CN113094584A - Method and device for determining recommended learning resources - Google Patents
Method and device for determining recommended learning resources Download PDFInfo
- Publication number
- CN113094584A CN113094584A CN202110405603.XA CN202110405603A CN113094584A CN 113094584 A CN113094584 A CN 113094584A CN 202110405603 A CN202110405603 A CN 202110405603A CN 113094584 A CN113094584 A CN 113094584A
- Authority
- CN
- China
- Prior art keywords
- determining
- user
- target
- target user
- algorithm
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
Abstract
The invention discloses a determination method and a determination device for recommending learning resources, and relates to the technical field of artificial intelligence. One embodiment of the method comprises: determining characteristic data of a target user; judging whether the characteristic data meet a first preset condition or not; determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition; and determining the recommended learning resources of the target user by using the target recommendation algorithm and the characteristic data of the target user. The implementation method can recommend the relevant learning resources to the user more accurately.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a method and a device for determining recommended learning resources.
Background
Internet education is a new education form combining internet science and technology with the education field. Users often cannot accurately locate the resources needed by themselves when confronted with massive learning resource information. The recommendation system is a personalized information recommendation system which recommends information, products and the like which are interested by a user to the user according to the information requirements, interests and the like of the user. However, the learning resource recommendation method in the existing recommendation system is often not accurate enough, and cannot meet the needs of users.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and an apparatus for determining recommended learning resources, which can recommend relevant learning resources to a user more accurately.
In a first aspect, an embodiment of the present invention provides a method for determining recommended learning resources, including:
determining characteristic data of a target user;
judging whether the characteristic data meet a first preset condition or not;
determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition;
and determining the recommended learning resources of the target user by using the target recommendation algorithm and the characteristic data of the target user.
Optionally, the feature data comprises: a user interest tag;
the judging whether the feature data meet a first preset condition includes:
and judging whether the number of the user interest tags is smaller than a first number threshold value.
Optionally, the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is smaller than a first number threshold, determining a clustering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining a user group of the target user by utilizing a clustering algorithm and the characteristic data of the target user;
and determining recommended learning resources of the target user according to the user group.
Optionally, the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining an association rule algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining the recommended learning resources of the target user according to the similar users.
Optionally, the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining a collaborative filtering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
Optionally, the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, judging whether the number of the users in the system meets a second preset condition, wherein the second preset condition comprises that: the number of users in the system is less than a second number threshold;
and determining a target recommendation algorithm according to a second judgment result corresponding to the second preset condition.
Optionally, the determining a target recommendation algorithm according to a second judgment result corresponding to the second preset condition includes:
if the second judgment result represents that the number of the users in the system is greater than or equal to the second number threshold, determining an association rule algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining the recommended learning resources of the target user according to the similar users.
Optionally, the determining a target recommendation algorithm according to a second judgment result corresponding to the second preset condition includes:
if the second judgment result represents that the number of the users in the system is smaller than the second number threshold, determining the collaborative filtering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
Optionally, the method further comprises:
acquiring a first training set, wherein the first training set comprises a plurality of learning resources;
acquiring first behavior data of a test user on the training set;
determining a first recommendation list of the test user according to the first behavior data;
determining a first behavior list of the test user by using the target recommendation algorithm, the feature data of the test user and the first training set;
and determining the accuracy of the determination method of the recommended learning resources according to the first recommendation list and the first behavior list.
Optionally, the method further comprises:
acquiring a second training set, wherein the second training set comprises a plurality of learning resources;
acquiring second behavior data of the test user on the training set;
determining a second recommendation list of the test user according to the second behavior data;
determining a second behavior list of the test user by using the target recommendation algorithm, the feature data of the test user and the second training set;
and determining the recall rate of the determination method of the recommended learning resources according to the second recommendation list and the second behavior list.
Optionally, the determining the feature data of the target user includes:
acquiring a log file of the target user;
pre-processing the log file, the pre-processing comprising one of: word segmentation processing, feature word recognition and invalid character filtering;
and performing statistical analysis on the processed log file to determine the characteristic data of the target user.
Optionally, the feature data comprises: explicit data and implicit data;
the explicit data includes at least one of: resource comment, resource approval, resource forwarding, resource downloading, user grading feedback, resource downloading, question making record, course resource searching, course interaction times, interaction time of each time and system online time;
the stealth data includes at least one of: browsing history, reading duration, viewing records and searching logs.
In a second aspect, an embodiment of the present invention provides an apparatus for determining a recommended learning resource, including:
the data determination module is used for determining the characteristic data of the target user;
the condition judgment module is used for judging whether the characteristic data meet a first preset condition or not;
the algorithm determining module is used for determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition;
and the resource determining module is used for determining the recommended learning resources of the target user by utilizing the target recommendation algorithm and the characteristic data of the target user.
In a third aspect, an embodiment of the present invention provides an electronic device, including:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any of the embodiments described above.
In a fourth aspect, an embodiment of the present invention provides a computer-readable medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method of any one of the above embodiments.
