CN116258587A

CN116258587A - Object attribute determining method, device, equipment and readable storage medium

Info

Publication number: CN116258587A
Application number: CN202310085597.3A
Authority: CN
Inventors: 张华�; 孙科伟
Original assignee: Industrial and Commercial Bank of China Ltd ICBC; ICBC Technology Co Ltd
Current assignee: Industrial and Commercial Bank of China Ltd ICBC; ICBC Technology Co Ltd
Priority date: 2023-01-29
Filing date: 2023-01-29
Publication date: 2023-06-13

Abstract

The application provides a method, a device, equipment and a readable storage medium for determining object attributes, belongs to the technical field of data processing, and can be used in the field of financial science and technology or other related fields. The method comprises the following steps: acquiring news information related to a plurality of objects; determining a plurality of news events related to a plurality of objects according to the news information; acquiring an object correlation graph, wherein the object correlation graph is used for indicating the influence degree of a historical event on each object; determining object characteristics of a plurality of objects according to the object correlation diagram; object attributes for a plurality of objects at a future time period are determined based on the plurality of news events and the object characteristics. The method improves the accuracy of determining the object attribute.

Description

Object attribute determining method, device, equipment and readable storage medium

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to a method, an apparatus, a device, and a readable storage medium for determining an object attribute.

Background

The data of the object (such as stock in the financial field) can be analyzed by a data analysis technology to determine the object attribute (such as the rising and falling condition of the stock).

In the related art, the object attribute of the future period may be obtained by predicting according to the history data of the object (such as the rising and falling of the stock price and the history data of the stock exchange). However, in the above-described process, prediction of the object attribute is performed based on the history data of the object, resulting in lower accuracy in determining the object attribute.

Disclosure of Invention

The application provides an object attribute determining method, an object attribute determining device, object attribute determining equipment and a readable storage medium, which are used for solving the problem of low accuracy in determining object attributes.

In a first aspect, the present application provides an object attribute determining method, including:

acquiring news information related to a plurality of objects;

determining a plurality of news events related to the plurality of objects according to the news information;

acquiring an object correlation graph, wherein the object correlation graph is used for indicating the influence degree of a historical event on each object;

determining object features of the plurality of objects according to the object correlation diagram;

and determining object attributes of the plurality of objects in future time periods according to the plurality of news events and the object characteristics.

In one possible implementation, determining object attributes of the plurality of objects at a future time period based on the plurality of news events and the object features includes:

determining an attention weight vector for the plurality of news events;

acquiring a feature vector of the object feature;

and determining object attributes of the plurality of objects in a future period according to the attention weight vector and the feature vector.

In one possible implementation, determining the attention weight vector for the plurality of news events includes:

determining a marked text corresponding to each news event;

determining the attention weight of each news event according to the marked text corresponding to each news event;

the attention weight vector is determined based on the attention weight of each news event.

In one possible implementation, for any one news event; determining the attention weight of the news event according to the marked text corresponding to the news event, wherein the determining comprises the following steps:

acquiring the event type of the news event;

determining the multi-head attention mechanism corresponding to the event type;

determining the weight of the marked text corresponding to the news event according to the multi-head attention mechanism;

and determining the attention weight of the news event according to the weight of the marked text corresponding to the news event.

In one possible implementation, determining object attributes of the plurality of objects in a future period according to the attention weight vector and the feature vector includes:

normalizing the attention weight vector to obtain a normalized attention weight vector;

normalizing the feature vector to obtain a normalized feature vector;

splicing the normalized attention weight vector and the normalized feature vector to obtain a predicted feature;

and processing the prediction features through a preset model to obtain object attributes of the objects in a future period.

In one possible implementation, obtaining the object-related graph includes:

determining a plurality of historical events within a historical period and occurrence moments of the plurality of historical events;

determining an attribute change sequence of the plurality of objects in the history period, wherein the attribute change sequence comprises a plurality of attribute values and change moments of the attribute values;

determining the influence degree of each historical event on the attribute value of each object according to the plurality of historical events, the occurrence time of the plurality of historical events and the attribute change sequence of each object;

and determining the object correlation graph according to the influence degree of each historical event on the attribute value of each object.

