Event atlas construction method and device, equipment, storage medium and program product

Abstract

The invention relates to the technical field of computers, and discloses an event atlas construction method and device, equipment, a storage medium and a program product, and the method comprises the steps: obtaining a currently generated first hot event and an existing second hot event; extracting event features of the first hot event from event data corresponding to the first hot event; and determining an association relationship between the first hotspot event and the second hotspot event based on the event characteristics, and constructing a hotspot event atlas by using the association relationship. By implementing the technical scheme disclosed by the invention, the automatic mining of the hot events and the construction accuracy of the hot event atlas are realized, and the real-time performance and accuracy of the hot events involved in the hot event atlas are ensured.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment, storage medium and program product for constructing an event graph. Background Art

[0002] On social multimedia platforms, a large amount of user-generated content is generated every day, and many hot events are included in these contents. Since social multimedia releases are constantly updated and hot events may be related to each other, this means that these release data need to be processed and analyzed in real time to ensure the timeliness and accuracy of hot events. However, how to extract event context and event connection points to aggregate related hot events into an event map has become an important technical challenge. Summary of the invention

[0003] In view of this, the present invention provides a method, device, equipment, storage medium and program product for constructing an event graph to solve the problem of difficulty in aggregating related hot events into an event graph.

[0004] In a first aspect, the present invention provides a method for constructing an event map, comprising: obtaining a currently generated first hot event and an existing second hot event; extracting event features of the first hot event from event data corresponding to the first hot event; determining a correlation between the first hot event and the second hot event based on the event features, and constructing a hot event map using the correlation relationship.

[0005] In a second aspect, the present invention provides a device for constructing an event graph, comprising: an acquisition module for acquiring a currently generated first hot event and an existing second hot event; a feature extraction module for extracting event features of the first hot event from event data corresponding to the first hot event; and a graph construction module for determining the correlation between the first hot event and the second hot event based on the event features, and constructing a hot event graph using the correlation relationship.

[0006] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the method for constructing an event graph of the above-mentioned first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0007] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for constructing an event graph of the first aspect or any corresponding embodiment thereof.

[0008] In a fifth aspect, the present disclosure provides a computer program product, including computer instructions, which are used to enable a computer to execute the method for constructing an event graph of the above-mentioned first aspect or any corresponding embodiment.

[0009] The event map construction method, device, equipment, storage medium and program product provided by the present disclosure extract event features from the event data of the first hot event currently generated, and use the event features to feature-associate the first hot event currently generated with the second hot event that has been generated in the historical time, so as to determine the association relationship between the first hot event and the second hot event. Thus, the hot event map is constructed using the association relationship, so that the event context and event association points can be extracted from the hot event currently generated, and the related hot events can be aggregated into a hot event map using the event context and event association points, thereby realizing the automatic discovery of hot events and the accuracy of the construction of the hot event map, and ensuring the real-time and accuracy of the hot events involved in the hot event map. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0011] Figure 1 is a schematic block diagram of constructing a hot event map according to an embodiment of the present invention;

[0012] Figure 2 is a flow chart of a method for constructing an event graph according to an embodiment of the present invention;

[0013] Figure 3 is a flow chart of another method for constructing an event graph according to an embodiment of the present invention;

[0014] Figure 4 is a schematic diagram of event clustering according to an embodiment of the present invention;

[0015] Figure 5 is a flow chart of another method for constructing an event graph according to an embodiment of the present invention;

[0016] Figure 6 is a detailed flow chart of a method for constructing an event graph according to an embodiment of the present invention;

[0017] Figure 7 is a structural block diagram of a device for constructing an event graph according to an embodiment of the present invention;

[0018] Figure 8 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0021] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0022] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0023] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0024] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0025] The content published on social multimedia platforms includes many hot events, which may be related to each other, and users' posts on social multimedia are constantly updated, which means that these hot data need to be processed and analyzed in real time to ensure the timeliness and accuracy of hot events.