One embodiment of the above invention has the following advantages or benefits: and determining a target recommendation algorithm according to whether the characteristic data of the target user meets a first preset condition, and determining recommended learning resources of the target user by using the target recommendation algorithm and the characteristic data of the target user. According to the embodiment of the invention, a fixed recommendation algorithm is not adopted, but a target recommendation algorithm is determined according to the characteristic data of the user, so that the obtained recommended learning resources are more accurate, and the requirements of the user can be better met.
Further effects of the above-mentioned non-conventional alternatives will be described below in connection with the embodiments.
Drawings
The drawings are included to provide a better understanding of the invention and are not to be construed as unduly limiting the invention. Wherein:
FIG. 1 is an exemplary system architecture diagram in which embodiments of the present invention may be employed;
fig. 2 is a schematic diagram illustrating a flow of a determination method for recommending learning resources according to an embodiment of the present invention;
fig. 3 is a schematic diagram illustrating a flow of another method for determining a recommended learning resource according to an embodiment of the present invention;
fig. 4 is a schematic diagram illustrating a flow of a determination method for recommending learning resources according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a determination apparatus for recommending learning resources according to an embodiment of the present invention;
fig. 6 is a schematic block diagram of a computer system suitable for use in implementing a terminal device or server of an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention are described below with reference to the accompanying drawings, in which various details of embodiments of the invention are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
Fig. 1 shows an exemplary system architecture 100 to which a determination method of a recommended learning resource or a determination apparatus of a recommended learning resource of an embodiment of the present invention can be applied.
As shown in fig. 1, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have installed thereon various communication client applications, such as shopping-like applications, web browser applications, search-like applications, instant messaging tools, mailbox clients, social platform software, office applications, etc. (by way of example only).
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 105 may be a server that provides various services, such as a back-office management server (for example only) that provides support for online educational websites browsed by users using the terminal devices 101, 102, 103. The server 105 may determine a target recommendation algorithm according to whether the feature data of the target user meets a first preset condition; and determining recommended learning resources of the target user by using a target recommendation algorithm.
It should be noted that the method for determining recommended learning resources provided by the embodiment of the present invention is generally executed by the server 105, and accordingly, the apparatus for determining recommended learning resources is generally disposed in the server 105.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
Fig. 2 is a schematic diagram of a flow of a determination method for recommending learning resources according to an embodiment of the present invention, as shown in fig. 2, the method includes:
step 201: characteristic data of the target user is determined.
Relevant feature data of the target user is collected to generate a user representation of the recommended task, the feature data including user attributes, user behavior data, or resources accessed by the user. Through establishing the user portrait, the recommendation system recommends learning resources for the user.
The characteristic data may be either explicit or implicit data, depending on the different types of input. For example, the most direct explicit feedback may be that the user directly enters content of interest. Implicit feedback may be an indirect inference of user preferences through observation of user behavior, or hybrid feedback may be obtained through a combination of explicit and implicit feedback.
Explicit data refers to data actively input by a user, such as comments, praise, forwarding, downloading and the like on content, and implicit data refers to browsing history, reading duration, viewing records, search logs and the like of the user. The back office creates a data set for each user using/accessing the site.
Specifically, the explicit data may include: resource comment, resource approval, resource forwarding, resource downloading, user grading feedback, resource downloading, question making record, course resource searching, course interaction times, interaction time of each time, system online time and the like. The stealth data may include: browsing history, reading duration, viewing records, search logs, etc.
Using the e-learning platform as an example, a user representation is a collection of personal information associated with a particular user. The information includes cognitive skills, intelligence level, learning style, hobbies and interactive behaviors of the user and the like. User profiles are typically used for information retrieval when building user models. In other words, the user representation roughly reflects the user model. The success of a recommendation system depends largely on its ability to characterize the user's interests. An accurate user model is essential to obtain an accurate recommendation result.
Because each user has a different preference for the product, the feature data set collected for each user is also quite different. The collected user characteristic data is more and more along with the time, and the recommended result is more and more accurate through a series of data analysis.
The user characteristic data can be represented by a data matrix, and the data matrix is subjected to standardization processing by using translation standard deviation transformation and maximum and minimum value transformation. The numerical value problem can be effectively avoided by the standardization processing of the data matrix, and the final data classification result is greatly influenced by different differences of orders of magnitude possibly caused by overlarge data; meanwhile, the influence of the characteristic data of the user on the K-means clustering and association rule analysis can be balanced through standardization processing, and better classification and recommendation effects are achieved.
Step 202: and judging whether the characteristic data meets a first preset condition or not.
The first preset condition can be set according to specific requirements and practical application environments. For example, the first preset condition may be: whether the number of user interest tags is less than a first number threshold, whether the number of users is less than a second number threshold, whether the number of learning resources is less than a third number threshold, and so on.
Step 203: and determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition.
The target recommendation algorithm can be set according to specific requirements and actual application environments. For example, the target recommendation algorithm may be a content-based recommendation algorithm, a collaborative filtering recommendation algorithm, a knowledge-based recommendation algorithm, and the like.
Step 204: and determining recommended learning resources of the target user by using a target recommendation algorithm and the characteristic data of the target user.
A learning resource is a resource used to assist a user in learning. The learning resources may include: video, audio, courseware, electronic books, question banks, analog test paper, and the like.
