Information recommendation method and device, electronic equipment and storage medium

Abstract

The present disclosure provides an information recommendation method and device, electronic equipment and storage medium, and relates to the field of big data processing. The specific implementation scheme is: determining a target object having a correlation relationship with an initial object based on a relationship network; determining a target keyword based on a keyword target group; wherein the keyword target group contains a subscription keyword of the initial object, and the keyword target group is one of a plurality of keyword candidate groups; generating and sending recommendation information corresponding to the target object based on the target keyword. The above method can ensure the accuracy of the recommendation information.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more particularly to the field of big data processing technology. Background Technology

[0002] With the development of computer technology, more and more application functions are being launched, providing users with richer channels for information access. However, this necessitates accurately identifying users and recommending information that meets their needs. Therefore, ensuring the accuracy of information recommendations for different users becomes a problem that needs to be solved. Summary of the Invention

[0003] This disclosure provides an information recommendation method, apparatus, electronic device, and storage medium.

[0004] According to a first aspect of this disclosure, an information recommendation method is provided, comprising:

[0005] Based on the relationship network, identify target objects that are associated with the initial object;

[0006] Based on the keyword target group, target keywords are determined; wherein, the keyword target group includes the subscription keywords of the initial object, and the keyword target group is one of multiple keyword candidate groups;

[0007] Based on the target keyword, generate and send recommendation information corresponding to the target object.

[0008] According to a second aspect of this disclosure, an information recommendation device is provided, comprising:

[0009] The object selection module is used to determine target objects that have a relationship with the initial object based on the relationship network;

[0010] The keyword determination module is used to determine target keywords based on a keyword target group; wherein, the keyword target group includes the subscription keywords of the initial object, and the keyword target group is one of multiple keyword candidate groups;

[0011] The information generation module is used to generate recommendation information corresponding to the target object based on the target keyword;

[0012] The information sending module is used to send recommendation information corresponding to the target object.

[0013] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0014] At least one processor; and

[0015] The memory is communicatively connected to the at least one processor; wherein,

[0016] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the aforementioned information recommendation method of the first aspect.

[0017] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing the computer to perform the aforementioned method.

[0018] According to a fifth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the aforementioned method.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description.

[0020] The solution provided in this embodiment can determine the target object based on the association relationship of the initial object in the relationship network, then determine the target keywords based on the keyword target group, and finally generate and send the recommendation information corresponding to the target object based on the target keywords. In this way, since the target object and the initial object have an association relationship, and the keyword target group is related to the subscription keywords of the initial object, the accuracy and relevance of the selected target keywords can be guaranteed to be high, thereby ensuring the accuracy of the recommendation information generated for the target object. Attached Figure Description

[0021] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0022] Figure 1 This is a flowchart illustrating an information recommendation method according to an embodiment of the present disclosure;

[0023] Figure 2 This is a schematic diagram of the composition structure of a relational network according to an embodiment of the present disclosure;

[0024] Figure 3 This is a schematic diagram of the weight values ​​of each candidate object in a relational network according to an embodiment of the present disclosure;

[0025] Figure 4 This is a schematic diagram of the association values ​​between candidate objects in a relational network according to an embodiment of the present disclosure;

[0026] Figure 5 This is a schematic diagram illustrating the composition of a keyword target group according to an embodiment of the present disclosure;

[0027] Figure 6This is a schematic diagram of a process for generating and sending recommendation information according to an embodiment of the present disclosure;

[0028] Figure 7 This is a schematic diagram of a process for determining a new target object according to an embodiment of the present disclosure;

[0029] Figure 8 This is a schematic diagram of a scenario for generating a relational subnetwork according to an embodiment of the present disclosure;

[0030] Figure 9 This is a schematic diagram illustrating a scenario in which a relational network is obtained by merging relational subnetworks according to an embodiment of the present disclosure;

[0031] Figure 10 This is a schematic diagram of a scenario for generating a keyword network according to an embodiment of the present disclosure;

[0032] Figure 11 This is a schematic diagram illustrating a scenario of obtaining a keyword candidate group based on a keyword network according to an embodiment of the present disclosure;

[0033] Figure 12 This is a schematic diagram illustrating an exemplary processing procedure of an information recommendation method according to an embodiment of the present disclosure;

[0034] Figure 13 This is a schematic diagram of the composition structure of an information recommendation device according to another embodiment of the present disclosure;

[0035] Figure 14 This is a schematic diagram of another component structure of an information recommendation device according to another embodiment of the present disclosure;

[0036] Figure 15 This is a block diagram of an electronic device used to implement the information recommendation method of the embodiments of this disclosure. Detailed Implementation

[0037] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0038] The first aspect of this disclosure provides an information recommendation method, such as... Figure 1 As shown, it includes:

[0039] S101: Based on the relationship network, identify the target object that has an association with the initial object;

[0040] S102: Determine target keywords based on keyword target groups; wherein, the keyword target groups include the subscription keywords of the initial object, and the keyword target groups are one of multiple keyword candidate groups;

[0041] S103: Based on the target keywords, generate and send recommendation information corresponding to the target object.

[0042] The solution provided in this embodiment can be applied to electronic devices, especially servers.

[0043] The relationship network may contain multiple candidate objects, each of which can be a node in the relationship network. Each candidate object in the relationship network is associated with at least one other candidate object. For example, see... Figure 2 A schematic diagram illustrating a relationship network. Figure 2 It includes candidate objects 1 to 8, where candidate object 1 is associated with candidate object 2, candidate object 3 and candidate object 4 respectively; candidate object 5 is associated with candidate object 4, candidate object 6 and candidate object 7 respectively; and candidate object 8 is associated only with candidate object 6. Figure 2 Candidate 2 is only associated with candidate 1.

[0044] The step of determining the target object that has an association with the initial object based on the relationship network can specifically be: taking the candidate objects that have an association with the initial object in the relationship network as the target objects.