[0026] At present, the discovery and mining of hot events are mainly carried out by constructing event graphs. However, constructing event graphs is a complex process involving multiple technical links, including data collection, content understanding, and relationship association. After collecting hot data, natural language processing is used to preprocess the text of the hot data to remove noise, stop words, punctuation, etc., and then entity recognition and event information extraction are performed on the preprocessed hot data. The identified entities are matched with the entities in the existing knowledge base, and the relationships between entities are identified to construct the associations between events to form an event graph. However, constructing an event graph through entity relationships cannot guarantee the accuracy of the construction of the event graph, and it is difficult for entity relationships to accurately determine the associations between events.

[0027] like Figure 1 As shown, the technical solution disclosed in the present invention is to extract event features of the currently collected hot events, and to perform reference disambiguation on the extracted event features to obtain accurate event features. The event features are used to associate the currently generated hot events with historical events (historical hot events or historical hot candidate events) that have occurred in the historical time, so as to identify the correlation between the hot events and determine whether the currently generated hot event is an extension or context of the historical event. If the currently generated hot event is not related to the historical event, the event density is clustered for the currently generated hot event and other events to generate a candidate hot event, and a hot event map of the candidate hot event is constructed to determine whether it is likely to become a hot event based on the heat of the event.

[0028] According to an embodiment of the present invention, an embodiment of a method for constructing an event graph is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0029] In this embodiment, a method for constructing an event graph is provided, which can be used in computer devices, such as mobile phones, tablet computers, and computers. Figure 2 is a flow chart of a method for constructing an event graph according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0030] Step S101, obtaining the currently generated first hot event and the existing second hot event.

[0031] The first hot event is the latest hot event obtained in real time from various authorized data sources. Specifically, the first hot event can be a major online event currently occurring in real time, and the data source can be an online major event database such as the homepage of a news website. The latest hot event highlighted on the homepage of the news website will be stored in the online major event database, and the computer device can access the content in the online major event database to extract the first hot event; the data source can also be an online hot database of social media, and there will be corresponding hot tags for the hot events currently occurring. The computer device can access the online hot database and obtain the current first hot event from it according to the hot tags.

[0032] Specifically, the first hot event may also be an event that has been discovered to be likely to become a hot topic, and the data source may be a press release database (for example, a database of press releases by official media and local media; a press release database by Internet users on the media, etc.). The computer device may access the press release database and discover the first hot event currently occurring from the press release database.

[0033] The second hot event is an event that has occurred in the historical time. The second hot event can be a historical event that has been determined as a hot spot, or a historical event that may become a hot spot. Specifically, the existing second hot events will be stored in the hot event database, and the computer device can access the hot event database to extract the second hot events that occurred in the historical time.

[0034] Step S102: extracting event features of the first hot event from event data corresponding to the first hot event.

[0035] Event data is used to characterize the occurrence data of hot events, which includes detailed description data of the occurrence of hot events, hot consumption data, strongly associated multimedia data, etc.

[0036] The event feature is used to characterize the event attributes of the first hot event, such as the event location, event name, event time, etc.

[0037] The event data corresponding to the first hot event is analyzed, key data such as the time, location, and event participants of the first hot event are identified from the event data, and event characteristics of the first hot event are determined based on the key data.

[0038] Step S103, determining the correlation between the first hot event and the second hot event based on the event characteristics, and constructing a hot event map using the correlation.

[0039] The hot event map is an event network structure generated based on the association relationship between hot events. The first hot event and the second hot event are matched with event features to determine whether the first hot event and the second hot event are associated. If the first hot event and the second hot event are associated, the association type between the first hot event and the second hot event is determined, and an event association link between the first hot event and the second hot event is constructed according to the association type, and the hot event map is constructed according to the event association link. If the first hot event and the second hot event are not associated, it can be determined that the first hot event is independent of the second hot event. At this time, an event node is created for the first hot event, and the hot event map is constructed according to the created event node. Thus, when a subsequent event is associated with the first hot event, the hot event map can be updated according to the association relationship between the subsequent event and the first hot event.

[0040] The method for constructing an event graph provided in this embodiment extracts event features from the event data of the first hot event currently generated, and uses the event features to associate the first hot event currently generated with the second hot event that has been generated in the historical time, so as to determine the association relationship between the first hot event and the second hot event. Thus, by using the association relationship to construct a hot event graph, it is possible to extract event contexts and event association points from the hot event currently generated, and aggregate related hot events into a hot event graph using the event contexts and event association points, thereby realizing the automatic discovery of hot events and ensuring the real-time and accuracy of the hot events involved in the hot event graph.