In the embodiment of the invention, a target recommendation algorithm is determined according to whether the characteristic data of the target user meets a first preset condition; and determining recommended learning resources of the target user by using a target recommendation algorithm and the characteristic data of the target user. According to the embodiment of the invention, a fixed recommendation algorithm is not adopted, but a target recommendation algorithm is determined according to the characteristic data of the user, so that the obtained recommended learning resources are more accurate. Therefore, the problem that accurate learning resources cannot be recommended to the user can be solved.
After the resource is recommended, the result can be measured and calculated through Precision (Precision) and Recall (Recall) indexes, and the effective evaluation on the recommendation effect can be realized.
In one embodiment of the invention, the method further comprises: acquiring a first training set, wherein the first training set comprises a plurality of learning resources; acquiring first behavior data of a test user on a first training set; determining a first recommendation list of the test user according to the first behavior data; determining a first behavior list of the test user by using a target recommendation algorithm, the feature data of the test user and a first training set; and determining the accuracy of the determination method for recommending the learning resources according to the first recommendation list and the first behavior list. The calculation of the accuracy is shown by the following formula:
where Precision represents accuracy, r (u) is a first list of behaviors of the test user on the first test set, which is predicted by the method of the embodiment of the present invention, and t (u) is a first list of recommendations made based on behaviors of the test user on the first training set.
In one embodiment of the invention, the method further comprises: acquiring a second training set, wherein the second training set comprises a plurality of learning resources; acquiring second behavior data of the test user on a second training set; determining a second recommendation list of the test user according to the second behavior data; determining a second behavior list of the test user by using a target recommendation algorithm, the feature data of the test user and a second training set; and determining the recall rate of the determination method of the recommended learning resources according to the second recommendation list and the second behavior list. The recall ratio is calculated as follows:
where Recall represents Recall, R (u) is a second list of behaviors of the test user on a second test set, which are predicted using the method of an embodiment of the present invention, and T (u) is a second list of recommendations based on behaviors of the test user on a second training set.
In one embodiment of the present invention, determining the characteristic data of the target user comprises: acquiring a log file of a target user; pre-processing the log file, the pre-processing comprising one of: word segmentation processing, feature word recognition and invalid character filtering; and performing statistical analysis on the processed log file to determine the characteristic data of the target user. By the method, the characteristic data of the target user can be accurately and completely extracted.
Aiming at the problem of information overload in an online platform of internet education, the embodiment of the invention clusters platform users by adopting a K-means clustering algorithm, thereby solving the problem of cold start in a recommendation system; through the course resource learning records of the users, a collaborative filtering algorithm is adopted, Apriori algorithm in the association rule mining technology is utilized to find users with similar interests, and the personalized resource recommendation list of the target user is generated according to the course resources which are interested by other users, so that dependence on user explicit data is avoided, and the implicit characteristic data of the users can be effectively analyzed. The method provided by the embodiment of the invention can provide personalized course learning resource recommendation for the user and meet the personalized requirements of the user. Fig. 3 is a schematic diagram of a flow of a determination method for recommending learning resources according to an embodiment of the present invention, as shown in fig. 3, the method includes:
step 301: determining feature data of the target user, the feature data comprising: a user interest tag.
Step 302: it is determined whether the number of user interest tags is less than a first number threshold.
If the first judgment result represents that the number of the user interest tags is smaller than the first number threshold, executing step 303; if the first determination result indicates that the number of the user interest tags is greater than or equal to the first number threshold, step 305 or step 307 is executed.
Step 303: and determining the clustering algorithm as a target recommendation algorithm, and determining a user group of the target user by using the clustering algorithm and the characteristic data of the target user.
The recommendation system has a cold start problem, and for mean clustering of new users or users whose interest tags do not meet the number requirement, clustering processing needs to be performed on information of existing users because the interest tags of the new users cannot be obtained. And clustering the personal information data of the users of the platform by adopting a mean clustering algorithm, wherein the users are divided into different clusters, and each cluster represents a user group with similar interest. The core idea of the K-means clustering (K-means clustering) algorithm is that an initial clustering center is selected at random, then the Euclidean distance from each sample point to the initial clustering center is calculated, and the Euclidean distances are distributed to the class represented by the clustering center with the maximum similarity according to the closest criterion; and finally, calculating the mean value of all sample points of each cluster, and updating the clustering center until the target criterion function is converged.
Step 304: and determining recommended learning resources of the target user according to the user group.
The recommended learning resources of the user can be determined by the following method.
And in the first mode, extracting a plurality of interest tags in front of a user group where the target user is located, and endowing the tags to the target user in the class cluster. And completing course resource recommendation of the user using the learning platform in the earlier stage through the extracted interest labels.
And secondly, determining that the users of the user group where the target user is located pay attention to learning or browse more learning resources, and recommending the learning resources to the users as recommended learning resources.
Step 305: and determining the association rule algorithm as a target recommendation algorithm, and determining similar users of the target user by using the association rule algorithm and the characteristic data of the target user.
Step 306: and determining recommended learning resources of the target user according to the similar users.
