe medicine purchasing robot and control method thereof |
| CN114049180A (en) * | 2021-12-13 | 2022-02-15 | 浙江创邻科技有限公司 | Commodity recommendation method and system based on graph model |
| CN114202395A (en) * | 2022-01-21 | 2022-03-18 | 福州大学 | Online commodity recommendation method and system based on key commodity identification |
| CN114202395B (en) * | 2022-01-21 | 2024-07-23 | 福州大学 | Online commodity recommendation method and system based on key commodity identification |
| CN114912015A (en) * | 2022-03-30 | 2022-08-16 | 阿里巴巴（中国）有限公司 | Object recommendation method, model training method, device, equipment and medium |
| WO2023245999A1 (en) * | 2022-06-20 | 2023-12-28 | 上海二三四五网络科技有限公司 | User multi-interest recommendation method and apparatus based on clustering technology, and device and medium |
| CN114780861A (en) * | 2022-06-20 | 2022-07-22 | 上海二三四五网络科技有限公司 | Clustering technology-based user multi-interest recommendation method, device, equipment and medium |
| CN117934113A (en) * | 2024-01-19 | 2024-04-26 | 广州千分位科技有限公司 | Commodity recommendation method and device based on big data |
| CN118134610A (en) * | 2024-05-08 | 2024-06-04 | 东莞巨正源科技有限公司 | Commodity recommendation method and system based on user information processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102254028A (en) | Personalized commodity recommendation method and system integrating attributes and structural similarity | |
| CN102789462B (en) | A kind of item recommendation method and system | |
| CN105868281B (en) | Location aware recommender system based on non-dominated ranking multi-target method | |
| US9569499B2 (en) | Method and apparatus for recommending content on the internet by evaluating users having similar preference tendencies | |
| CN110020128B (en) | Search result ordering method and device | |
| CN103377250B (en) | Top k based on neighborhood recommend method | |
| CN107562818A (en) | Information recommendation system and method | |
| CN101853463A (en) | Collaborative filtering recommending method and system based on client characteristics | |
| CN106600372A (en) | Commodity recommending method and system based on user behaviors | |
| CN101454771A (en) | System and method of segmenting and tagging entities based on profile matching using a multi-media survey | |
| US11321724B1 (en) | Product evaluation system and method of use | |
| CN104731866B (en) | Personalized food recommendation method based on position | |
| CN101482884A (en) | Cooperation recommending system based on user predilection grade distribution | |
| CN112131480A (en) | Personalized commodity recommendation method and system based on multilayer heterogeneous attribute network representation learning | |
| CN103136683A (en) | Method and device for calculating product reference price and method and system for searching products | |
| CN113592588B (en) | E-commerce platform commodity recommendation system and method based on big data collaborative filtering technology | |
| CN104778237A (en) | Individual recommending method and system based on key users | |
| CN111949887A (en) | Item recommendation method and device and computer-readable storage medium | |
| CN115860880B (en) | Personalized commodity recommendation method and system based on multi-layer heterogeneous graph convolution model | |
| CN107545457B (en) | Automobile racing product type determination method and device | |
| CN109977299A (en) | A kind of proposed algorithm of convergence project temperature and expert's coefficient | |
| CN110347923B (en) | Traceable fast fission type user portrait construction method | |
| CN115829683A (en) | Power integration commodity recommendation method and system based on inverse reward learning optimization | |
| An et al. | A two-stage multiple-factor aware method for travel product recommendation | |
| CN105590234A (en) | Method and system for recommending commodities to target users |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20111123 |