In one possible implementation, determining object features of the plurality of objects according to the object correlation graph includes:

randomly sequencing the plurality of objects to obtain a random sequencing sequence;

traversing the object correlation graph according to the random sorting sequence and a random walk algorithm to obtain an object sequence;

and processing the object sequence through a preset neural network to obtain the object characteristics.

In a second aspect, the present application provides an object attribute determining apparatus, including a first acquisition module, a first determination module, a second acquisition module, a second determination module, and a third determination module, where,

the first acquisition module is used for acquiring news information related to a plurality of objects;

the first determining module is used for determining a plurality of news events related to the plurality of objects according to the news information;

the second acquisition module is used for acquiring an object correlation graph, wherein the object correlation graph is used for indicating the influence degree of a historical event on each object;

the second determining module is used for determining object characteristics of the plurality of objects according to the object correlation diagram;

the third determining module is configured to determine object attributes of the plurality of objects in a future period according to the plurality of news events and the object features.

In one possible implementation manner, the third determining module is specifically configured to:

determining an attention weight vector for the plurality of news events;

acquiring a feature vector of the object feature;

determining a marked text corresponding to each news event;

acquiring the event type of the news event;

determining the multi-head attention mechanism corresponding to the event type;

normalizing the feature vector to obtain a normalized feature vector;

In one possible implementation manner, the second obtaining module is specifically configured to:

In one possible implementation manner, the second determining module is specifically configured to:

In a third aspect, an embodiment of the present application provides an electronic device, including: a memory and a processor, wherein the memory is configured to store,

the memory stores computer-executable instructions;

the processor executing computer-executable instructions stored in the memory causes the processor to perform the object property determination method of any one of the first aspects.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium having stored therein computer-executable instructions for implementing the object property determination method of any one of the first aspects when the computer-executable instructions are executed by a processor.

In a fifth aspect, embodiments of the present application provide a computer program product comprising a computer program which, when executed by a processor, implements the object property determination method of any one of the first aspects.

According to the method, the device, the equipment and the readable storage medium for determining the object attribute, after news information related to a plurality of objects is acquired, a plurality of news events related to the plurality of objects can be determined according to the news information. Object characteristics of a plurality of objects can be determined according to the object correlation diagram, object attributes of the plurality of objects in future time periods are determined through a plurality of news events and the object characteristics, and accuracy of determining the object attributes is improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.

Fig. 1 is a schematic diagram of an application scenario provided in an embodiment of the present application;

fig. 2 is a flow chart of a method for determining object attributes according to an embodiment of the present application;

FIG. 3 is a flowchart illustrating another method for determining object properties according to an embodiment of the present disclosure;

FIG. 4 is a flowchart illustrating another method for determining object properties according to an embodiment of the present disclosure;

FIG. 5 is a schematic diagram of a process for determining object properties according to an embodiment of the present application;

fig. 6 is a schematic structural diagram of an object attribute determining apparatus according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Specific embodiments thereof have been shown by way of example in the drawings and will herein be described in more detail. These drawings and the written description are not intended to limit the scope of the inventive concepts in any way, but to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as detailed in the accompanying claims.

Fig. 1 is a schematic diagram of an application scenario provided in an embodiment of the present application. Referring to fig. 1, the terminal device 101, the database 102, and the processing device 103 are included.

Different news information may be stored in the terminal device 101, and the news information may refer to news information occurring within a certain period of time. Database 102 may store historical events for a plurality of objects (e.g., stocks, funds, etc.) and the time of occurrence of the historical events.

The processing device 103 may acquire news information in the terminal device 101, and process the news information to obtain a plurality of news events. The processing device 103 may obtain historical events of a plurality of objects in the database 102, determine object features based on the historical events of the plurality of objects, and process the plurality of news events and the object features to determine object attributes of the plurality of objects.

In the embodiment of the application, after acquiring the news information related to the plurality of objects, a plurality of news events related to the plurality of objects may be determined according to the news information. Object features of the plurality of objects may be determined from the object correlation map, and object attributes of the plurality of objects at a future time period may be determined from the plurality of news events and the object features.