In order to complete the recommendation of the relevant resources of the users, an Apriori algorithm in an association rule technology is adopted to mine association rules among the users. The target of algorithmic analysis includes two terms: frequent sets of items and association rules are discovered. First a frequent set of items needs to be found before association rules can be obtained. The two input parameters of the Apriori algorithm are the minimum support (min _ support) and the data set, respectively. The algorithm will first generate a list of all the item sets for a single item and then scan the data records to see which item sets meet the minimum support requirement and the sets that do not meet the minimum support are removed. The remaining sets are then combined to generate a set of items containing two elements. Next, the transaction record is rescanned to remove the set of items that do not meet the minimum support. This process is repeated until all sets of items have been removed.
In order to complete the subsequent collaborative filtering analysis process based on users, an Apriori algorithm commonly used in association rule mining technology is adopted to mine association rules among users. The support degree of the association rule is defined as the following formula:
support(A→B)＝P(A∪B)
wherein, P (A U B represents the probability that the transaction contains the union of the set A and the set B. the confidence of the association rule is defined as the following formula:
in order to reduce the computational complexity of generating a frequent item set, the candidate item set needs to be pruned with support, and the process needs to follow two laws:
if a set is a frequent item set, then all of its subsets are frequent item sets. For example: assuming that a set A, B is a frequent item set, i.e., A, B appears on a record at the same time with a frequency greater than or equal to the minimum support (min _ support), its subsets A and B must appear with a frequency greater than or equal to the minimum support, i.e., its subsets are frequent item sets.
If a collection is not a frequent item set, then all of its supersets are not frequent item sets. For example: assuming that the set { A } is not a frequent item set, i.e., A occurs less frequently than the minimum support, then any superset thereof, e.g., { A, B } must occur less frequently than the minimum support, and thus its superset must also not be a frequent item set.
After the association rule is utilized to excavate the association user with the strong association rule with the target user, the score of the target user on the target resource can be predicted directly according to the score of the association user on the target resource, the target user does not generate target behaviors on the target resource, and the target behaviors can include browsing, learning, collecting, concerning, evaluating, downloading and the like. And then selecting a plurality of learning resources with the highest prediction scores to the target user. Specifically, the score of the associated user on the target resource is determined, the value of the similarity between the target user and the associated user is used as a weight, the weighted sum of the scores of the associated users on the target resource is calculated, and the weighted sum is used as the predicted score of the target user on the target resource.
In one embodiment of the invention, the method for determining the recommended learning resources of the target user comprises the following steps:
s01: and establishing a user scoring matrix, wherein the user scoring matrix comprises the scores of the users for the learning resources. The input data of the collaborative filtering algorithm is usually expressed as a user-rating matrix, the rating value can be the browsing times, the learning times, the collection times and the like of the user, and the display rating can also be adopted, such as the direct rating or the learning times of the user on the resource course as the rating.
S02: at least one nearest neighbor of the target user is determined. The method mainly finishes the search of the nearest neighbors of the target user, and calculates the nearest neighbor set which is most similar to the target user by calculating the similarity between the target user and other users. The process is completed in two steps: firstly, the similarity between users is obtained by adopting a cosine value similarity calculation method, and secondly, the nearest neighbor is selected according to the following method. The following eligible users may be set as "nearest neighbors," including: selecting users with similarity greater than a set threshold; selecting a plurality of previous users with the maximum similarity; select a number of users with a similarity greater than a predetermined threshold, and so on.
S03: the step 305 is executed, and a correlation rule algorithm is adopted to generate a similar user result set of the target user. And mining users with strong association rules with the target user by adopting an association rule algorithm, and adding the users into the neighbor set obtained in the S02 to generate a similar user result set of the target user.
S04: and generating a related recommended resource result, wherein the calculation method comprises the following steps: and taking the similarity value between the target user and the neighbor user in the similar user result set generated in the step S03 as a weight, and then determining the prediction score of the target user on the target resource according to the score of the neighbor user on the target resource. For example, the difference between the resource's score of the neighboring user and all the scores of the neighboring users is weighted and averaged. And carrying out weighted average on the difference value between the score of the target resource by the neighbor user and the score average value (or median) of the neighbor user, or calculating the average value of the weighted sum of the scores of the target resource by all the associated users, and taking the average value of the weighted sum as the predicted score of the target user on the target resource, and the like. The method is used for predicting the scores of the target users on the resources which do not generate the evaluation, and then selecting a plurality of items with the highest predicted scores to recommend to the target users.
Step 307: and determining the collaborative filtering algorithm as a target recommendation algorithm, and determining the recommended learning resources of the target user by utilizing the collaborative filtering algorithm and the characteristic data of the target user.
And in the first mode, the recommended learning resources of the target user are determined according to the similar user result set of the target user. The basic idea of this approach is to search the nearest neighbors of the target user based on the similarity between the user and the curriculum learning resources, and then generate recommendations to the target user based on the scores of the nearest neighbors. The method mainly comprises the following steps:
s11: and establishing a user scoring matrix, wherein the user scoring matrix comprises the scores of the users for the learning resources. The input data of the collaborative filtering algorithm is usually expressed as a user-rating matrix, the rating value can be the browsing times, the learning times, the collection times and the like of the user, and the display rating can also be adopted, such as the direct rating or the learning times of the user on the resource course as the rating.