[0045] The number of candidate objects associated with the initial object may be one or more. If the initial object is associated with only one candidate object, then that candidate object is one target object. If the initial object is associated with multiple candidate objects, each of the multiple candidate objects can be used as the target object. Figure 2 For example, assuming the initial object is candidate object 1, then the initial object has the following relationships: candidate object 2, candidate object 3, and candidate object 4; and candidate object 2, candidate object 3, and candidate object 4 can be the target object selected in this instance.

[0046] It should be noted that when there are multiple target objects, the processing for any one of them is the same. Therefore, this embodiment will be described for any one of the target objects in the following description.

[0047] The keyword target group can be determined by: obtaining the subscription keywords of the initial object; and selecting a keyword candidate group containing the subscription keywords from the plurality of keyword candidate groups as the keyword target group.

[0048] Each of the multiple keyword candidate groups contains one or more candidate keywords; and different keyword candidate groups may contain different candidate keywords.

[0049] Determining the target keyword from the keyword target group can specifically involve using the subscription keyword of the initial object in the keyword target group, and one or more other candidate keywords in the keyword target group, as the target keyword. The number of target keywords is not limited in this embodiment.

[0050] The step of generating and sending recommendation information corresponding to the target object based on the target keywords may include: generating and sending recommendation information corresponding to the target object based on the target keywords and relevant information of the target object. The relevant information of the target object may be at least one of the target object's network address, the target object's identifier, and the target object's account.

[0051] As can be seen, by adopting the above scheme, the target object can be determined based on the association relationship of the initial object in the relationship network, then the target keywords can be determined based on the keyword target group, and finally, recommendation information corresponding to the target object can be generated and sent based on the target keywords. In this way, since the target object and the initial object are associated, and the keyword target group is related to the subscription keywords of the initial object, the accuracy and relevance of the selected target keywords can be guaranteed to be high, thereby ensuring the accuracy of the recommendation information generated for the target object.

[0052] In one implementation, the method for selecting the initial object may include:

[0053] Obtain multiple candidate objects in the relationship network and their corresponding weight values;

[0054] The candidate object whose weight value is greater than the preset weight threshold is selected from the plurality of candidate objects as the initial object.

[0055] Specifically, the relationship network may contain multiple candidate objects, and each candidate object is associated with at least one other candidate object. Furthermore, each candidate object in the relationship network will also have a corresponding weight value; wherein, the weight value corresponding to each candidate object can be used to represent the activity level of the candidate object.

[0056] The step of selecting candidate objects whose weight values ​​are greater than a preset weight threshold from the plurality of candidate objects as the initial objects can be:

[0057] At least one candidate object whose weight value is greater than a preset weight threshold is selected from the plurality of candidate objects as the initial object.

[0058] The number of initial objects can be one or more, and this embodiment does not limit it. When there are multiple initial objects, since the processing for each initial object is the same, this embodiment will not go into detail. In the following embodiments, one initial object will be used as an example for explanation.

[0059] The aforementioned preset weight threshold can be set according to the actual situation. It can be a specific value, such as 30; or it can be a ratio value. When the preset weight threshold is a ratio value, anything greater than the preset weight threshold value means selecting the top ratio of candidate objects from the highest weight values. The ratio value can be set according to the actual situation, such as 20%.

[0060] To be specific, with Figure 3 For example, in the relationship network, candidate object 1 has a weight of 20, candidate object 2 has a weight of 7, candidate object 3 has a weight of 10, candidate object 4 has a weight of 3, candidate object 5 has a weight of 30, candidate object 6 has a weight of 21, candidate object 7 has a weight of 12, and candidate object 8 has a weight of 6. Assuming the preset weight threshold is an integer 25, then... Figure 3 Candidate object 1 can be one initial object, and candidate object 5 can be another initial object. Assuming the preset weight threshold is 20%, that is, selecting the top 25% of candidate objects with the largest weight values, these will also be the initial objects for candidate object 1 and candidate object 5 respectively.

[0061] As can be seen, by adopting the above scheme, candidate objects with weight values ​​greater than a preset weight threshold can be selected from the relationship network as the initial objects. This ensures that the selected initial objects are those with high activity levels or high loyalty to the target product. Selecting subsequent target objects based on these initial objects guarantees the accuracy of target object location, thus making the generated recommendation information for the target objects more accurate.

[0062] In one implementation, determining the target object that has an association with the initial object based on the relationship network can be done in the following two ways:

[0063] Method 1: All candidate objects in the relationship network that have an association with the initial object are taken as the target object.

[0064] by Figure 3 For example, assuming the initial object is candidate object 1, all objects that are associated with candidate object 1 include candidate object 2, candidate object 3, and candidate object 4. The objects corresponding to candidate object 2, candidate object 3, and candidate object 4 can all be used as target objects.

[0065] Method 2: Based on the relationship network, determine at least one candidate target object that has an association with the initial object; determine the target object based on the association values ​​between the initial object and the at least one candidate target object in the relationship network.

[0066] The step of determining at least one candidate target object based on the association relationship of the initial object in the relationship network may include: taking objects that the initial object is associated with in the relationship network as the at least one candidate target object.

[0067] The step of determining the target object based on the association values ​​between the initial object and the at least one candidate target object in the relationship network can be as follows: based on the relationship network, determine the association values ​​between the initial object and the at least one candidate target object respectively; from the at least one candidate target object, select the candidate target object whose association value is greater than a preset association threshold value as the target object.

[0068] Specifically, the relationship network may also include the association value between each candidate object and its other associated candidate objects. This association value can be used to represent the degree of association between each candidate object and other candidate objects. Based on this association value, the stability of the social relationship between a candidate object and other candidate objects can be determined.

[0069] The preset association threshold value can be set according to the actual situation, such as a numerical value or a proportional value.

[0070] For example, see Figure 4 With the initial object as Figure 3 Taking candidate object 1 as an example, the association value between candidate object 1 and candidate object 2 is 20, the association value between candidate object 1 and candidate object 3 is 10, and the association value between candidate object 1 and candidate object 4 is 5. Assuming the preset association threshold is 15, that is... Figure 4 Candidate object 2 is a candidate target object whose association value is greater than the preset association threshold value, and is therefore selected as the target object. For example, the preset association threshold value can be 60%, meaning that the top 60% of candidate objects with the highest association values ​​with candidate object 1 are selected as the target objects. Figure 4 Candidate object 2 and candidate object 3 are respectively designated as the target object.

