CN115659013A - User recommendation method and device - Google Patents
User recommendation method and device Download PDFInfo
- Publication number
- CN115659013A CN115659013A CN202211263075.XA CN202211263075A CN115659013A CN 115659013 A CN115659013 A CN 115659013A CN 202211263075 A CN202211263075 A CN 202211263075A CN 115659013 A CN115659013 A CN 115659013A
- Authority
- CN
- China
- Prior art keywords
- user
- preference
- feature
- feature set
- database
- 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
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses a user recommendation method and a user recommendation device, wherein the method comprises the following steps: obtaining historical operation data of a first user; and selecting a user set with the feature set having the highest similarity with the preference feature set from a user database, wherein the user set comprises user information of at least one user. The scheme is based on the user database, data processing based on text content is achieved, intelligent matching of users with different genres is achieved through user feature extraction, feature learning and user recommendation, and accuracy of individual user recommendation functions is greatly improved.
Description
Technical Field
The invention relates to a data processing technology, in particular to a user recommendation method and device.
Background
In the friend-making platform, as most users do not want to expose their true information on the internet too much, the feature extraction of the users is not accurate, which results in that the accuracy of the friend-making recommendation function using the content-based recommendation algorithm is low, and most users with incomplete information are directly filtered. Therefore, most of the existing friend-making recommendation platforms in the industry adopt the following technical schemes: 1. and analyzing the correlation degree of a certain user with other users according to the basic information of the user by using a friend making platform based on a recommendation algorithm of demographics, and recommending the user according to the preference of the user with high correlation degree. 2. And recommending users with higher popularity to the users 3 by using the friend making platform based on the popularity recommendation algorithm according to the preference of most users, wherein the proportion of the picture data in the characteristic weight of most friend making platforms is higher, namely the head portrait and the picture of the users are the key points of the characteristics of the users.
However, the above solutions have their limitations and disadvantages, and the accuracy of the recommended users is still not good enough. As for a friend-making platform using a demographic-based recommendation algorithm, it requires a large number of user bases and its recommendation function is less accurate due to inaccuracy of internet user information; aiming at a friend making platform using a recommendation algorithm based on popularity, the recommendation method only pushes popular users and cannot realize a customized recommendation function; the recommendation algorithm of most friend-making platforms has a higher weight occupied by the picture data, that is, the correlation between the feature data set of the user and the picture data thereof is higher. However, the authenticity of pictures in internet applications is low, resulting in a low correlation between the user's feature data and its real information.
Disclosure of Invention
In view of this, the present invention provides the following technical solutions:
a user recommendation method includes:
obtaining historical operation data of a first user;
determining a set of preference features for the first user based on the historical operational data;
and selecting a user set with the highest similarity between the feature set and the preference feature set from a user database, wherein the user set comprises user information of at least one user.
Optionally, the method further comprises:
and carrying out feature extraction on the user information of all users in the user database in advance to obtain a feature set of each user.
Optionally, the pre-extracting the features of the user information of all the users in the user database to obtain a feature set of each user includes:
extracting key information in the user information of each user in the user database based on a TF-IDF algorithm;
and processing based on the key information to obtain a feature set of each user.
Optionally, the determining a set of preference features of the first user based on the historical operational data includes:
and constructing a user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the user portrait.
Optionally, the selecting, from the user database, a user set with a feature set having the highest similarity to the preference feature set includes:
and respectively carrying out Euclidean similarity calculation on the feature sets of all the users in the user database and the preference feature set to obtain a user set with the highest similarity.
Optionally, after selecting the user set with the feature set having the highest similarity to the preference feature set from the user database, the method further includes:
and displaying the user information of each user in the user set based on the high and low output of the similarity between the feature set and the preference feature set.
A user recommendation device comprising:
the historical data acquisition module is used for acquiring historical operation data of a first user;
a preference feature determination module for determining a set of preference features for the first user based on the historical operating data;
and the recommendation user determining module is used for selecting a user set with a feature set having the highest similarity with the preference feature set from a user database, wherein the user set comprises user information of at least one user.
Optionally, the method further comprises:
and the user characteristic extraction module is used for extracting the characteristics of the user information of all the users in the user database in advance to obtain a characteristic set of each user.
Optionally, the user feature extraction module is specifically configured to: extracting key information in the user information of each user in the user database based on a TF-IDF algorithm; and processing based on the key information to obtain a feature set of each user.