The method shown in the present application will be described below by way of specific examples. It should be noted that the following embodiments may exist alone or in combination with each other, and for the same or similar content, the description will not be repeated in different embodiments.

Fig. 2 is a flow chart of an object attribute determining method according to an embodiment of the present application. Referring to fig. 2, the method may include:

s201, news information related to a plurality of objects is acquired.

The execution body of the embodiment of the application may be a processing device, or may be an object attribute determining apparatus provided in the processing device. The object property determining means may be implemented by software or by a combination of software and hardware.

News information related to a plurality of objects within a certain preset period can be acquired in the terminal device according to the plurality of objects. For example, assuming that a plurality of objects are a plurality of stocks, the news information is news information related to the plurality of stocks.

News information may include news information of news media stories, and related information of a plurality of objects released by users on various types of software.

For any one object, at least one piece of relevant news information of the object can be acquired.

The news information may be associated with one or more objects. For example, assuming that there are an object 1 and an object 2, the related news information of the object 1 may be news information 1, and the related news information of the object 2 may also be news information 1.

For example, assuming that there are 3 objects, namely, object 1, object 2 and object 3, 5 pieces of news information related to the object may be obtained from the news information, namely, news information 1-5, and the correspondence between the object and the news information may be as shown in table 1.

TABLE 1

Object(s)	News information
		Object 1	News information 1, news information 2, news information 3
Object 2	News information 1, news information 3
		Object 3	News information 4, news information 5

S202, determining a plurality of news events related to a plurality of objects according to the news information.

News information can be identified, and news events corresponding to each piece of news information can be determined.

For example, the determined plurality of news events may be as shown in Table 2.

TABLE 2

News information	News event
		News information 1	News event 1
News information 2	News event 2
		……	……

S203, acquiring an object correlation diagram.

The object correlation graph comprises a plurality of objects and edges between the objects, wherein the edges are used for indicating the influence degree of each historical event on each object.

For example, if the influence of the history event 1 on the stock prices of the stock 1 and the stock 2 is large, the influence of the history event 1 on the stock 1 and the stock 2 is high. If the influence of the history event 1 on the stock prices of the stock 1 and the stock 2 is small, the influence of the history event 1 on the stock 1 and the stock 2 is small.

S204, determining object characteristics of a plurality of objects according to the object correlation diagram.

Object features may be used to represent feature information of an object.

For any one object, the object features of the object may be determined from the object correlation map.

S205, determining object attributes of a plurality of objects in a future period according to the news events and the object characteristics.

News events may affect object properties of multiple objects, and object properties of multiple objects at future time periods may be determined based on the multiple news events and object characteristics.

According to the object attribute determining method, after news information related to a plurality of objects is acquired, a plurality of news events related to the plurality of objects can be determined according to the news information. An object correlation graph is obtained, the object correlation graph can be used for indicating weight relationships among objects, and object characteristics of a plurality of objects are determined according to the object correlation graph. According to the news events and the object characteristics, object attributes of the objects in future time periods are determined, and accuracy of determining the object attributes is improved.

Fig. 3 is a flowchart of another object attribute determining method according to an embodiment of the present application. Referring to fig. 3, the method may include:

s301, acquiring news information related to a plurality of objects.

S302, determining a plurality of news events related to a plurality of objects according to the news information.

The execution of S301-S302 may refer to the execution of S201-S202, and will not be described here.

S303, determining a plurality of history events in the history period and occurrence moments of the plurality of history events.

A plurality of historical events within a historical period, and the time of occurrence of the plurality of historical events, may be obtained in a database.

For example, it is assumed that 3 history events, namely, a history event 1, a history event 2, and a history event 3, may be determined in the history period, the occurrence time corresponding to the history event 1 is time 1, the occurrence time corresponding to the history event 2 is time 2, and the occurrence time corresponding to the history event 3 is time 3, which may be shown in table 3.

TABLE 3 Table 3

Historical events	Time of occurrence
		Historical event 1	Time of occurrence 1
Historical event 2	Time of occurrence 2
		Historical event 3	Time of occurrence 3

S304, determining attribute change sequences of a plurality of objects in a history period.