[0071] It should be understood that one or more target objects can be identified based on the above two methods, and the processing method for each target object is the same. Therefore, this embodiment will describe any one of the target objects in the following description, without going into detail about each one.

[0072] As can be seen, by adopting the above scheme, the target object can be determined based on the association value between the initial object and other candidate target objects in the relationship network. In this way, the target object can be selected more accurately, and the accuracy of the recommendation information generated for the target object will also be higher.

[0073] In one implementation, determining the target keywords based on the keyword target group may include: prioritizing the multiple candidate keywords in the keyword target group to obtain N priority target keywords; where N is an integer greater than or equal to 2.

[0074] Here, N can be set according to the actual situation. For example, it can be equal to 3, or equal to 2, or other values. We will not exhaustively list them here.

[0075] The target keyword group can be one of multiple candidate keyword groups obtained from the keyword network.

[0076] In this embodiment, the method for determining the keyword target group may include:

[0077] Obtain the keywords subscribed by the initial user; among the multiple keyword candidate groups, select at least one keyword candidate group that contains the keywords subscribed by the initial user as the keyword target group.

[0078] It should be understood that the keywords subscribed by the initial user can be one or more. Typically, the keywords subscribed by an initial user may appear in the same keyword candidate group or in multiple keyword candidate groups. In this case, all of these multiple keyword candidate groups can be regarded as the keyword target group.

[0079] The keyword target group may include multiple candidate keywords, each of which corresponds to a node value. The node value of each candidate keyword can be used to represent the number of times the candidate keyword has been subscribed. In other words, the acceptance level of the candidate keyword among users can be known through the node value of the candidate keyword. A higher node value indicates that the candidate keyword is highly accepted by users, while a lower node value indicates that the candidate keyword is poorly accepted by users.

[0080] For example, see Figure 5 The keyword target group includes candidate keyword 1 to candidate keyword 5, and each candidate keyword has a corresponding node value, for example... Figure 5 The diagram shows that the node value of candidate keyword 1 is 20, the node value of candidate keyword 2 is 10, the node value of candidate keyword 3 is 1, the node value of candidate keyword 4 is 50, and the node value of candidate keyword 5 is 5.

[0081] The step of prioritizing the multiple candidate keywords in the keyword target group to obtain N priority target keywords can be achieved by determining the N priority target keywords based on the node values ​​corresponding to the multiple candidate keywords. For example, taking N as 2, the candidate keyword with the highest first proportion of node values ​​can be taken as the highest priority target keyword, and the remaining candidate keywords can be taken as the lowest priority target keywords.

[0082] By prioritizing the multiple candidate keywords in the target keyword group as described above, N priority target keywords are obtained, which can provide a more accurate range division for the subsequent generation and sending of the recommendation information corresponding to the target object, thus ensuring recommendation efficiency.

[0083] Based on the above embodiments, taking N equal to 3 as an example, an exemplary description is provided, in which the priority division of the multiple candidate keywords in the keyword target group to obtain N priority target keywords includes:

[0084] The subscription keywords of the initial object in the keyword target group are taken as the target keywords with the first priority.

[0085] Candidate keywords with node values ​​greater than a preset node threshold in the keyword target group are designated as target keywords with the second priority.

[0086] Candidate keywords whose node values ​​are less than or equal to the preset node threshold value in the keyword target group are designated as target keywords with the third priority.

[0087] In other words, the subscription keywords of the initial object can be used as the highest priority target keywords. This is because the target object and the initial object are related, so using the subscription keywords of the initial object as the first priority target keywords allows the recommendation information generated based on these first priority target keywords to more accurately meet the needs of the target object. The number of subscription keywords of the initial object is not limited and can include one or more.

[0088] Furthermore, the solution provided in this embodiment takes into account that the needs of the target object may not be exactly the same as those of the initial object. Therefore, the target keywords with second priority and third priority are further selected through the keyword target group. The selection method is based on the preset node threshold value.

[0089] The preset node threshold value can be set according to actual conditions; for example, it can be an integer value or a proportional value. See also... Figure 5 Assuming a preset node threshold of 40, then keywords exceeding this threshold are candidate keyword 4, and correspondingly, candidate keyword 4 can be considered the target keyword with the second priority. Assuming a preset node threshold of the top 10%, then candidate keywords exceeding this threshold are... Figure 5 In the context of keyword 4, candidate keyword 4 can be used as the target keyword with the second priority.

[0090] Furthermore, other candidate keywords that are less than or equal to the preset threshold value can be used as the target keywords with the third priority.

[0091] Additionally, the target keywords of the second or third priority may include the subscription keywords of the initial object. In this case, the subscription keywords of the initial object can be deleted.

[0092] Finally, it should be noted that the number of the first priority target keywords, the second priority target keywords, and the third priority target keywords can all be one or more, and this embodiment does not limit their possible number.

[0093] As can be seen, by adopting the above scheme, multiple priority target keywords can be determined based on the keyword target group. Thus, when generating recommendation information, the corresponding recommendation information can be generated according to the priority order, thereby ensuring that the recommendation information that is more likely to meet the needs of the target user is recommended first, thus ensuring the accuracy of the recommendation.

[0094] Accordingly, based on the candidate keywords, recommendation information corresponding to the target object is generated and sent, such as... Figure 6 As shown, it may include:

[0095] S601: Based on the target keyword with the i-th priority among the N priorities, generate and send the i-th recommendation information corresponding to the target object; i is an integer greater than or equal to 1 and less than N;

[0096] S602: Upon receiving a rejection message for the i-th recommendation, generate and send the (i+1)-th recommendation for the target object based on the target keyword of the (i+1)-th priority among the N priorities.

[0097] The i-th priority is any one of the N priorities.

[0098] In other words, recommendation information corresponding to the target object is generated sequentially based on the order of the target keywords from high to low among the N priorities.