Optionally, the preference characteristic determining module is specifically configured to: and constructing a user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the user portrait.
The technical scheme shows that the embodiment of the invention discloses a user recommendation method and a device, and the method comprises the following steps: obtaining historical operation data of a first user; and selecting a user set with the feature set having the highest similarity with the preference feature set from a user database, wherein the user set comprises user information of at least one user. The scheme is based on the user database, data processing based on text content is achieved, intelligent matching of users with different genres is achieved through user feature extraction, feature learning and user recommendation, and accuracy of individual user recommendation functions is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flowchart of a user recommendation method according to an embodiment of the present invention;
FIG. 2 is a flowchart of extracting a user feature set according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a recommendation process logic of a user recommendation method disclosed in an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a user recommendation device according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device disclosed in an embodiment of the present application.
Detailed Description
For the sake of reference and clarity, the descriptions, abbreviations or abbreviations of the technical terms used hereinafter are summarized as follows:
is characterized in that: a label corresponding to each object, such as a person's height, a category of the article, etc.
Feature extraction: and extracting corresponding characteristic data for each object, such as extracting abstract, key words, reference materials and the like of the periodical.
And (3) feature learning: the method comprises the steps of recording preference data of a certain user for different features, and learning a preference feature set of the user. For example, if a user prefers sports news and does not like politics news, the characteristics of the sports news intersect with the user's set of preference characteristics.
And (3) recommendation generation: and recommending a group of candidate users with the highest relevance for a certain user by comparing the preference feature set of the user with the features of the candidate objects.
And (3) complete matching: and in a certain data set, extracting the data with the matching content strictly consistent with the matched content.
Structured data and unstructured data: structured data is a range of values, such as height, weight, and gender, that fall within a particular range of values. The unstructured data is data in a non-specific value range, such as personalized signatures of users and article contents.
Euclidean similarity algorithm: the similarity of the two data objects is judged by calculating the absolute path of the two data objects in the multidimensional space.
Robustness: the finger algorithm still has higher accuracy under the condition that a small part of data is wrong.
TF-IDF: the term frequency-inverse document frequency, an abbreviation, is a commonly used weighting technique for information retrieval and data mining. Where TF is Term Frequency (Term Frequency) and IDF is Inverse text Frequency index (Inverse Document Frequency).
The Rocchio algorithm is an efficient classification algorithm and is widely applied to the fields of text classification, query expansion and the like. The optimal solution is obtained by a method of constructing a prototype vector.
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart of a user recommendation method disclosed in an embodiment of the present invention. Referring to fig. 1, the user recommendation method may include:
step 101: historical operating data of the first user is obtained.
The historical operation data may be operations of the user browsing and searching for the interested user. The historical operating data can reflect the preference characteristics of the first user to a certain extent.
Step 102: determining a set of preference features for the first user based on the historical operational data.
For example, the user a may have a feature that is preferred by the first user if the user a is browsing the information introduction of the user a in detail; alternatively, the user may search for keywords, such as age value, height value, etc., to determine the friend-making criteria of the first user.
Step 103: and selecting a user set with the highest similarity between the feature set and the preference feature set from a user database, wherein the user set comprises user information of at least one user.
In this step, similarity calculation is performed on feature sets of all candidate recommenders, which satisfy basic requirements, such as age requirements and height requirements, in the user database and the preference feature set respectively to obtain the similarity corresponding to each candidate recommender, n candidate recommenders with the highest similarity are selected as final recommenders, and a set formed by the n final recommenders is the user set.
The user recommendation method in this embodiment is based on a user database, realizes data processing based on text content, and realizes customized friend-making recommendation schemes for different users by extracting, analyzing and comparing features of individual users. The accuracy of the user recommendation function is greatly improved, the satisfaction of the user on the platform is improved, and the user friend making target is realized more quickly.
In one implementation, the user recommendation method may further include: and carrying out feature extraction on the user information of all users in the user database in advance to obtain a feature set of each user.
In the implementation, the user information of each user can be immediately subjected to feature extraction after being put in storage, so that the user can be timely matched with a recommended user or matched and recommended to other users.
Fig. 2 is a flowchart of extracting a user feature set according to an embodiment of the present invention. With reference to fig. 2, the pre-extracting the features of the user information of all users in the user database to obtain the feature set of each user may specifically include:
step 201: and extracting key information in the user information of each user in the user database based on a TF-IDF algorithm.
Step 202: and processing based on the key information to obtain a feature set of each user.