S12: at least one nearest neighbor of the target user is determined. The method mainly finishes the search of the nearest neighbors of the target user, and calculates the nearest neighbor set which is most similar to the target user by calculating the similarity between the target user and other users. The process is completed in two steps: firstly, the similarity between users is obtained by adopting a cosine value similarity calculation method, and secondly, the nearest neighbor is selected according to the following method. The following eligible users may be set as "nearest neighbors," including: selecting users with similarity greater than a set threshold; selecting a plurality of previous users with the maximum similarity; select a number of users with a similarity greater than a predetermined threshold, and so on.
S13: and generating a similar user result set of the target user by adopting an association rule algorithm. If Apriori algorithm is adopted to dig out users with strong association rules with the target user, the users are added into the neighbor set obtained in the step S02 to generate a similar user result set of the target user.
S14: and generating a related recommended resource result, wherein the calculation method comprises the following steps: and taking the similarity value between the target user and the neighbor user in the similar user result set generated in the step S03 as a weight, and then determining the prediction score of the target user on the target resource according to the score of the neighbor user on the target resource. For example, the difference between the resource's score of the neighboring user and all the scores of the neighboring users is weighted and averaged. And carrying out weighted average on the difference value between the score of the target resource by the neighbor user and the average value (or the median) of all the scores of the neighbor user, or calculating the average value of the weighted sum of the scores of the target resource by each associated user, and taking the average value of the weighted sum as the predicted score of the target user on the target resource, and the like. The method is used for predicting the scores of the target users on the resources which are not evaluated, and then selecting a plurality of learning resources with the highest predicted scores to recommend the learning resources to the target users.
S15: and adopting a collaborative filtering algorithm, such as an ALS algorithm, to iteratively calculate the personalized resource recommendation result of the user. The specific process is as follows: (1) decomposing the sparse scoring matrix into a product of a user characteristic vector matrix and a resource and article characteristic vector matrix; (2) alternately using a least square method to gradually calculate the feature vector of the user-resource object, so that the mean square error is minimum; (3) and predicting the grade of a certain user on a certain resource item through the user-resource item characteristics to generate the personalized recommended learning resource.
In one embodiment of the present invention, a final set of recommended learning resources for the target user is generated based on the sets of recommended learning resources generated in S14 and S15. Specifically, a union or an intersection of the two learning resource sets may be extracted, or a plurality of learning resources may be extracted from the two learning resource sets respectively, and the extracted learning resources are used as a final recommended learning resource set of the target user. The embodiment of the present invention does not limit how to generate the final recommended learning resource set of the target user according to the recommended learning resource sets generated in S14 and S15.
And secondly, determining recommended learning resources of the target user by adopting an ALS algorithm as a target recommendation algorithm. Specifically, a user-resource-scoring data matrix is constructed, the preference degree of each user to resources is predicted through matrix decomposition, and recommendation of personalized course resources is completed.
The ALS algorithm belongs to User-Item collaborative filtering, namely a hybrid collaborative filtering algorithm. It considers both user and article aspects. The relationship between the user and the resource object can be abstracted as the following triples:<User,Item,Rating>. Wherein User represents User, Item represents resource Item, and Rating is UserAnd scoring the resource item to represent the user's preference for the resource item. Assuming that m users and n resource items are included in the system, a scoring matrix R is definedm×nWherein the element ruiIndicating the rating of the ith user for the ith resource item. Since it is also not possible for one user to score all resource items, the R matrix is a sparse matrix.
For the data characteristics, it can be assumed that there are several associated dimensions (age, region, position, etc.) between the platform user and the resource item, and the following formula can be obtained by mapping the scoring matrix R to these feature dimensions:
where k represents the dimension of the implicit factor, and generally takes a value of 20 to 200. X is a user characteristic vector matrix, and Y is a resource and article characteristic vector matrix. In order to make the low rank matrices X and Y as close to R as possible, the following squared error function formula needs to be minimized:
wherein r isuiRepresents the u user's score, x, for the i resource itemuIs the feature vector of the u-th user, yiIs the feature vector of the ith resource item. The specific process is as follows: decomposing the sparse scoring matrix into a product of a user characteristic vector matrix X and a resource and article characteristic vector matrix Y; alternately using a least square method to gradually calculate the feature vector of the user-resource object, so that the mean square error is minimum; and predicting the grade of a certain user on a certain resource item through the user-resource item characteristics to generate the personalized recommended resource.
The method of the embodiment of the invention can obtain the following beneficial effects: the method adopts a clustering algorithm to realize the generation of the recommended resources of the new user, and solves the cold start problem of the recommendation system; strong association rules among resource articles are mined by adopting an association rule analysis technology, and relevant resource recommendation results of users are generated; the calculation complexity of the collaborative filtering recommendation algorithm is reduced by adopting the ALS algorithm in the collaborative filtering, the implicit feedback characteristic of the user is utilized, the data sparsity existing in the traditional recommendation technology is avoided, and the accuracy and the real-time performance of the resource recommendation result are improved.
Fig. 4 is a schematic diagram of a flow of a determination method for recommending learning resources according to an embodiment of the present invention, as shown in fig. 4, the method includes:
step 401: determining feature data of the target user, the feature data comprising: a user interest tag.
Step 402: it is determined whether the number of user interest tags is less than a first number threshold.
If the first judgment result represents that the number of the user interest tags is smaller than the first number threshold, executing step 403; if the first determination result indicates that the number of the user interest tags is greater than or equal to the first number threshold, step 404 is executed.
Step 403: determining a clustering algorithm as a target recommendation algorithm; and determining the recommended learning resources of the target user by using a clustering algorithm and the characteristic data of the target user.
Step 404: it is determined whether the number of users in the system is less than a second number threshold.
If the second determination result indicates that the number of users in the system is greater than or equal to the second number threshold, execute step 405; if the second determination result indicates that the number of users in the system is less than the second number threshold, step 407 is executed.
Step 405: and determining the association rule algorithm as a target recommendation algorithm, and determining similar users of the target user by using the association rule algorithm and the characteristic data of the target user.
Step 406: and determining recommended learning resources of the target user according to the similar users.
Step 407: and determining the collaborative filtering algorithm as a target recommendation algorithm, and determining the recommended learning resources of the target user by utilizing the collaborative filtering algorithm and the characteristic data of the target user.
In the embodiment of the invention, the target recommendation algorithm is determined according to the number of the interest tags of the users and the number of the users in the system respectively. And under the condition that the number of users in the system is large, determining the associated user of the target user through the association rule, wherein the attribute difference between the determined associated user and the target user is small. Therefore, the corresponding recommended learning resources of the target user can be predicted more accurately through the association rule algorithm.
Under the condition that the number of users in the system is small, if the association rule determines the associated user with the target user, the attribute difference between the determined associated user and the target user is large, and the corresponding recommended learning resource of the target user cannot be predicted accurately through the associated user. Therefore, under the condition that the number of users in the system is small, a collaborative filtering algorithm is adopted to predict the corresponding recommended learning resources of the target user more accurately.
Fig. 5 is a schematic structural diagram of a determination apparatus for recommending learning resources according to an embodiment of the present invention, and as shown in fig. 5, the method includes:
a data determining module 501, configured to determine feature data of a target user;
a condition determining module 502, configured to determine whether the feature data meets a first preset condition;
the algorithm determining module 503 is configured to determine a target recommendation algorithm according to a first determination result corresponding to a first preset condition;
the resource determining module 504 is configured to determine recommended learning resources of the target user by using a target recommendation algorithm and feature data of the target user.
Optionally, the characteristic data comprises: a user interest tag;
the condition determining module 502 is specifically configured to:
it is determined whether the number of user interest tags is less than a first number threshold.
Optionally, the algorithm determining module 503 is specifically configured to:
if the first judgment result represents that the number of the user interest tags is smaller than a first number threshold, determining a clustering algorithm as a target recommendation algorithm;
the resource determining module 504 is specifically configured to:
determining a user group of the target user by using a clustering algorithm and the characteristic data of the target user;
and determining recommended learning resources of the target user according to the user group.
Optionally, the condition determining module 502 is specifically configured to:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining an association rule algorithm as a target recommendation algorithm;
the resource determining module 504 is specifically configured to:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining recommended learning resources of the target user according to the similar users.
Optionally, the algorithm determining module 503 is specifically configured to:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining the collaborative filtering algorithm as a target recommendation algorithm;
the resource determining module 504 is specifically configured to:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
Optionally, the algorithm determining module 503 is specifically configured to:
if the first judgment result represents that the number of the user interest tags is larger than or equal to the first number threshold, judging whether the number of the users in the system meets a second preset condition, wherein the second preset condition comprises the following steps: the number of users in the system is less than a second number threshold;
and determining a target recommendation algorithm according to a second judgment result corresponding to a second preset condition.
Optionally, the algorithm determining module 503 is specifically configured to:
if the second judgment result represents that the number of the users in the system is larger than or equal to a second number threshold, determining the association rule algorithm as a target recommendation algorithm;
the resource determining module 504 is specifically configured to:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining recommended learning resources of the target user according to the similar users.
Optionally, the algorithm determining module 503 is specifically configured to:
if the second judgment result represents that the number of the users in the system is smaller than a second number threshold, determining the collaborative filtering algorithm as a target recommendation algorithm;
the resource determining module 504 is specifically configured to:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
Optionally, the method further comprises:
an accuracy determining module 505, configured to obtain a first training set, where the first training set includes a plurality of learning resources;
acquiring first behavior data of a test user on a training set;
determining a first recommendation list of the test user according to the first behavior data;
determining a first behavior list of the test user by using a target recommendation algorithm, the feature data of the test user and a first training set;
and determining the accuracy of the determination method for recommending the learning resources according to the first recommendation list and the first behavior list.
Optionally, the method further comprises:
a recall rate determining module 506, configured to obtain a second training set, where the second training set includes a plurality of learning resources;
acquiring second behavior data of the test user on the training set;
determining a second recommendation list of the test user according to the second behavior data;
determining a second behavior list of the test user by using a target recommendation algorithm, the feature data of the test user and a second training set;
and determining the recall rate of the determination method of the recommended learning resources according to the second recommendation list and the second behavior list.
Optionally, the data determining module 501 is specifically configured to:
acquiring a log file of a target user;
pre-processing the log file, the pre-processing comprising one of: word segmentation processing, feature word recognition and invalid character filtering;
and performing statistical analysis on the processed log file to determine the characteristic data of the target user.
Optionally, the characteristic data comprises: explicit data and implicit data;
the explicit data includes at least one of: resource comment, resource approval, resource forwarding, resource downloading, user grading feedback, resource downloading, question making record, course resource searching, course interaction times, interaction time of each time and system online time;
the stealth data includes at least one of: browsing history, reading duration, viewing records and searching logs.
An embodiment of the present invention provides an electronic device, including:
one or more processors;
a storage device for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method of any of the embodiments described above.
Referring now to FIG. 6, a block diagram of a computer system 600 suitable for use with a terminal device implementing an embodiment of the invention is shown. The terminal device shown in fig. 6 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 6, the computer system 600 includes a Central Processing Unit (CPU)601 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage section 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data necessary for the operation of the system 600 are also stored. The CPU 601, ROM 602, and RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.
The following components are connected to the I/O interface 605: an input portion 606 including a keyboard, a mouse, and the like; an output portion 607 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 608 including a hard disk and the like; and a communication section 609 including a network interface card such as a LAN card, a modem, or the like. The communication section 609 performs communication processing via a network such as the internet. The driver 610 is also connected to the I/O interface 605 as needed. A removable medium 611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 610 as necessary, so that a computer program read out therefrom is mounted in the storage section 608 as necessary.
In particular, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 609, and/or installed from the removable medium 611. The computer program performs the above-described functions defined in the system of the present invention when executed by the Central Processing Unit (CPU) 601.
It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules described in the embodiments of the present invention may be implemented by software or hardware. The described modules may also be provided in a processor, which may be described as: a processor comprises a data determining module, a condition judging module, an algorithm determining module and a resource determining module. Where the names of these modules do not in some cases constitute a limitation of the module itself, for example, the data determination module may also be described as a "module that determines the characteristic data of the target user".
As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to comprise:
determining characteristic data of a target user;
judging whether the characteristic data meet a first preset condition or not;
determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition;
and determining the recommended learning resources of the target user by using the target recommendation algorithm and the characteristic data of the target user.
According to the technical scheme of the embodiment of the invention, a target recommendation algorithm is determined according to whether the characteristic data of the target user meets a first preset condition; and determining recommended learning resources of the target user by using a target recommendation algorithm and the characteristic data of the target user. According to the embodiment of the invention, a fixed recommendation algorithm is not adopted, but a target recommendation algorithm is determined according to the characteristic data of the user, so that the obtained recommended learning resources are more accurate, and the requirements of the user can be better met.
The above-described embodiments should not be construed as limiting the scope of the invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions can occur, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (15)
1. A method for determining a recommended learning resource, comprising:
determining characteristic data of a target user;
judging whether the characteristic data meet a first preset condition or not;
determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition;
and determining the recommended learning resources of the target user by using the target recommendation algorithm and the characteristic data of the target user.
2. The method of claim 1, wherein the characterization data comprises: a user interest tag;
the judging whether the feature data meet a first preset condition includes:
and judging whether the number of the user interest tags is smaller than a first number threshold value.
3. The method according to claim 2, wherein the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is smaller than a first number threshold, determining a clustering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining a user group of the target user by utilizing a clustering algorithm and the characteristic data of the target user;
and determining recommended learning resources of the target user according to the user group.
4. The method according to claim 2, wherein the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining an association rule algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining the recommended learning resources of the target user according to the similar users.
5. The method according to claim 2, wherein the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, determining a collaborative filtering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
6. The method according to claim 2, wherein the determining a target recommendation algorithm according to the first determination result corresponding to the first preset condition includes:
if the first judgment result represents that the number of the user interest tags is larger than or equal to a first number threshold, judging whether the number of the users in the system meets a second preset condition, wherein the second preset condition comprises that: the number of users in the system is less than a second number threshold;
and determining a target recommendation algorithm according to a second judgment result corresponding to the second preset condition.
7. The method according to claim 6, wherein the determining a target recommendation algorithm according to the second determination result corresponding to the second preset condition includes:
if the second judgment result represents that the number of the users in the system is greater than or equal to the second number threshold, determining an association rule algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
determining similar users of the target user by using an association rule algorithm and the characteristic data of the target user;
and determining the recommended learning resources of the target user according to the similar users.
8. The method according to claim 6, wherein the determining a target recommendation algorithm according to the second determination result corresponding to the second preset condition includes:
if the second judgment result represents that the number of the users in the system is smaller than the second number threshold, determining the collaborative filtering algorithm as a target recommendation algorithm;
the determining the recommended learning resources of the target user by using the target recommendation algorithm and the feature data of the target user includes:
and determining the recommended learning resources of the target user by utilizing a collaborative filtering algorithm and the characteristic data of the target user.
9. The method of claim 1, further comprising:
acquiring a first training set, wherein the first training set comprises a plurality of learning resources;
acquiring first behavior data of a test user on the first training set;
determining a first recommendation list of the test user according to the first behavior data;
determining a first behavior list of the test user by using the target recommendation algorithm, the feature data of the test user and the first training set;
and determining the accuracy of the determination method of the recommended learning resources according to the first recommendation list and the first behavior list.
10. The method of claim 1, further comprising:
acquiring a second training set, wherein the second training set comprises a plurality of learning resources;
acquiring second behavior data of the test user on the second training set;
determining a second recommendation list of the test user according to the second behavior data;
determining a second behavior list of the test user by using the target recommendation algorithm, the feature data of the test user and the second training set;
and determining the recall rate of the determination method of the recommended learning resources according to the second recommendation list and the second behavior list.
11. The method of claim 1, wherein determining the characteristic data of the target user comprises:
acquiring a log file of the target user;
pre-processing the log file, the pre-processing comprising one of: word segmentation processing, feature word recognition and invalid character filtering;
and performing statistical analysis on the processed log file to determine the characteristic data of the target user.
12. The method of claim 1, wherein the characterization data comprises: explicit data and implicit data;
the explicit data includes at least one of: resource comment, resource approval, resource forwarding, resource downloading, user grading feedback, resource downloading, question making record, course resource searching, course interaction times, interaction time of each time and system online time;
the stealth data includes at least one of: browsing history, reading duration, viewing records and searching logs.
13. An apparatus for determining a recommended learning resource, comprising:
the data determination module is used for determining the characteristic data of the target user;
the condition judgment module is used for judging whether the characteristic data meet a first preset condition or not;
the algorithm determining module is used for determining a target recommendation algorithm according to a first judgment result corresponding to the first preset condition;
and the resource determining module is used for determining the recommended learning resources of the target user by utilizing the target recommendation algorithm and the characteristic data of the target user.
14. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-12.
15. A computer-readable medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-12.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110405603.XA CN113094584A (en) | 2021-04-15 | 2021-04-15 | Method and device for determining recommended learning resources |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110405603.XA CN113094584A (en) | 2021-04-15 | 2021-04-15 | Method and device for determining recommended learning resources |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113094584A true CN113094584A (en) | 2021-07-09 |
Family
ID=76677727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110405603.XA Pending CN113094584A (en) | 2021-04-15 | 2021-04-15 | Method and device for determining recommended learning resources |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113094584A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113724040A (en) * | 2021-08-17 | 2021-11-30 | 卓尔智联(武汉)研究院有限公司 | Course recommendation method, electronic device and storage medium |
-
2021
- 2021-04-15 CN CN202110405603.XA patent/CN113094584A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113724040A (en) * | 2021-08-17 | 2021-11-30 | 卓尔智联(武汉)研究院有限公司 | Course recommendation method, electronic device and storage medium |
| CN113724040B (en) * | 2021-08-17 | 2023-11-28 | 卓尔智联(武汉)研究院有限公司 | Course recommendation method, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444428B (en) | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium | |
| Karimi et al. | News recommender systems–Survey and roads ahead | |
| WO2022041979A1 (en) | Information recommendation model training method and related device | |
| US10380158B2 (en) | System for determining and optimizing for relevance in match-making systems | |
| CN110717098B (en) | Meta-path-based context-aware user modeling method and sequence recommendation method | |
| CN110110233B (en) | Information processing method, device, medium and computing equipment | |
| CN111966914B (en) | Content recommendation method and device based on artificial intelligence and computer equipment | |
| US11874798B2 (en) | Smart dataset collection system | |
| CN111754278A (en) | Article recommendation method and device, computer storage medium and electronic equipment | |
| CN111429161B (en) | Feature extraction method, feature extraction device, storage medium and electronic equipment | |
| CN112819024B (en) | Model processing method, user data processing method and device and computer equipment | |
| CN114036398A (en) | Content recommendation and ranking model training method, device, equipment and storage medium | |
| CN110059172B (en) | Method and device for recommending answers based on natural language understanding | |
| CN114817692A (en) | Method, device and equipment for determining recommended object and computer storage medium | |
| CN113094584A (en) | Method and device for determining recommended learning resources | |
| CN115618024A (en) | Multimedia recommendation method and device and electronic equipment | |
| CN112632275B (en) | Crowd clustering data processing method, device and equipment based on personal text information | |
| CN112948561B (en) | Method and device for automatically expanding question-answer knowledge base | |
| CN112052402B (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN114707068A (en) | Method, device, equipment and medium for recommending intelligence base knowledge | |
| CN110929528B (en) | Method, device, server and storage medium for analyzing emotion of sentence | |
| CN111127057B (en) | Multi-dimensional user portrait recovery method | |
| CN113327147A (en) | Method and device for displaying article information | |
| CN112395487B (en) | Information recommendation method and device, computer readable storage medium and electronic equipment | |
| CN113220994B (en) | User personalized information recommendation method based on target object enhanced representation |
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 |