The attribute change sequence may include a plurality of attribute values and change times of the attribute values.

For example, the sequence of attribute changes for a plurality of objects may be as shown in Table 4.

TABLE 4 Table 4

S305, determining the influence degree of each historical event on the attribute value of each object according to the plurality of historical events, the occurrence time of the plurality of historical events and the attribute change sequence of each object.

For any one object, according to the occurrence time of a plurality of historical events and the change time of the object, determining the historical event which can influence the attribute value of the object; and determining the influence degree of the historical event on the attribute value of the object according to the historical event influencing the attribute value of the object and the attribute value of the object.

If the occurrence time of a certain historical event is earlier than or equal to the change time of the attribute values of the plurality of objects, it can be determined that the historical event can affect the attribute values of the plurality of objects.

For example, assuming that the occurrence time 1 of the history event 1 is earlier than or equal to the change time 12 of the object 1, the change time 23 of the object 2, the history event 1 may affect the attribute values of the object 1 and the object 2.

The degree of influence of the history event on the attribute values of the plurality of objects can be determined by the change condition of the attribute values of the plurality of objects.

For example, assume that there are 3 historical events, as shown in table 3, historical event 1, historical event 2, and historical event 3, respectively, and that there are 3 objects, as shown in table 5, object 1, object 2, and object 3, respectively. Assuming that the occurrence time 2 of the history event 2 is earlier than the change time 12 of the attribute value 12 of the object 1 and the change time 22 of the attribute value 22 of the object 2, and that the history event 2 has an influence on the attribute value 12 of the object 1 and the attribute value 22 of the object 2, the influence degree 21 of the history event 2 on the object 1 can be determined according to the change situation of the attribute value 12 of the object 1 and the influence degree 22 of the history event on the object 2 can be determined according to the change situation of the attribute value 22 of the object 2.

TABLE 5

Any one object may be affected by at least one historical event.

For example, assume that there are 3 historical events, as shown in table 3, historical event 1, historical event 2, and historical event 3, respectively, and that there are 3 objects, as shown in table 5, object 1, object 2, and object 3, respectively. Assuming that the occurrence time 1 of the history event 1 is earlier than the change time 21 of the object 2, the degree 12 of influence of the history event 1 on the object 2 can be determined according to the change condition of the attribute value 21 of the object 2. Assuming that the occurrence time 1 of the history event 2 is earlier than the change time 23 of the object 2, the influence degree 22 of the history event 2 on the object 2 can be determined according to the change condition of the attribute value 23 of the object 2. That is, object 2 may be affected by historical event 1 and historical event 2.

S306, determining an object correlation diagram according to the influence degree of each historical event on the attribute value of each object.

Each object may be used as a node of the object correlation graph, and the degree of influence of each history event on the attribute value of each object may be used as an edge of the object correlation graph.

For example, assuming that there are 3 objects, namely, object 1, object 2, and object 3, respectively, the degree of influence of the history event 1 on the object 1 and the object 2 is the degree of influence 11 and the degree of influence 12, respectively, and the degree of influence of the history event 2 on the object 2 and the object 3 is the degree of influence 22 and the degree of influence 23, respectively, an object correlation map including the objects 1-3, and edges between the object 1 and the object 2 are determined according to the degree of influence 11 and the degree of influence 12, and edges between the object 2 and the object 3 are determined according to the degree of influence 22 and the degree of influence 23 can be determined.

S307, randomly ordering the plurality of objects to obtain a random ordering sequence.

For example, assume that there are 3 objects, namely, object 1, object 2, and object 3, and the random order obtained by randomly ordering the 3 objects is object 2, object 3, and object 1.

And S308, traversing the object correlation diagram according to a random walk algorithm according to the random ordering sequence to obtain an object sequence.

For any one object in the random ordering sequence, the object can be used as a starting point of a random walk algorithm; traversing the object correlation diagram according to the starting point according to a random walk algorithm to obtain an object sequence value of the starting point; the sequence of objects may be determined from a plurality of object sequence values for a plurality of objects in a randomly ordered order.