The TF-IDF algorithm is often used for extracting key words in documents. The Term Frequency (TF) is the quotient of the number of occurrences of a word in an article and the total number of articles. The Inverse Document Frequency (IDF) is to add different weights to each word, so as to filter some words which are more and less important, such as "the", "i", "y", and so on, and the product of the two is the importance degree of a word in the article. And taking the maximum product of 10 words as the key words of the signature and friend making requirements of the user, and combining all structured data of the user as a characteristic data set of the user.
Logic expression:
word frequency (TF) = number of occurrences of a word in a text/total number of words in a text;
inverse Document Frequency (IDF) = log (total number of texts of all users/number of documents containing a certain word + 1);
TF-IDF = word frequency and inverse document frequency.
Let n i,j For word i appears in all unstructured text sets j of a certain userNumber of times, n *,j Is the total number of words in the text set. Let N be the total number of the user's unstructured text sets, N i Is the total number of occurrences of i in all text sets. Then there is a change in the number of,
wherein TF-IDF represents the degree of importance of the word i.
In one implementation, the determining the set of preference features for the first user based on the historical operational data may include: and constructing a user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the user portrait.
The rocchi algorithm is often used to calculate the user preference characteristics for articles. And setting corresponding feedback weights to calculate a corresponding query vector through the feedback of a certain user to other users of different genres. And performing Euclidean similarity calculation on the query vector and the feature vectors of other users with different characteristics to judge the possible preference degree of the user to a certain user with different characteristics, and taking N users with the highest similarity as a result set to output as result data.
Logic expression:
preference feature vector of a certain user = (weight of positive feedback × (sum of feature vectors of all the anisotropic users that the user is interested in))/(total number of all the anisotropic users that the user is interested in) - (weight of negative feedback × (sum of feature vectors of all the anisotropic users that the user is not interested in))/(total number of all the anisotropic users that the user is not interested in).
Let beta, gamma be weight values of positive and negative feedback, I γ ，I nγ A set of users that are indicative of the user's interest and disinterest. Then there is a change in the number of,
wherein, the first and the second end of the pipe are connected with each other,
a preference feature vector characterizing the user.
In one implementation, the selecting a user set with a feature set having the highest similarity to the preference feature set from the user database may include: and respectively carrying out Euclidean similarity calculation on the feature sets of all users in the user database and the preference feature set to obtain a user set with the highest similarity.
Euclidean similarity, is often used to calculate the absolute distance between two objects in either the low or high dimension. The smaller the absolute distance between two objects is, the higher the similarity between the two objects is, and the higher the association probability is.
Logic expression:
euclidean similarity = (sum of squares of differences between dimensional coordinates of user preference feature vectors).
And setting p and q as a user preference feature set and a feature set of a certain user respectively, and setting the maximum dimension of p and q as n.
Then there is a list of the number of,
according to the scheme, the user recommendation process comprises three parts, namely feature extraction, feature learning and recommendation generation. In the feature extraction process, key information is extracted from unstructured data and structured labels such as user signatures and the puppet standard by using a TF-IDF algorithm to calculate a feature set of each user; the feature learning process uses a Rocchio algorithm to generate a preference feature set of a certain user according to the operation feedback of the user; in the recommendation generation process, the Euclidean similarity calculation is carried out on the candidate different-sex user characteristic data and the preference characteristic set of the user generated in the characteristic learning process, and n different-sex users most relevant to the preference characteristics of the user are returned to the user.
In other implementations, after the selecting, by the user recommendation method, the user set with the feature set having the highest similarity to the preference feature set from the user database, the method may further include: and displaying the user information of each user in the user set based on the high and low output of the similarity between the feature set and the preference feature set.
The implementation is that the information of n users returned by the recommendation process is displayed on a user page and an administrator page. The user page can acquire n different users with the highest matching degree with the preference of the user page, and continuously perform interesting and uninteresting operations on each user so as to gradually and accurately improve the recommendation accuracy; the administrator page may obtain the n heterogeneous users with the highest matching degree with the selected role preference and recommend the n heterogeneous users to the target user.