[0099] The step of generating and sending the i-th recommendation information corresponding to the target object based on the target keyword of the i-th priority among the N priorities can specifically be: generating the i-th recommendation information corresponding to the target object based on the relevant information of the target object and the target keyword of the i-th priority. The relevant information of the target object can be at least one of the target object's network address, the target object's account, the target object's identifier name, etc.; that is, using the target keyword of the i-th priority as the specific content of the i-th recommendation information, and using the network address of the device currently used by the target object and the target object's account as the address-related information of the i-th recommendation information, the i-th recommendation information corresponding to the target object can be obtained. It should be understood that sending the i-th recommendation information corresponding to the target object can be: sending the i-th recommendation information of the target object to the device used by the target object, for example, it can be any one of the terminal devices used by the target object, such as mobile phones, personal computers, and tablet computers.

[0100] After sending the i-th recommendation information corresponding to the target object to the device used by the target object, it can be determined whether the target object accepts or rejects the i-th recommendation information. For example, if the target keyword of the i-th priority contained in the i-th recommendation information is displayed on the device used by the target object, and the target object does not want to subscribe to any of the target keywords of the i-th priority, it can click the reject button (which can be a virtual button). Accordingly, the electronic device executing this embodiment can receive a rejection message for the i-th recommendation information. If the target object wants to subscribe to any one or more target keywords of the i-th priority, it can click the accept button (which can be a virtual button). Accordingly, the electronic device executing this embodiment can receive an accept message for the i-th recommendation information.

[0101] Furthermore, upon receiving the acceptance information for the i-th recommendation information, the subsequent processing of generating and sending the i+1 recommendation information corresponding to the target object based on the target keyword with the (i+1)-th priority can be skipped.

[0102] Upon receiving a rejection message for the i-th recommendation, the i+1 recommendation information corresponding to the target object can be generated based on the target keyword with priority i+1. After generating the i+1 recommendation information for the target object, the i+1 recommendation information can also be sent to the device used by the target object. For example, it can be a terminal device used by the target object, such as any one of a mobile phone, personal computer, or tablet computer.

[0103] The process of generating the i+1 recommendation information corresponding to the target object based on the target keyword with the i+1 priority, and the process of sending the i+1 recommendation information to the device used by the target object, are similar to the aforementioned process of generating the i-th recommendation information corresponding to the target object based on the target keyword with the i-th priority, and the process of sending the i-th recommendation information to the device used by the target object, and will not be repeated here.

[0104] Regarding the above processing, it should be pointed out by example that this solution can divide the candidate keywords in the keyword target group into three priorities, and when generating corresponding recommendation information, determine whether to generate and send recommendation information of the next priority based on the detected operation of recommendation information for different priorities. In actual processing, the keyword target group can also be divided into more or fewer priorities, all of which are within the protection scope of this embodiment, but are not exhaustively listed.

[0105] As can be seen, by adopting the above scheme, recommendation information corresponding to the target object can be generated sequentially based on target keywords of different priorities. In this way, the recommendation information corresponding to the target object can be generated more specifically, avoiding the problem of inconvenience in searching or selecting caused by generating recommendation information corresponding to the target object based on a large number of target keywords at once. This can generate the recommendation information corresponding to the target object more efficiently and accurately.

[0106] In one implementation, after generating and sending recommendation information corresponding to the target object based on the target keywords, such as... Figure 7 As shown, the method further includes:

[0107] S701: Upon receiving acceptance information for the recommendation information, the target object is used as the new initial object;

[0108] S702: Based on the relationship network, determine a new target object that has an association with the new initial object.

[0109] If the target object has subscribed to at least one target keyword contained in the recommendation information, it can receive an acceptance message for the recommendation information.

[0110] The process of determining the new target object that has an association with the new initial object based on the relationship network can include the following two methods:

[0111] Method 1: Take all objects in the relationship network that have an association with the new initial object as the new target object.

[0112] Method 2: Based on the association relationship of the new initial object in the relationship network, determine at least one candidate target object; based on the association value between the new initial object and the at least one candidate target object, determine the new target object.

[0113] The step of determining at least one candidate target object based on the association relationship of the new initial object in the relationship network may include: taking objects that are associated with the new initial object in the relationship network as the at least one candidate target object.

[0114] Determining the new target object based on the association values ​​between the new initial object and the at least one candidate target object can be achieved by: determining the association values ​​between the new initial object and the at least one candidate target object based on the relationship network; and selecting candidate target objects from the at least one candidate target object whose association values ​​are greater than a preset association threshold as the new target object.

[0115] The specific processing methods described above are similar to those described above for determining the target object based on the association relationship of the initial object in the relationship network, so they will not be repeated here.

[0116] It should be noted that after determining the new target object, the process may further include: determining new target keywords based on the new keyword target group; wherein the new keyword target group includes the subscription keywords of the new initial object; the keyword target group is one of at least one keyword candidate group; and generating and sending recommendation information corresponding to the new target object based on the new target keywords. The above processing is similar to the aforementioned process of determining target keywords and generating and sending recommendation information corresponding to the target object, and will not be repeated here.

[0117] By adopting the solution provided in this embodiment, not only can more accurate recommendation information be generated for the selected target objects, but the selected target objects can also be used as new initial objects to select more new target objects to generate and send corresponding recommendation information. In this way, more objects can be introduced into the system, and then relationship networks can be generated and keyword candidate groups can be updated based on more objects, further improving the accuracy of subsequent recommendation information generation.

[0118] In one implementation, the method for constructing the aforementioned relationship network is described, which may specifically include:

[0119] The relationship network is obtained by merging multiple relational subnetworks.

[0120] Based on the occurrence counts of each candidate object in the multiple relational sub-networks, the weight values ​​corresponding to each candidate object in the relational network are determined; based on the occurrence counts of the association relationships between the candidate objects in the multiple relational sub-networks, the association values ​​between the candidate objects in the relational network are determined.