For example, assuming that there are 3 objects, the random ordering order is object 2, object 3, object 1, and the object 2 is taken as a starting point, and the object correlation diagram is traversed according to a random walk algorithm to obtain an object sequence value 2 corresponding to the object 2; traversing the object correlation diagram by taking the object 3 as a starting point according to a random walk algorithm to obtain an object sequence value 3 corresponding to the object 3; traversing the object correlation diagram by taking the object 1 as a starting point according to a random walk algorithm to obtain an object sequence value 1 corresponding to the object 1; from the object sequence values 1-3 and the random order, an object sequence can be obtained, which can be shown in table 6.

TABLE 6

Object(s)	Object sequence value
		Object 2	Object sequence value 2
Object 3	Object sequence value 3
		Object 1	Object sequence value 1

S309, processing the object sequence through a preset neural network to obtain object characteristics.

The object sequence can be used as input of a preset neural network to obtain the object characteristics. The object feature may include a plurality of objects and object feature values corresponding to the plurality of objects.

Alternatively, the preset neural network may be a graph neural network.

For example, assuming that there are 3 objects, namely, object 1, object 2, and object 3, respectively, the object feature value corresponding to object 1 is object feature value 1, the object feature value corresponding to object 2 is object feature value 2, and the object feature value corresponding to object 3 is object feature value 3, the object feature can be determined, and the object feature can be shown in table 7.

TABLE 7

Object(s)	Object feature value
		Object 1	Object feature value 1
Object 2	Object feature value 2
		Object 3	Object feature value 3

S310, determining object attributes of a plurality of objects in a future period according to the news events and the object characteristics.

The execution of S310 may refer to the execution of S205, and will not be described herein.

According to the object attribute determining method provided by the embodiment of the application, the object correlation graph can be determined according to a plurality of historical events, the occurrence time of the plurality of historical events and the attribute change sequence of each object. And obtaining an object sequence according to the random ordering sequence of the plurality of objects and the object correlation diagram. And processing the object sequence through a preset neural network to obtain object characteristics. According to the news events and the object characteristics, object attributes of the objects in future time periods are obtained, and accuracy of determining the object attributes is improved.

On the basis of any one of the above embodiments, a process of determining object attributes of a plurality of objects in a future period according to a plurality of news events and object features in step S310 will be described in further detail with reference to fig. 4.

Fig. 4 is a flowchart of an object attribute determining method according to an embodiment of the present application. Referring to fig. 4, the method may include:

s401, determining a marked text corresponding to each news event.

If the text identical to the preset text can be determined in the news information corresponding to the news event, the text can be determined as the marked text corresponding to the news event.

For example, assuming that there are a preset text 1, a preset text 2, and a preset text 3, in news information corresponding to a news event 1, the same text as the preset text 1 and the preset text 3 can be determined, the preset text 1 and the preset text 2 can be determined as mark text corresponding to the news event 1.

S402, acquiring the event type of the news event.

Each news event may have its corresponding event type, which may include a company performance type, a litigation outcome type, a rumor clarification type, and so forth.

S403, determining a multi-head attention mechanism corresponding to the event type.

Each event type may have its corresponding preset multi-headed attention mechanism.

S404, determining the weight of the marked text corresponding to the news event according to the multi-head attention mechanism.

For example, assuming that the marked text corresponding to the news event 1 is marked text 1 and marked text 2, and the event type of the news event 1 is a company performance type, the weight of the marked text 1 of the news event 1 may be determined to be weight 1, and the weight of the marked text 2 may be determined to be weight 2 according to a multi-head attention mechanism of the preset company performance type.

S405, determining the attention weight of the news event according to the weight of the marked text corresponding to the news event.

The weights of the marked texts corresponding to the news event can be added to obtain the attention weight of the news event.

For example, assuming that the marked text corresponding to the news event 1 is marked text 1 and marked text 2, the weight of the marked text 1 is 3, and the weight of the marked text 2 is 1, the attention weight of the news event 1 is 4.

S406, determining attention weight vectors according to the attention weight of each news event.

For example, assuming that there are 3 news events, namely news event 1, news event 2, and news event 3, respectively, the attention weight of news event 1 is attention weight 1, the attention weight of news event 2 is attention weight 2, and the attention weight of news event 3 is attention weight 3, it may be determined that the attention weight vectors are { attention weight 1, attention weight 2, attention weight 3}.

S407, obtaining a feature vector of the object feature.

Feature vectors of object features may be obtained from the object features. For example, assuming that the object feature is as shown in table 7, the feature vector of the object feature may be determined as { object feature value 1, object feature value 2, object feature value 3}.

S408, carrying out normalization processing on the attention weight vector to obtain a normalized attention weight vector.

S409, carrying out normalization processing on the feature vector to obtain a normalized feature vector.

The attention weight vector and the feature vector select the same normalization processing method, so that the attention weight vector and the feature vector have uniform dimension.

Normalization processing methods may include linear normalization, standard deviation normalization, nonlinear normalization, and the like.

S410, splicing the normalized attention weight vector and the normalized feature vector to obtain the predicted feature.

For example, assuming that the normalized attention weight vector is { attention weight 1, attention weight 2, attention weight 3}, the normalized feature vector is { object feature value 1, object feature value 2, object feature value 3}, a prediction feature may be determined, and the prediction feature may be { attention weight 1, attention weight 2, attention weight 3, object feature value 1, object feature value 2, object feature value 3}.

S411, processing the prediction features through a preset model to obtain object attributes of a plurality of objects in a future period.

The prediction features can be used as input of a preset model to obtain object attributes of a plurality of objects. The preset model can be a multi-layer perceptron, and can be optimized through a cross entropy loss function.

If the object is a stock, the object attributes may include stock price up, stock price down, and stock price flat.

According to the object attribute determining method provided by the embodiment of the application, the attention weight vector of the news event can be determined according to the marked text corresponding to each news event and the event type of the news event. After the feature vector of the object feature is obtained, the normalized attention weight vector and the normalized feature vector can be spliced to obtain the predicted feature, and the predicted feature is processed through a preset model to obtain the object attributes of a plurality of objects. In the process, the object attribute is determined according to the attention weight vector of the news event and the feature vector of the object, so that the accuracy of determining the object attribute is improved.

Next, a detailed description will be given of an object attribute determining method shown in the embodiment of the present application by way of a specific example with reference to fig. 5.

Fig. 5 is a schematic diagram of a process for determining an object attribute according to an embodiment of the present application. Referring to fig. 5, in the actual application process, news events may be determined through news information, a preset multi-head attention mechanism is determined according to event types corresponding to the news events, and attention weight vectors are obtained according to the multi-head attention mechanism.

And determining an object correlation diagram according to the historical event information in the preset period and the attribute change sequences of the plurality of objects. After traversing the object correlation diagram through a random walk algorithm to obtain an object sequence, the object sequence can be processed through a preset neural network to obtain object characteristics.

The method comprises the steps of obtaining feature vectors of object features according to the object features, normalizing attention weight vectors and the feature vectors, then splicing to obtain predicted features, and processing the preset features through a preset multi-layer perceptron to obtain object attributes of a plurality of objects.

Fig. 6 is a schematic structural diagram of an object attribute determining apparatus according to an embodiment of the present application. Referring to fig. 6, the object property determining apparatus 10 may include a first acquiring module 11, a first determining module 12, a second acquiring module 13, a second determining module 14, and a third determining module 15, wherein,

the first acquiring module 11 is configured to acquire news information related to a plurality of objects;

the first determining module 12 is configured to determine a plurality of news events related to a plurality of objects according to news information;

the second obtaining module 13 is configured to obtain an object correlation diagram, where the object correlation diagram is used to indicate a degree of influence of a historical event on each object;

the second determining module 14 is configured to determine object features of a plurality of objects according to the object correlation map;

the third determining module 15 is configured to determine object attributes of the plurality of objects in a future period according to the plurality of news events and the object features.

In one possible implementation, the third determining module 15 is specifically configured to:

determining an attention weight vector for a plurality of news events;

acquiring a feature vector of an object feature;

object attributes of the plurality of objects at a future time period are determined based on the attention weight vector and the feature vector.

determining a marked text corresponding to each news event;

an attention weight vector is determined based on the attention weight of each news event.

acquiring event types of news events;

determining a multi-head attention mechanism corresponding to the event type;

determining the weight of a marked text corresponding to the news event according to the multi-head attention mechanism;

normalizing the feature vector to obtain a normalized feature vector;

and processing the prediction features through a preset model to obtain object attributes of a plurality of objects in a future period.

In one possible implementation, the second acquisition module 13 is specifically configured to:

determining a plurality of historical events in a historical period and occurrence moments of the plurality of historical events;

determining an attribute change sequence of a plurality of objects in a history period, wherein the attribute change sequence comprises a plurality of attribute values and change moments of the attribute values;

and determining an object correlation graph according to the influence degree of each historical event on the attribute value of each object.

In one possible implementation, the second determining module 14 is specifically configured to:

randomly sequencing a plurality of objects to obtain a random sequencing sequence;

traversing the object correlation diagram according to a random walk algorithm according to the random ordering sequence to obtain an object sequence;

and processing the object sequence through a preset neural network to obtain object characteristics.

The object attribute determining apparatus provided in the embodiment of the present application may execute the technical solution shown in the foregoing method embodiment, and its implementation principle and beneficial effects are similar, and will not be described herein again.

Fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application, referring to fig. 7, the electronic device 20 may include a processor 21 and a memory 22. The processor 21, the memory 22, and the like are illustratively interconnected by a bus 23.

Memory 22 stores computer-executable instructions;

the processor 21 executes computer-executable instructions stored in the memory 22, causing the processor 21 to perform the object property determination method as shown in the method embodiments described above.

Accordingly, embodiments of the present application provide a computer-readable storage medium having stored therein computer-executable instructions for implementing the object property determination method of the above-described method embodiments when the computer-executable instructions are executed by a processor.

Accordingly, embodiments of the present application may also provide a computer program product, including a computer program, which, when executed by a processor, may implement the object attribute determining method shown in the foregoing method embodiments.

It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should be noted that the data processing method, apparatus, device, storage medium and program product of the present disclosure may be used in the field of data processing technology, and may also be used in the field of financial technology or other related fields. The application fields of the data processing method, apparatus, device, storage medium and program product of the present disclosure are not limited.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.

Claims

1. An object attribute determining method, comprising:

acquiring news information related to a plurality of objects;

2. The method of claim 1, wherein determining object properties of the plurality of objects at a future time period based on the plurality of news events and the object features comprises:

determining an attention weight vector for the plurality of news events;

acquiring a feature vector of the object feature;

3. The method of claim 2, wherein determining the attention weight vector for the plurality of news events comprises:

determining a marked text corresponding to each news event;

4. A method according to claim 3, wherein for any one news event; determining the attention weight of the news event according to the marked text corresponding to the news event, wherein the determining comprises the following steps:

acquiring the event type of the news event;

determining a multi-head attention mechanism corresponding to the event type;

5. The method of any of claims 2-4, wherein determining object properties of the plurality of objects at a future time period from the attention weight vector and the feature vector comprises:

normalizing the feature vector to obtain a normalized feature vector;

6. The method of any of claims 1-5, wherein obtaining an object correlation map comprises:

7. The method of any of claims 1-6, wherein determining object features of the plurality of objects from the object correlation map comprises:

8. An object attribute determining device is characterized by comprising a first acquisition module, a first determination module, a second acquisition module, a second determination module and a third determination module, wherein,

9. An electronic device, comprising: a memory and a processor, wherein the memory is configured to store,

the memory stores computer-executable instructions;

the processor executing computer-executable instructions stored in the memory, causing the processor to perform the object property determination method of any one of claims 1 to 7.

10. A computer readable storage medium having stored therein computer executable instructions for implementing the object property determination method of any one of claims 1 to 7 when the computer executable instructions are executed by a processor.

11. A computer program product comprising a computer program which, when executed by a processor, implements the object property determination method of any one of claims 1 to 7.