Fig. 3 is a schematic diagram of a recommendation process logic of the user recommendation method disclosed in the embodiment of the present invention. With reference to fig. 3, the implementation of the scheme of the application includes four parts of single user feature extraction, user preference feature learning, recommendation generation, and page display: the single user feature extraction is mainly used for extracting key features of all users in a single database. Acquiring key information such as height, age, address and friend-making requirements of a user; user preference feature learning, namely constructing an exclusive user picture according to the preference of a certain user to obtain a preference feature set of the user picture; recommendation generation, which is to analyze similarity mainly according to a preference data set of a certain user and other heterosexual feature data in a single database to obtain a group of heterosexual users with higher similarity; and page display, which is mainly to present the opposite sex data in the recommendation generation step to the corresponding user.
While, for purposes of simplicity of explanation, the foregoing method embodiments have been described as a series of acts or combination of acts, it will be appreciated by those skilled in the art that the present invention is not limited by the illustrated ordering of acts, as some steps may occur in other orders or concurrently with other steps in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required by the invention.
The method is described in detail in the embodiments disclosed above, and the method of the present invention can be implemented by various types of apparatuses, so that the present invention also discloses an apparatus, and the following detailed description will be given of specific embodiments.
Fig. 4 is a schematic structural diagram of a user recommendation device according to an embodiment of the present invention. Referring to fig. 4, the user recommendation device 40 may include:
a historical data obtaining module 401, configured to obtain historical operation data of the first user.
A preference feature determination module 402 for determining a set of preference features for the first user based on the historical operational data.
A recommended user determining module 403, configured to select, from a user database, a user set with a feature set having the highest similarity with the preference feature set, where the user set includes user information of at least one user.
The user recommendation device in the embodiment is based on the user database, realizes data processing based on text content, and realizes customized friend-making recommendation schemes for different users through feature extraction, analysis and comparison of individual users. The accuracy of the user recommendation function is greatly improved, the satisfaction of the user on the platform is improved, and the user friend making target is realized more quickly.
In one implementation, the user recommendation apparatus may further include: and the user characteristic extraction module is used for extracting the characteristics of the user information of all the users in the user database in advance to obtain a characteristic set of each user.
In one implementation, the user feature extraction module is specifically configured to: extracting key information in the user information of each user in the user database based on a TF-IDF algorithm; and processing based on the key information to obtain a feature set of each user.
In one implementation, the preference feature determination module is specifically configured to: and constructing a user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the user portrait.
In one implementation, the recommended user determination module is specifically configured to: and respectively carrying out Euclidean similarity calculation on the feature sets of all users in the user database and the preference feature set to obtain a user set with the highest similarity.
In one implementation, the user recommendation apparatus may further include: and the recommendation output module is used for outputting and displaying the user information of each user in the user set based on the similarity between the feature set and the preference feature set.
The user recommendation device in any of the above embodiments includes a processor and a memory, and the historical data obtaining module, the preference feature determining module, the recommended user determining module, and the user feature extracting module in the above embodiments: the recommendation output module and the like are stored in the memory as program modules, and the processor executes the program modules stored in the memory to realize corresponding functions.
The processor comprises a kernel, and the kernel calls the corresponding program module from the memory. The kernel can be provided with one or more, and the processing of the return visit data is realized by adjusting the kernel parameters.
The memory may include volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), including at least one memory chip.
In an exemplary embodiment, a computer-readable storage medium, which can be directly loaded into an internal memory of a computer and contains a software code, is provided, and the computer program can be loaded into the computer and executed to implement the steps of any of the above-mentioned embodiments of the user recommendation method.
In an exemplary embodiment, a computer program product, which can be directly loaded into an internal memory of a computer and contains a software code, is provided, and can implement the steps of any of the above-described user recommendation methods when the computer program is loaded into the computer and executed.
Further, the embodiment of the invention provides the electronic equipment. Fig. 5 is a schematic structural diagram of an electronic device disclosed in an embodiment of the present application. Referring to fig. 5, the electronic device 50 includes at least one processor 501, and at least one memory 502, a bus 503 connected to the processor; the processor and the memory complete mutual communication through a bus; the processor is used for calling the program instructions in the memory to execute the user recommendation method.
In the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
It is further noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A user recommendation method, comprising:
obtaining historical operation data of a first user;
determining a set of preference features for the first user based on the historical operational data;
and selecting a user set with the highest similarity between the feature set and the preference feature set from a user database, wherein the user set comprises user information of at least one user.
2. The user recommendation method of claim 1, further comprising:
and carrying out feature extraction on the user information of all users in the user database in advance to obtain a feature set of each user.