[0121] The generation of the j-th relational subnetwork among the plurality of relational subnetworks can be achieved by generating a j-th relational subnetwork containing the plurality of candidate objects when there are interactive operations between them; where j is an integer greater than or equal to 1. The interactive operation can be any one of a sharing operation, a recommendation operation, or an invitation operation.

[0122] It should be noted that different relational subnetworks in the plurality of relational subnetworks may contain the same multiple candidate objects, or may contain at least some different candidate objects.

[0123] For example, see Figure 8Suppose candidate a invites candidate b to participate in activity 1, candidate b invites candidate c to participate in activity 1, candidate a invites candidate d to participate in activity 1, and candidate c invites candidate e to participate in activity 1. Then, we can construct relational subnetwork 1. See also... Figure 8 If candidate a invites candidate d and candidate f to participate in activity 2, candidate b invites candidate c to participate in activity 2, and candidate c invites candidate g to participate in activity 2, then we can construct the relational subnetwork 2.

[0124] The process of merging multiple relational subnetworks to obtain the relational network may include:

[0125] If the x-th relation subnetwork and the y-th relation subnetwork have the same candidate object, the same candidate object is used as the common candidate object, and the x-th relation subnetwork and the y-th relation subnetwork are merged. x and y are different, and x and y are both integers greater than or equal to 1. This process is repeated until all relation subnetworks are merged to obtain the relation network.

[0126] See Figure 9 Relational subnetworks 1 and 2 both contain candidate objects a, b, c, and e. These candidate objects can be merged as common candidate objects to obtain... Figure 9 The diagram illustrates the merged relationship network.

[0127] The step of determining the weight values ​​of each candidate object in the relation network based on the occurrence counts of each candidate object in the multiple relation sub-networks can be as follows: Obtain the occurrence count of the r-th candidate object in the relation network within the multiple relation sub-networks, and use this occurrence count as the weight value of the r-th candidate object in the relation network; where r is an integer greater than or equal to 1. It should be understood that the r-th candidate object is any one of all candidate objects in the relation network, and the processing for each candidate object is the same, therefore it will not be elaborated upon individually. For example, see... Figure 9 If candidate node 'a' appears twice in both relational subnetworks, its corresponding weight value is 2. Other candidate nodes can also have their weight values ​​determined in the same way. The specific weight values ​​for each candidate node are as follows: Figure 9 The examples shown are simply not listed one by one.

[0128] The association value among the candidate objects in the relational network is determined based on the frequency of occurrence of the association relationships among the candidate objects in the multiple relational sub-networks. For example, see [link to relevant documentation]. Figure 9In this relationship network, the association between candidate object a and candidate object d appears twice. Therefore, the association value between candidate object a and candidate object d is [value missing]. Figure 9 The method for determining the association values ​​between other candidate nodes is the same as shown in the example of "2", but they are not listed one by one.

[0129] Through the above processing, the association relationships between all the candidate objects collected can be obtained to generate corresponding relationship subnetworks. Then, by merging the various relationship subnetworks, the relationship network can be obtained. Furthermore, the weight values ​​of the candidate objects in the relationship network and the association values ​​between the candidate objects can be determined. Thus, it can provide an accurate reference for selecting target objects based on the relationship network, ensuring the accuracy of the recommendation information obtained in subsequent processing.

[0130] In one implementation, the method for generating the aforementioned multiple keyword candidate groups is described, including:

[0131] A keyword network is generated based on multiple candidate keywords subscribed by multiple candidate objects;

[0132] Based on the subscription counts corresponding to the multiple candidate keywords, the node values ​​of the multiple candidate keywords in the keyword network are determined; based on the occurrence counts of the association relationships between the candidate keywords, the association values ​​of the multiple candidate keywords in the keyword network are determined.

[0133] Based on the association relationships and association values ​​among the multiple candidate keywords in the keyword network, the multiple keyword candidate groups are generated.

[0134] The number of candidate objects can be two or more, and this embodiment does not limit it.

[0135] The generation of a keyword network based on multiple candidate keywords subscribed to by multiple candidate objects can be achieved by: establishing a connection relationship between at least two candidate keywords subscribed to by each candidate object, and merging the at least two candidate keywords with connection relationships corresponding to each candidate object to obtain the keyword network.

[0136] Among the plurality of candidate objects, each candidate object may contain one or more candidate keywords that it has subscribed to. If the qth candidate object among the plurality of candidate objects contains only one candidate keyword, then that candidate keyword can exist independently. If the qth candidate object among the plurality of candidate objects contains at least two candidate keywords, then the at least two candidate keywords are associated with each other.

[0137] For example, see Figure 10Candidate object a has three candidate keywords, namely candidate keyword 1 to candidate keyword 3, which are related to each other; candidate object b has candidate keyword 2 and candidate keyword 5, which are also related to each other; candidate object c has two keywords, candidate keyword 4 and candidate keyword 6, which are related to each other; candidate object d has two keywords, candidate keyword 1 and candidate keyword 6, which are related to each other. Candidate keyword 1 can be a common candidate keyword, and the two sets of candidate keywords for candidate objects a and d are merged; candidate keyword 2 can be a common candidate for two sets of candidate keywords, and these two sets of candidate keywords are merged. This process is repeated until the keyword network is obtained.

[0138] The node value of each candidate keyword in the keyword network is determined based on the subscription count corresponding to each candidate keyword. For example, Figure 10 If candidate keyword 1 appears twice, then the node value of candidate keyword 1 in the keyword network is 2. If candidate keyword 2 appears twice, then the node value of candidate keyword 2 in the keyword network is 2. It should be understood that the node values ​​of the other candidate keywords are determined in the same way, but not exhaustively.

[0139] Determining the association value of the multiple candidate keywords in the keyword network based on the frequency of occurrence of the association relationship between the candidate keywords includes: determining the association value between any two candidate keywords in the keyword network based on the frequency of occurrence of the association relationship between any two candidate keywords. For example, Figure 10 If the association between candidate keyword 1 and candidate keyword 2 occurs once, then the association value between keyword 1 and keyword 2 is 1.

[0140] The step of generating multiple candidate keyword groups based on the association relationships and association values ​​among the multiple candidate keywords in the keyword network may include: calculating the distance between any two candidate keywords based on the association relationships and association values ​​among at least two candidate keywords; and generating the at least one candidate keyword group based on the distance between any two candidate keywords. The method for calculating the distance between any two candidate keywords can be the k-means clustering algorithm; of course, other algorithms can also be used, but this embodiment does not exhaustively list them.

[0141] The process of generating the at least one keyword candidate group based on the distance between any two candidate keywords can be described as follows: Taking the positions where the distance between any two candidate keywords exceeds a preset distance threshold as grouping positions, at least one keyword candidate group is obtained. Figure 11 For example, if the distance between candidate keyword 8 and other keywords is greater than a preset distance threshold, then candidate keyword 8 and its associated candidate keyword 7 are grouped together as a single keyword candidate group; if the distance between candidate keyword 1 and candidate keyword 6 is greater than a preset distance threshold, then candidate keyword 1 and candidate keyword 6 are separated into two keyword candidate groups. After this division, Figure 11 Three keyword candidate groups can be obtained, namely: Figure 11 The keyword candidate group 1, keyword candidate group 2 and keyword candidate group 3 are shown in the figure.

[0142] As can be seen, by adopting the above scheme, candidate keywords can be associated, and at least one keyword candidate group can be obtained based on the association relationship and association value between at least two candidate keywords. In this way, more accurate information can be provided for the selection of target keywords based on the keyword target group of the initial object and for the generation of corresponding recommendation information for the target object, thus ensuring the accuracy of the subsequent generation of recommendation information.

[0143] Finally combined Figure 12 Taking the initial object as the initial user, the target object as the target user, and the candidate object as the candidate user as an example, this embodiment provides an implementation scenario. In this scenario, the solution provided by this embodiment can be applied to the recommendation processing of academic papers. Specifically:

[0144] S1201: Generate a relational network.

[0145] Specifically, firstly, based on the interaction behavior of each candidate user in multiple sharing activities, multiple relational sub-networks corresponding to the multiple sharing activities are obtained; the relational sub-networks contain the association relationships between the multiple candidate users, which can be obtained based on the sharing behavior or sharing relationships between each candidate user;

[0146] The multiple relational subnetworks are merged to obtain a relational network. The number of times each candidate user appears in the multiple relational subnetworks can be used as the weight value of the candidate user in the relational network. The number of times the relationship between any two candidate users that are related can be used as the association value between the two candidate users in the relational network. For example, the number of activities that candidate user a invites candidate user b to participate in can be used as the association value between candidate user a and candidate user b.

[0147] S1202: Generate keyword network.

[0148] Specifically, the candidate keywords subscribed to by each candidate user are obtained, and the relationship between the different candidate keywords subscribed to by each candidate user is established. Finally, the relationship between all candidate keywords of all candidate users is obtained to generate the keyword network. For example, if candidate user A subscribes to candidate keyword a and candidate keyword b, then candidate keyword a and candidate keyword b are related.

[0149] The node value of the multiple candidate keywords in the keyword network can be determined based on the number of subscriptions corresponding to the multiple candidate keywords respectively; the association value of the multiple candidate keywords in the keyword network can be determined based on the number of occurrences of the association relationship between the candidate keywords. For example, if candidate user A subscribes to candidate keyword a and candidate keyword b, then candidate keyword a and candidate keyword b are associated. If P candidate users simultaneously subscribe to candidate keyword a and candidate keyword b, then the association value between candidate keyword a and candidate keyword b is P, where P is an integer greater than or equal to 2.

[0150] Based on the association relationships and association values ​​among the multiple candidate keywords in the keyword network, multiple keyword candidate groups are generated. For example, the association relationships among multiple candidate keywords are used as cohesion relationships. The multiple keyword candidate groups of the keyword network are obtained through k-means. At this time, each keyword candidate group can be considered as content that candidate users with similar or related cognitive ranges care about or pay attention to.

[0151] The execution of S1201 and S1202 can be done in any order; S1201 and S1202 can be executed simultaneously; or S1201 can be executed first, followed by S1202; or S1202 can be executed first, followed by S1201.

[0152] After generating the aforementioned relationship network and multiple keyword candidate groups, the following operations can be performed:

[0153] S1203: Extract the candidate users whose weight values ​​rank in the top 20% of the relation network as the initial users, and determine the keyword target group.

[0154] Since a higher weight value for a candidate user in the relationship network indicates a higher level of loyalty or activity, the candidate users whose weight values ​​rank in the top 20% of the relationship network can be extracted as the initial users, and the keyword target group of the initial users' subscription keywords can be determined.

[0155] S1204: Based on the relationship network, determine the target users who have an association with the initial user.

[0156] Then, based on the keyword target group, target keywords can be determined; based on the target keywords, recommendation information corresponding to the target user can be generated and sent, and this recommendation information may contain the target keywords. The rules for generating and sending the recommendation information corresponding to the target user can be as follows: the initial user's subscription keywords are used as the first priority target keywords to generate and send the recommendation information corresponding to the target user; secondly, candidate keywords in the keyword target group with node values ​​greater than a preset node threshold are used as the second priority target keywords, and recommendation information corresponding to the target user is generated and sent based on these second priority target keywords; finally, candidate keywords in the keyword target group with node values ​​less than or equal to the preset node threshold are used as the third priority target keywords, and recommendation information corresponding to the target user is generated and sent based on these third priority target keywords.

[0157] After the target user receives the recommendation information through the device they are using, they can choose whether to subscribe on the device they are using.

[0158] After completing the above processing, if the target user accepts the subscription to the target keywords in the recommendation information, an acceptance message for the recommendation information can be received. Correspondingly, upon receiving the acceptance message for the recommendation information, the target object is used as a new initial object, and the aforementioned processes S1203-S1204 are repeated. This processing can also improve the relationship network and the keyword network, thereby further enhancing the accuracy of generating and sending recommendation information to new target users.

[0159] The second aspect of this disclosure also provides an information recommendation device, such as... Figure 13 As shown, it includes:

[0160] The object selection module 1301 is used to determine target objects that have an association relationship with the initial object based on the relationship network;

[0161] The keyword determination module 1302 is used to determine target keywords based on a keyword target group; wherein, the keyword target group includes the subscription keywords of the initial object, and the keyword target group is one of multiple keyword candidate groups;

[0162] The information generation module 1303 is used to generate recommendation information corresponding to the target object based on the target keywords;

[0163] The information sending module 1304 is used to send recommendation information corresponding to the target object.

[0164] The object selection module 1301 is used to obtain multiple candidate objects in the relationship network and their respective weight values; and select candidate objects whose weight values ​​are greater than a preset weight threshold value from the multiple candidate objects as the initial objects.

[0165] The object selection module 1301 is used to determine at least one candidate target object that has an association relationship with the initial object based on the relationship network; and to determine the target object based on the association values ​​between the initial object and the at least one candidate target object in the relationship network.

[0166] The keyword determination module 1302 is used to prioritize the multiple candidate keywords in the keyword target group to obtain N priority target keywords; N is an integer greater than or equal to 2.

[0167] The information generation module 1303 is used to generate the i-th recommendation information corresponding to the target object based on the target keyword of the i-th priority among the N priorities; i is an integer greater than or equal to 1 and less than N; and when a rejection information for the i-th recommendation information is received, to generate the (i+1)-th recommendation information corresponding to the target object based on the target keyword of the (i+1)-th priority among the N priorities.

[0168] The information sending module 1304 is used to send the i-th recommendation information corresponding to the target object, and to send the (i+1)-th recommendation information corresponding to the target object.

[0169] The keyword determination module 1302 is used to select the subscription keyword of the initial object in the keyword target group as the first priority target keyword;

[0170] Candidate keywords with node values ​​greater than a preset node threshold in the keyword target group are designated as target keywords with the second priority.

[0171] Candidate keywords whose node values ​​are less than or equal to the preset node threshold value in the keyword target group are designated as target keywords with the third priority.

[0172] exist Figure 13 On the basis of, such as Figure 14 As shown, the device further includes:

[0173] Information receiving module 1305 is used to receive acceptance information for the recommendation information;

[0174] The object selection module 1301 is used to, upon receiving acceptance information for the recommendation information, use the target object as a new initial object; and determine a new target object that has an association with the new initial object based on the relationship network.

[0175] The device further includes:

[0176] The relationship network generation module 1306 is used to merge multiple relationship sub-networks to obtain the relationship network; determine the weight values ​​corresponding to the multiple candidate objects in the relationship network based on the occurrence frequency of the multiple candidate objects in the multiple relationship sub-networks; and determine the association value between the multiple candidate objects in the relationship network based on the occurrence frequency of the association relationship between the multiple candidate objects in the multiple relationship sub-networks.

[0177] The device further includes:

[0178] The keyword candidate cluster generation module 1307 is used to generate a keyword network based on multiple candidate keywords subscribed to by multiple candidate objects; determine the node value of the multiple candidate keywords in the keyword network based on the subscription count corresponding to each of the multiple candidate keywords; determine the association value of the multiple candidate keywords in the keyword network based on the occurrence count of the association relationship between the candidate keywords; and generate the multiple keyword candidate clusters based on the association relationship between the multiple candidate keywords in the keyword network and the association value.

[0179] By adopting the above scheme, target objects can be determined based on the association relationships of initial objects in the relationship network. Then, target keywords can be determined based on a keyword target group. Finally, recommendation information corresponding to the target object can be generated and sent based on the target keywords. In this way, since the target object and the initial object are associated, and the keyword target group is related to the initial object's subscribed keywords, the accuracy and relevance of the selected target keywords can be guaranteed to be high, thus ensuring that the recommendation information generated for the target object is more accurate.

[0180] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0181] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0182] Figure 15A schematic block diagram of an example electronic device 1500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0183] like Figure 15 As shown, the electronic device 1500 includes a computing unit 1501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1502 or a computer program loaded from a storage unit 1508 into a random access memory (RAM) 1503. The RAM 1503 may also store various programs and data required for the operation of the electronic device 1500. The computing unit 1501, ROM 1502, and RAM 1503 are interconnected via a bus 1504. An input / output (I / O) interface 1505 is also connected to the bus 1504.

[0184] Multiple components in electronic device 1500 are connected to I / O interface 1505, including: input unit 1506, such as keyboard, mouse, etc.; output unit 1507, such as various types of monitors, speakers, etc.; storage unit 1508, such as disk, optical disk, etc.; and communication unit 1509, such as network card, modem, wireless transceiver, etc. Communication unit 1509 allows electronic device 1500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0185] The computing unit 1501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1501 performs the various methods and processes described above. For example, in some embodiments, the various methods described above can be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage unit 1508. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 1500 via ROM 1502 and / or communication unit 1509. When the computer program is loaded into RAM 1503 and executed by the computing unit 1501, one or more steps of the various methods described above can be performed. Alternatively, in other embodiments, the computing unit 1501 can be configured to perform the various methods described above by any other suitable means (e.g., by means of firmware).

[0186] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0187] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0188] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0189] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0190] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0191] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0192] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0193] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for information recommendation, comprising: determining a target object having a correlation with an initial object based on a relation network; determining a target keyword based on a keyword target group, wherein the keyword target group comprises a subscription keyword of the initial object, the keyword target group is one of a plurality of keyword candidate groups, and the target keyword comprises the subscription keyword; and generating and sending recommendation information corresponding to the target object to the target object based on the target keyword, so that the target object determines whether to subscribe to the recommendation information. 2.The method of claim 1, further comprising: obtaining a plurality of candidate objects in the relation network and respective weight values corresponding to the plurality of candidate objects; and selecting a candidate object having a weight value greater than a preset weight threshold value from the plurality of candidate objects as the initial object. The method of claim 1, wherein the target object having a correlation with the initial object is determined based on the relation network, comprising: determining at least one candidate target object having a correlation with the initial object based on the relation network; and determining the target object based on correlation values between the initial object and the at least one candidate target object in the relation network. The method of claim 1, wherein the target keyword is determined based on the keyword target group, comprising: performing priority division on a plurality of candidate keywords in the keyword target group to obtain N target keywords of N priorities, wherein N is an integer greater than or equal to 2. The method of claim 1, wherein the recommendation information corresponding to the target object is generated and sent based on the candidate keyword, comprising: generating and sending an i th recommendation information corresponding to the target object based on a target keyword of an i th priority in the N priorities, wherein i is an integer greater than or equal to 1 and less than N; and in a case where rejection information for the i th recommendation information is received, generating and sending an i+1 th recommendation information corresponding to the target object based on a target keyword of an i+1 th priority in the N priorities. The method of claim 1, wherein the priority division on the plurality of candidate keywords in the keyword target group to obtain the N target keywords of the N priorities, comprising: taking the subscription keyword of the initial object in the keyword target group as a target keyword of a first priority; taking a candidate keyword having a node value greater than a preset node threshold value in the keyword target group as a target keyword of a second priority; and taking the candidate keyword having the node value less than or equal to the preset node threshold value in the keyword target group as a target keyword of a third priority. 7.The method of claim 1, further comprising: in a case where acceptance information for the recommendation information is received, taking the target object as a new initial object; and determining a new target object having a correlation with the new initial object based on the relation network.

3. The method of claim 1, wherein, 8.The method of any one of claims 1-7, further comprising: merging a plurality of relation sub-networks to obtain the relation network; determining weight values corresponding to a plurality of candidate objects in the relation network based on respective occurrence numbers of the plurality of candidate objects in the plurality of relation sub-networks. ​ ​ 4. The method of claim 1, wherein, ​ ​ 5. The method of claim 4, wherein, ​ ​ ​ 6. The method of claim 4, wherein, ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ determine the association values between the plurality of candidate objects in the relationship network based on the occurrence times of the association relationships between the plurality of candidate objects in the plurality of relationship sub-networks.

9. The method of any one of claims 1-7, further comprising: generating a keyword network based on a plurality of candidate keywords subscribed by a plurality of candidate objects; determining node values of the plurality of candidate keywords in the keyword network based on a plurality of subscription times corresponding to the plurality of candidate keywords respectively; determining association values of the plurality of candidate keywords in the keyword network based on the occurrence times of the association relationships between the plurality of candidate keywords; generating the plurality of keyword candidate groups based on the association relationships and the association values between the plurality of candidate keywords in the keyword network.

10. An information recommendation apparatus, comprising: an object selection module configured to determine a target object having an association relationship with an initial object based on a relationship network; a keyword determination module configured to determine a target keyword based on a keyword target group; wherein the keyword target group comprises a subscription keyword of the initial object, the keyword target group is one of a plurality of keyword candidate groups, and the target keyword comprises the subscription keyword; an information generation module configured to generate recommendation information corresponding to the target object based on the target keyword; an information sending module configured to send the recommendation information to the target object, so that the target object determines whether to subscribe to the recommendation information.

11. The apparatus of claim 10, wherein: the object selection module is configured to acquire a plurality of candidate objects in the relationship network and respective weight values corresponding to the plurality of candidate objects, and select a candidate object having a weight value greater than a preset weight threshold value from the plurality of candidate objects as the initial object.

12. The apparatus of claim 10, wherein: the object selection module is configured to determine at least one candidate target object having an association relationship with the initial object based on the relationship network, and determine the target object based on association values between the initial object and the at least one candidate target object in the relationship network.

13. The apparatus of claim 10, wherein: the keyword determination module is configured to divide a plurality of candidate keywords in the keyword target group into N priority target keywords, wherein N is an integer greater than or equal to 2.

14. The apparatus of claim 13, wherein: the information generation module is configured to generate an i-th recommendation information corresponding to the target object based on an i-th priority target keyword in the N priority target keywords, wherein i is an integer greater than or equal to 1 and less than N; in a case where rejection information for the i-th recommendation information is received, generate an i+1-th recommendation information corresponding to the target object based on an i+1-th priority target keyword in the N priority target keywords; the information sending module is configured to send the i-th recommendation information corresponding to the target object, and send the i+1-th recommendation information corresponding to the target object.

15. The apparatus of claim 13, wherein: ​ The keyword determining module is configured to determine the subscription keyword of the initial object in the keyword target group as a first-priority target keyword. The keyword determining module is configured to determine a candidate keyword in the keyword target group with a node value greater than a preset node threshold value as a second-priority target keyword. The keyword determining module is configured to determine the candidate keyword in the keyword target group with a node value less than or equal to the preset node threshold value as a third-priority target keyword.

16. The apparatus of claim 10, further comprising: an information receiving module configured to receive acceptance information for the recommendation information; the object selecting module is configured to determine the target object as a new initial object if the acceptance information for the recommendation information is received. The object selecting module is configured to determine a new target object having an association relationship with the new initial object based on the relationship network.

17. The apparatus of any one of claims 10-16, further comprising: a relationship network generating module configured to combine a plurality of relationship sub-networks to obtain the relationship network; determine a weight value corresponding to each of a plurality of candidate objects in the relationship network based on a number of occurrences of the candidate object in the plurality of relationship sub-networks; determine an association value between the plurality of candidate objects in the relationship network based on a number of occurrences of an association relationship between the plurality of candidate objects in the plurality of relationship sub-networks.

18. The apparatus of any one of claims 10-16, further comprising: a keyword candidate group generating module configured to generate a keyword network based on a plurality of candidate keywords subscribed by a plurality of candidate objects; determine a node value of each of the plurality of candidate keywords in the keyword network based on a number of subscriptions of the candidate keyword; determine an association value of the plurality of candidate keywords in the keyword network based on a number of occurrences of an association relationship between the plurality of candidate keywords; generate the plurality of keyword candidate groups based on the association relationship and the association value between the plurality of candidate keywords in the keyword network.

19. An electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-9.

20. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to enable the computer to perform the method of any one of claims 1-9.

21. A computer program product comprising a computer program which, when executed by a processor, implements the method of any one of claims 1-9.

Images