[0041] In this embodiment, a method for constructing an event graph is provided, which can be used in computer devices, such as mobile phones, tablet computers, and computers. Figure 3 is a flow chart of a method for constructing an event graph according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0042] Step S201, obtain the currently generated first hot event and the existing second hot event. Figure 2 Step S101 of the illustrated embodiment will not be described in detail here.

[0043] Step S202: extracting event features of the first hot event from event data corresponding to the first hot event. Figure 2 Step S102 of the illustrated embodiment will not be described in detail here.

[0044] Step S203, determining the correlation between the first hot event and the second hot event based on the event characteristics, and constructing a hot event map using the correlation.

[0045] Specifically, the above step S203 includes:

[0046] Step S2031 , content association is performed on the first hot event and the second hot event based on the event features to generate a content association result.

[0047] The content of the first hot event and the second hot event are matched according to the event characteristics to determine the content similarity between the first hot event and the second hot event. The content similarity is used to determine the relevance of the first hot event and the second hot event in terms of content, and the content relevance result of the first hot event and the second hot event is obtained.

[0048] Step S2032: determine the association relationship between the first hot event and the second hot event according to the content association result.

[0049] The correlation relationship is used to characterize whether there is a correlation between the first hot event and the second hot event. The first hot event and the second hot event are analyzed in terms of background content, event content, etc. in combination with the content correlation result to determine the correlation relationship between the first hot event and the second hot event in terms of background content and event content.

[0050] In some optional implementations, the above step S2032 includes:

[0051] Step a1: determine whether the first hot event is an extension of the second hot event based on the content association result.

[0052] Step a2: if the first hot event is an extension of the second hot event, then determine that the first hot event and the second hot event are associated with each other through extension.

[0053] Determine whether the content association between the first hot event and the second hot event is an association of event content in combination with the content association result. If it is determined that the content association between the first hot event and the second hot event is an association of event content, it can be determined that the event content represented by the first hot event is a continuation or extension of the event content represented by the second hot event, that is, the first hot event is an extension of the second hot event. At this time, it can be determined that the association relationship between the first hot event and the second hot event is an extended association, that is, the first hot event is an extended event of the second hot event.

[0054] By detecting whether the first hot event is a continuation or extension of the second hot event, the extended correlation relationship between the first hot event and the second hot event can be determined, so that the extended correlation relationship can be used to construct a hot event map, so as to organize related hot events in the same hot event map, thereby ensuring the accuracy of the construction of the hot event map.

[0055] In some optional implementations, the above step S2032 includes:

[0056] Step b1, determining whether there is a background association between the first hot event and the second hot event based on the content association result.

[0057] Step b2: If there is a background association between the first hot event and the second hot event, it is determined that there is a context association between the first hot event and the second hot event.

[0058] Combined with the content association result, determine whether the content association between the first hot event and the second hot event is an association of background content. If it is determined that the content association between the first hot event and the second hot event is an association of background content, it can be determined that the first hot event and the second hot event belong to the same series of events, that is, the first hot event is the contextual development of the second hot event. At this time, it can be determined that the association relationship between the first hot event and the second hot event is a contextual association, that is, the first hot event is the contextual event of the second hot event.

[0059] By detecting whether the first hot event and the second hot event only have background association, the context association relationship between the first hot event and the second hot event can be determined, so as to extract the context from the related hot events and organize the hot events with context association in the same hot event map, thereby realizing the construction of the hot event map using the context association relationship, and further ensuring the accuracy of the construction of the hot event map.

[0060] In some optional implementations, the above step S2032 includes:

[0061] Step c1: if it is determined based on the content association result that the first hot event and the second hot event are not associated, the first hot event is determined as a first hot candidate event.

[0062] Step c2, obtaining the existing second hotspot candidate events.

[0063] Step c3: clustering the first hotspot candidate event and the second hotspot candidate event to generate a target candidate event.

[0064] The first hotspot candidate event indicates a newly generated event that has the possibility of becoming a hotspot, that is, the first hotspot event has the possibility of becoming a hotspot. If it is determined that the first hotspot event and the second hotspot event are not related in terms of background content, event content, etc. in combination with the content association result, it can be determined that the first hotspot event and the second hotspot event are not related. At this time, the first hotspot event is determined as the first hotspot candidate event.

[0065] The second hotspot candidate events are events that may become hotspots in the historical time. Specifically, the existing second hotspot candidate events are stored in the hotspot event database, and the computer device can access the hotspot event database to extract the second hotspot candidate events that occurred in the historical time.

[0066] The newly generated first hot candidate event is density clustered together with the second hot candidate event already in the hot event database, and the pre-trained large language model and clustering algorithm are used to identify related events to obtain various related event clusters. Based on the event clustering results, an event name is generated for each related event cluster to represent the main event of the related event cluster, and the main event of the related event cluster is determined as the target candidate event.

[0067] like Figure 4 As shown, by aggregating the currently generated hot candidate events with the existing hot candidate events, so as to aggregate the events with related hot spots and generate corresponding target candidate events, it is possible to extract the core content from the aggregated target candidate events, and use the extracted core content to construct a hot event map, so as to facilitate real-time mining of hot events.

[0068] In some optional implementations, the above method further includes:

[0069] Step d1, obtaining the heat index of the target candidate event.

[0070] Step d2, determining the event heat of the target candidate event based on the heat index.

[0071] Step d3: Update the target candidate event to a hot event according to the event heat.

[0072] The heat index is an indicator used to measure the heat of an event, which may include event exposure, social media interaction (such as likes, comments, shares), and event search volume, etc. Specifically, the heat data of the target candidate event within a specific time range is collected, the heat data is analyzed, and the heat index of the target candidate event is extracted from the heat data.

[0073] The heat calculation is performed according to each heat index of the target candidate event to obtain the event heat corresponding to the target candidate event. Specifically, the event heat can be obtained by weighting and summing each heat index according to its importance; it can also be obtained by setting a high weight for the index with higher importance and using an exponential weighted summation method; it can also be obtained by adding the ranking of each heat index and obtaining the total ranking as the event heat.

[0074] The calculated event heat is analyzed to determine whether the event heat of the target candidate event exceeds the preset heat value (the pre-set minimum heat value that can be used as a hot event). If the event heat of the target candidate event exceeds the preset heat value, the target candidate event is determined as a hot event and the target candidate event is updated to the hot event database.

[0075] As time and user interactions change, event popularity will change dynamically. Therefore, the hot events in the hot event database can be continuously updated based on the event popularity to ensure the comprehensiveness, accuracy and timeliness of the hot events in the hot event database.

[0076] Step S2033, constructing a hot event map based on the association type of the association relationship.

[0077] The association type indicates the type of association relationship, such as the extended association, context association, etc. described above. The first hot event and the second hot event are used as event nodes, and an event relationship chain is constructed between the first hot event and the second hot event in combination with the association type of the association relationship. The event nodes are connected according to the event relationship chain to generate a hot event map.

[0078] In some optional implementations, the above step S2033 includes:

[0079] Step e1, classifying the first hot event according to the association type, and determining the event type of the first hot event.

[0080] Step e2: determining the event node of the first hot event based on the event type.

[0081] Step e3, construct a hot event map according to event nodes.

[0082] The event type of the first hot event can be determined in combination with the association type, and the event type includes extended events, context events, and hot candidate events. Thus, according to the association type between the first hot event and the second hot event, the first hot event is classified into extended events, context events, and hot candidate events to obtain the event type of the first hot event.

[0083] The event node is used to characterize the association relationship between the first hot event and the second hot event and the newly generated target candidate event. Specifically, if there is an association between the first hot event and the second hot event, an event node is generated according to the event type of the first hot event, a relationship chain between the first hot event and the second hot event is constructed based on the event node, and the first hot event is updated to the hot event map according to the relationship chain. If there is no association between the first hot event and the second hot event, an independent event node is generated according to the event type of the first hot event, and a hot event map for the first hot event is constructed with the first hot event as the center.

[0084] By classifying the first hot event, event nodes are generated using the event type of the first hot event, and a hot event map is constructed according to each event node. Thus, similar but non-repetitive hot events can be divided according to event types, thereby realizing automatic division of hot events, and then facilitating accurate construction of a hot event map according to the division results of hot events.

[0085] The event graph construction method provided in this embodiment realizes automatic discovery of hot events and ensures the real-time nature of hot events by using event features to associate content with different hot events, so as to determine the association relationship between hot events based on the content association results. In combination with the association type of the association relationship, a relationship chain between hot events is constructed to ensure the accuracy of the relationship chain in the hot event graph. By connecting each hot event according to the relationship chain to form a hot event graph, the comprehensiveness, accuracy and timeliness of the hot data in the hot event graph are ensured.

[0086] In this embodiment, a method for constructing an event graph is provided, which can be used in computer devices, such as mobile phones, tablet computers, and computers. Figure 5 is a flow chart of a method for constructing an event graph according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:

[0087] Step S301, obtain the currently generated first hot event and the existing second hot event. Figure 3 Step S201 of the illustrated embodiment will not be described in detail here.

[0088] Step S302: extracting event features of the first hot event from event data corresponding to the first hot event.

[0089] The event characteristics include at least one of the event subject, location, time and keywords. Specifically, the above step S302 includes:

[0090] Step S3021, identify the subject of the event data and determine the event subject corresponding to the first hot event.

[0091] The event subject is the name of the participating object of the first hot event, such as a person's name, an organization name, etc. As described above, the event data includes detailed description data for the hot event, hot consumption data, strongly associated multimedia data, etc. The event data is converted into a text representation to obtain the text information corresponding to the event data. Then, the object entities contained in the text information are identified to determine one or more object entities contained in the text information. By analyzing each object entity, the participating object entity of the first hot event, that is, the event subject of the first hot event, is determined.

[0092] In a specific example, text conversion and entity recognition of event data can be achieved through a pre-trained large language model. The event data of the first hot event is input into the large language model, the event data is converted into text information through the large language model, and object entity recognition is performed on the text information to determine the event subject of the first hot event.

[0093] Step S3022: perform location identification on the event data to determine the location where the first hot event occurs.

[0094] The event data is converted into text representation to obtain text information corresponding to the event data. Then, the location entity is identified in the text information to determine the location entity contained in the text information. By analyzing each location entity, the location where the first hot event occurred is determined.

[0095] In a specific example, text conversion and entity recognition of event data can be achieved through a pre-trained large language model. The event data of the first hot event is input into the large language model, the event data is converted into text information through the large language model, and the location entity recognition is performed on the text information to determine the occurrence location of the first hot event.

[0096] Step S3023, perform time identification on the event data to determine the time when the first hot event occurs.

[0097] The event data is converted into text representation to obtain text information corresponding to the event data. The time information is extracted from the text information by using named entity recognition (NER) or time expression recognition to determine the time when the first hot event occurred.

[0098] Step S3024: perform semantic analysis on the event data to determine multiple keywords corresponding to the first hot event.

[0099] The event data is converted into text representation to obtain the text information corresponding to the event data. The text information is preprocessed to remove irrelevant information such as stop words, and then semantic analysis is performed on the preprocessed text information. Multiple keywords related to the first hot event are extracted from the text information by combining word frequency statistics, TF-IDF and other methods, and a keyword cloud is generated based on the frequency of the keywords.

[0100] In some optional implementations, the above method further includes: if the event features include ambiguous features, performing reference disambiguation processing on the ambiguous features.

[0101] An ambiguous feature is a feature that refers to multiple entities in an event feature. In order to ensure that each event feature can point to a certain entity, it is necessary to perform reference disambiguation on the ambiguous feature. Specifically, reference disambiguation includes pronoun disambiguation and noun phrase disambiguation, among which pronoun disambiguation is used to disambiguate the object entity referred to by the pronoun, and noun phrase disambiguation is used to disambiguate the object entity referred to by the noun phrase.

[0102] In a specific example, a machine learning model, such as a hidden Markov model (HMM), a conditional random field (CRF), or a neural network, can be used to learn reference relationships to perform reference disambiguation on ambiguous features in event features.

[0103] In another specific example, predefined rules can be used to parse the object entity referred to by the ambiguous feature to achieve the purpose of reference disambiguation. Of course, reference disambiguation can also be achieved by combining with an external knowledge base or common sense reasoning, and the specific method of reference disambiguation is not limited here.

[0104] By performing reference disambiguation on ambiguous features in event features, the accuracy of the reference of event features is ensured, thereby ensuring the accuracy of the association relationship.

[0105] Step S303: Determine the correlation between the first hot event and the second hot event based on the event characteristics, and construct a hot event map using the correlation. Figure 3 Step S203 of the illustrated embodiment will not be described in detail here.

[0106] The event graph construction method provided in this embodiment extracts event features corresponding to the first hot event from multiple dimensions, ensuring that the event features can accurately represent the event content of the first hot event, thereby improving the accuracy of determining the association relationship.

[0107] As a specific application example of this application, Figure 6As shown in the figure, when a new hot word event is added to the social media platform, the existing historical hot events such as the text of the historical hot word event, the hot word associated with the hot word, and the strongly associated video of the hot word are recalled from the hot word event database. Then, the new hot word event is associated with the historical hot word event to determine whether the new hot word event is associated with the historical hot word event. If there is an association, the event features are extracted from the new hot word event, the association type between the new hot word event and the historical hot word event is analyzed, and the event node of the event corresponding to the new hot word event is created according to the association type, and the hot event map is updated using the event node. If there is no association, other hot candidate events are obtained from the hot pool, the new hot word event is clustered with other hot candidate events, and the hot event is supplemented and recalled according to the clustering results to determine whether the new hot word event can become a new hot event. If the new hot word event can become a new hot event, a hot event map is constructed based on the newly generated hot event.

[0108] In this embodiment, a device for constructing an event graph is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated here. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0109] This embodiment provides a device for constructing an event graph, such as Figure 7 As shown, including:

[0110] The acquisition module 401 is used to acquire the currently generated first hot event and the existing second hot event.

[0111] The feature extraction module 402 is used to extract the event feature of the first hot event from the event data corresponding to the first hot event.

[0112] The graph construction module 403 is used to determine the correlation between the first hot event and the second hot event based on event features, and construct a hot event graph using the correlation.

[0113] In some optional implementations, the graph construction module 403 includes:

[0114] The content association unit is used to perform content association on the first hot event and the second hot event based on the event features to generate a content association result.

[0115] The association relationship determination unit is used to determine the association relationship between the first hot event and the second hot event according to the content association result.

[0116] The construction unit is used to construct a hot event map based on the association type of the association relationship.

[0117] In some optional implementations, the association relationship determination unit includes:

[0118] The extension judgment subunit is used to determine whether the first hot event is an extension of the second hot event based on the content association result.

[0119] The extended association determination subunit is used to determine that the first hot event and the second hot event are associated with each other if the first hot event is an extension of the second hot event.

[0120] In some optional implementations, the association relationship determination unit includes:

[0121] The background association judgment subunit is used to determine whether there is background association between the first hot event and the second hot event based on the content association result.

[0122] The context association determination subunit is used to determine that the first hot event and the second hot event are contextually associated if there is a background association between the first hot event and the second hot event.

[0123] In some optional implementations, the association relationship determination unit includes:

[0124] The candidate event determination subunit is used to determine the first hot event as a first hot event candidate event if it is determined based on the content association result that the first hot event and the second hot event are not associated.

[0125] The candidate event acquisition subunit is used to acquire the existing second hotspot candidate events.

[0126] The clustering subunit is used to cluster the first hotspot candidate event and the second hotspot candidate event to generate a target candidate event.

[0127] In some optional implementations, the above-mentioned graph construction module 403 further includes:

[0128] The index acquisition unit is used to obtain the heat index of the target candidate event.

[0129] The heat determination unit is used to determine the event heat of the target candidate event based on the heat index.

[0130] The updating unit is used to update the target candidate event to a hot event according to the heat of the event.

[0131] In some optional embodiments, the building block includes:

[0132] The classification subunit is used to classify the first hot event according to the association type and determine the event type of the first hot event.

[0133] The event node determination subunit is used to determine the event node of the first hot event based on the event type.

[0134] The construction subunit is used to construct a hot event map according to event nodes.

[0135] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0136] The event graph construction device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0137] The embodiment of the present invention also provides a computer device having the above Figure 7 The device for constructing the event graph shown.

[0138] See also Figure 8 , Figure 8 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 8 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 10 is taken as an example.

[0139] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0140] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0141] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0142] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0143] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 8 The example of connecting through bus is taken in the following.

[0144] The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display and a plasma display. In some optional embodiments, the display device can be a touch screen.

[0145] The computer device also includes a communication interface, which is used for the computer device to communicate with other devices or a communication network.

[0146] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0147] A part of the present disclosure may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present disclosure through the operation of the computer. Those skilled in the art should understand that the existence of computer program instructions in computer-readable media includes, but is not limited to, source files, executable files, installation package files, etc., and accordingly, the way in which computer program instructions are executed by a computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0148] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for constructing an event graph, characterized in that: The method comprises: Obtain the currently generated first hot event and the existing second hot event; Extracting event features of the first hot event from event data corresponding to the first hot event; Based on the event characteristics, the correlation relationship between the first hot event and the second hot event is determined, and a hot event map is constructed using the correlation relationship.

2. The method according to claim 1, characterized in that The determining the association relationship between the first hot event and the second hot event based on the event characteristics, and constructing a hot event map using the association relationship, includes: Based on the event features, content association is performed on the first hot event and the second hot event to generate a content association result; Determine the association relationship between the first hot event and the second hot event according to the content association result; Based on the association type of the association relationship, the hot event map is constructed.

3. The method according to claim 2, characterized in that The determining the association relationship between the first hot event and the second hot event according to the content association result includes: Determining whether the first hot event is an extension of the second hot event based on the content association result; If the first hot event is an extension of the second hot event, it is determined that the first hot event and the second hot event are associated with each other through extension.

4. The method according to claim 2, characterized in that: The determining the association relationship between the first hot event and the second hot event according to the content association result includes: Determine whether the first hot event and the second hot event have background association based on the content association result; If there is a background association between the first hot event and the second hot event, it is determined that there is a context association between the first hot event and the second hot event.

5. The method according to claim 2, characterized in that: The determining the association relationship between the first hot event and the second hot event according to the content association result includes: If it is determined based on the content association result that the first hot event and the second hot event are not associated, determining the first hot event as a first hot candidate event; Obtain the existing second hotspot candidate events; The first hotspot candidate event and the second hotspot candidate event are clustered to generate a target candidate event.

6. The method according to claim 5, characterized in that Also includes: Obtaining a heat index of the target candidate event; Determining the event heat of the target candidate event based on the heat index; The target candidate event is updated as a hot event according to the event heat.

7. The method according to any one of claims 2 to 6, characterized in that: The step of constructing the hot event map based on the association type of the association relationship includes: Classifying the first hot event according to the association type, and determining the event type of the first hot event; Determine an event node of the first hot event based on the event type; The hot event map is constructed according to the event nodes.

8. The method according to claim 1, characterized in that The event feature includes at least one of an event subject, a location, a time, and a keyword; and extracting the event feature of the first hot event from the event data corresponding to the first hot event includes: Performing subject identification on the event data to determine the event subject corresponding to the first hot event; and / or, Performing location identification on the event data to determine the location where the first hot event occurs; and / or, Performing time identification on the event data to determine the time when the first hot event occurred; and / or, Perform semantic analysis on the event data to determine multiple keywords corresponding to the first hot event.

9. The method according to claim 8, characterized in that Also includes: If the event features include ambiguous features, reference disambiguation processing is performed on the ambiguous features.

10. A device for constructing an event graph, characterized in that: The device comprises: An acquisition module, used to acquire the currently generated first hot event and the existing second hot event; A feature extraction module, used to extract event features of the first hot event from event data corresponding to the first hot event; A graph construction module is used to determine the correlation between the first hot event and the second hot event based on the event characteristics, and to construct a hot event graph using the correlation.

11. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the event graph construction method described in any one of claims 1 to 9 by executing the computer instructions.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for constructing an event graph according to any one of claims 1 to 9.

13. A computer program product, characterized in that It comprises computer instructions, and the computer instructions are used to make a computer execute the method for constructing an event graph according to any one of claims 1 to 9.

Images