3. The user recommendation method according to claim 2, wherein the pre-extracting the features of the user information of all users in the user database to obtain the feature set of each user comprises:
extracting key information in the user information of each user in the user database based on a TF-IDF algorithm;
and processing based on the key information to obtain a feature set of each user.
4. The user recommendation method of claim 1, wherein said determining a set of preference features for the first user based on the historical operational data comprises:
and constructing the user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the first user portrait.
5. The user recommendation method according to claim 1, wherein the selecting a user set with a feature set having a highest similarity with the preference feature set from a user database comprises:
and respectively carrying out Euclidean similarity calculation on the feature sets of all users in the user database and the preference feature set to obtain a user set with the highest similarity.
6. The user recommendation method according to claim 1, further comprising, after selecting the user set with the feature set having the highest similarity with the preference feature set from the user database:
and displaying the user information of each user in the user set based on the high and low output of the similarity between the feature set and the preference feature set.
7. A user recommendation device, comprising:
the historical data acquisition module is used for acquiring historical operation data of a first user;
a preference feature determination module for determining a set of preference features for the first user based on the historical operational data;
and the recommended user determining module is used for selecting a user set with the feature set having the highest similarity with the preference feature set from a user database, wherein the user set comprises user information of at least one user.
8. The user recommendation device of claim 7, further comprising:
and the user characteristic extraction module is used for extracting the characteristics of the user information of all the users in the user database in advance to obtain a characteristic set of each user.
9. The user recommendation device of claim 8, wherein the user feature extraction module is specifically configured to: extracting key information in the user information of each user in the user database based on a TF-IDF algorithm; and processing based on the key information to obtain a feature set of each user.
10. The user recommendation device of claim 7, wherein the preference feature determining module is specifically configured to: and constructing a user portrait of the first user based on the historical operation data through a Rocchio algorithm to obtain a preference characteristic set of the user portrait.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211263075.XA CN115659013A (en) | 2022-10-14 | 2022-10-14 | User recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211263075.XA CN115659013A (en) | 2022-10-14 | 2022-10-14 | User recommendation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115659013A true CN115659013A (en) | 2023-01-31 |
Family
ID=84987942
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211263075.XA Pending CN115659013A (en) | 2022-10-14 | 2022-10-14 | User recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115659013A (en) |
-
2022
- 2022-10-14 CN CN202211263075.XA patent/CN115659013A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9846744B2 (en) | Media discovery and playlist generation | |
| CN109885773B (en) | Personalized article recommendation method, system, medium and equipment | |
| US9489401B1 (en) | Methods and systems for object recognition | |
| US9519685B1 (en) | Tag selection, clustering, and recommendation for content hosting services | |
| Figueiredo et al. | Assessing the quality of textual features in social media | |
| US20100223258A1 (en) | Information retrieval system and method using a bayesian algorithm based on probabilistic similarity scores | |
| EP1995669A1 (en) | Ontology-content-based filtering method for personalized newspapers | |
| CN107066589B (en) | Entity semantics and word frequency ordering method and device based on comprehensive knowledge | |
| CN105426528A (en) | Retrieving and ordering method and system for commodity data | |
| CN111259173B (en) | Search information recommendation method and device | |
| CN107193883B (en) | Data processing method and system | |
| CN112084405A (en) | Searching method, searching device and computer storage medium | |
| KR101088710B1 (en) | Method and Apparatus for Online Community Post Searching Based on Interactions between Online Community User and Computer Readable Recording Medium Storing Program thereof | |
| Donaldson | A hybrid social-acoustic recommendation system for popular music | |
| Zaharieva et al. | Retrieving Diverse Social Images at MediaEval 2017: Challenges, Dataset and Evaluation. | |
| KR20110043369A (en) | Association analysis method for music recommendation | |
| Tomeo et al. | Addressing the cold start with positive-only feedback through semantic-based recommendations | |
| CN115659013A (en) | User recommendation method and device | |
| CN114282119A (en) | Scientific and technological information resource retrieval method and system based on heterogeneous information network | |
| CN112307312A (en) | Article recommendation method and device | |
| CN111046221B (en) | Song recommendation method, device, terminal equipment and storage medium | |
| KR101137491B1 (en) | System and Method for Utilizing Personalized Tag Recommendation Model in Web Page Search | |
| CN115017293A (en) | Document recommendation method based on LDA topic model | |
| CN116881549A (en) | Information providing method and device, electronic equipment and storage medium | |
| CN116977024A (en) | Receipt address-based bill recognition method, device and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |