CN110162637B - Information map construction method, device and equipment - Google Patents
Information map construction method, device and equipment Download PDFInfo
- Publication number
- CN110162637B CN110162637B CN201910114989.1A CN201910114989A CN110162637B CN 110162637 B CN110162637 B CN 110162637B CN 201910114989 A CN201910114989 A CN 201910114989A CN 110162637 B CN110162637 B CN 110162637B
- Authority
- CN
- China
- Prior art keywords
- information
- node
- sub
- map
- graph
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses an information map construction method, an information map construction device and information map construction equipment, wherein the information map construction method comprises the following steps: extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and stitching the extracted map elements to generate an information map corresponding to the input information; wherein at least a part of sub-spectrums in the multi-class sub-spectrums have the same information node, and extracting the spectrum element corresponding to the input information from the multi-class sub-spectrums according to the input information comprises: determining a node attribute value of a target information node according to the input information; and extracting a map element corresponding to the determined node attribute value from each class of sub-maps. By extracting and splicing the map elements corresponding to the input information in the multiple types of sub-maps, the sub-maps of multiple different types can be effectively associated based on the input information, and the information of multiple types associated with the input information can be returned.
Description
Technical Field
The present disclosure relates to the field of map construction, and more particularly, to an information map construction method, an information map construction apparatus, and an information map construction device.
Background
With the wide application of artificial intelligence in civil and commercial fields, the construction of the map plays an increasingly important role in the processes of business big datamation, intellectualization and the like, so that the construction of the map, in particular the construction of the information map, is also faced with higher requirements. At present, the construction of the information spectrum is concentrated on the construction of a general information spectrum in a user searching recommendation service and a knowledge base obtained based on manual arrangement, such as the knowledge of Microsoft's people cube and hundred degrees, which is successfully used in the scenes of intelligent question-answering, machine translation and the like.
However, the information patterns obtained by current construction are single-category information patterns, the information patterns are independent, and unified association is not formed among the information patterns of a plurality of categories. When specific query information is input, the returned result is mostly single or single category information, and all information related to the specific query information cannot be obtained.
Therefore, there is a need for an information map construction method capable of returning a plurality of categories of information associated with input information based on the input information efficiently when information map construction is realized based on the input information, particularly when construction of a financial information map is realized.
Disclosure of Invention
In view of the above, the present disclosure provides an information map construction method, an information map construction apparatus, and an information map construction device. By utilizing the information map construction method provided by the invention, on the basis of realizing information map construction based on the input information, a plurality of sub-maps of different categories can be effectively associated based on the input information, and a plurality of categories of information associated with the input information can be returned.
According to an aspect of the present disclosure, an information map construction method is provided, including: extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and stitching the extracted map elements to generate an information map corresponding to the input information; wherein at least a part of sub-spectrums in the multi-class sub-spectrums have the same information node, and extracting the spectrum element corresponding to the input information from the multi-class sub-spectrums according to the input information comprises: determining a node attribute value of a target information node according to the input information; and extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
In some embodiments, each class of sub-spectrum has at least two information nodes, and each class of sub-spectrum includes a plurality of map elements; for each information node, each map element has a node attribute value for that information node.
In some embodiments, stitching the extracted map elements to generate an information map corresponding to the input information includes: and taking each determined node attribute value as a link point, and linking map elements extracted from the multi-class sub-maps and corresponding to the node attribute values.
In some embodiments, the multiple classes of sub-patterns include at least two classes of concept sub-patterns, company sub-patterns, stock sub-patterns, investment sub-patterns, product sub-patterns, event sub-patterns, industry sub-patterns.
In some embodiments, the target information node comprises at least one of: company name, unit name, person name, industry name, product name, stock code, stock name, concept name.
In some embodiments, the multiple classes of sub-spectrums include at least event sub-spectrums, and the input information is event information, wherein determining a node attribute value of the target information node from the input information includes: for the input event information, determining a node attribute value of a target information node corresponding to the event information based on the event sub-map.
In some embodiments, before extracting the spectrum element corresponding to the input information from the multi-class sub-spectrum according to the input information, constructing the multi-class sub-spectrum is further included, where constructing each class of sub-spectrum in the multi-class sub-spectrum includes: setting at least two information nodes and setting the corresponding relation between the information nodes; extracting node attribute values of the information nodes from external data; correlating the extracted node attribute values to form a map element; denoising the obtained map elements to obtain a sub-map.
In some embodiments, the at least two information nodes comprise company name information nodes, and denoising the resulting map elements comprises: for each node attribute value of the company name information node, a plurality of node attribute values conforming to the company simple full scale corresponding relation are associated based on the pre-established company simple full scale corresponding relation.
In some embodiments, the at least two inodes comprise a name inode and denoising the resulting map element comprises: dividing a plurality of attribute values of company name information nodes associated with the same person name information node, and identifying the node attribute value of the person name information node by utilizing the node attribute value of each company name information node, thereby eliminating ambiguity of the person name information node.
According to another aspect of the present disclosure, there is provided an information map construction apparatus including: a map element extraction module configured to extract a map element corresponding to the input information from the multi-class sub-graph according to the input information; an information map generation module configured to splice the extracted map elements to generate an information map corresponding to the input information.
In some embodiments, each class of sub-spectrum has at least two information nodes, and each class of sub-spectrum includes a plurality of map elements; for each information node, each map element has a node attribute value for that information node; wherein, the atlas element extraction module includes: the target information node determining module is configured to determine a node attribute value of the target information node according to the input information; and a map element correspondence module configured to extract, from each class of sub-maps, a map element corresponding to the determined node attribute value.
In some embodiments, the information profile generation module comprises: and the map element linking module is configured to link the map elements extracted from the multi-class sub-maps and corresponding to the node attribute values by taking each determined node attribute value as a linking point.
According to another aspect of the present disclosure, there is provided an information graph construction apparatus, wherein the apparatus comprises a processor and a memory containing a set of instructions which, when executed by the processor, cause the information graph construction apparatus to perform operations comprising: extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and stitching the extracted map elements to generate an information map corresponding to the input information. Wherein at least a part of sub-spectrums in the multi-class sub-spectrums have the same information node, and extracting the spectrum element corresponding to the input information from the multi-class sub-spectrums according to the input information comprises: determining a node attribute value of a target information node according to the input information; and extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
In some embodiments, each class of sub-spectrum has at least two information nodes, and each class of sub-spectrum includes a plurality of map elements; for each information node, each map element has a node attribute value for that information node. By utilizing the information map construction method provided by the invention, on the basis of realizing information map construction based on the input information, a plurality of sub-maps of different categories can be effectively associated based on the input information, and a plurality of categories of information associated with the input information can be returned.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings may be obtained according to these drawings without making creative efforts to one of ordinary skill in the art. The following drawings are not intended to be drawn to scale on actual dimensions, emphasis instead being placed upon illustrating the principles of the disclosure.
FIG. 1 illustrates an exemplary flowchart of an information graph construction method 100 according to an embodiment of the present disclosure;
FIG. 2 illustrates a partial schematic diagram of a corporate sub-map 200 in accordance with an embodiment of the present disclosure;
FIG. 3 illustrates a flowchart of an exemplary method 300 of extracting a atlas element corresponding to input information from a multi-class sub-spectrum according to input information in accordance with an embodiment of the present disclosure;
FIG. 4 illustrates a schematic diagram of extracting a map element 400 corresponding to input information from a multi-class sub-spectrum according to the input information, according to an embodiment of the present disclosure;
FIG. 5 is a schematic diagram of linking, with each determined node attribute value as a link point, a graph element 500 extracted from the multi-class sub-graph that corresponds to the node attribute value;
FIG. 6 illustrates a schematic diagram of determining node attribute values 600 for a target information node corresponding to event information based on event sub-graphs, according to an embodiment of the present disclosure;
FIG. 7 illustrates an exemplary flowchart of a method 700 of constructing a sub-map according to an embodiment of the present disclosure;
FIG. 8 illustrates an exemplary flow chart of a method 800 of corporate name disambiguation in an embodiment of the present disclosure;
FIG. 9 illustrates an exemplary flow chart of a method 900 of name disambiguation in an embodiment of the present disclosure;
FIG. 10 illustrates an exemplary block diagram of an information graph construction apparatus 110 according to an embodiment of the present disclosure;
fig. 11 shows an exemplary block diagram of an information graph construction apparatus 950 according to an embodiment of the disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this application and in the claims, the terms "a," "an," "the," and/or "the" are not specific to the singular, but may include the plural, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
Flowcharts are used in this application to describe the operations performed by systems according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Fig. 1 shows an exemplary flowchart of an information graph construction method 100 according to an embodiment of the present disclosure.
First, in step S101, a map element corresponding to input information is extracted from a multi-class sub-map according to the input information. After extracting a plurality of map elements corresponding to the input information, further, in step S102, the extracted map elements are spliced to generate an information map corresponding to the input information.
The input information may be query information directly input by the user, or may be information that the computer system queries itself in response to user input information or control information. The embodiment of the disclosure is not limited by the source of the input information and the input mode thereof. For example, the query information input by the user in the web search bar may be the query information of the user, or the information generated by preprocessing the query information of the user by the computer may be the query information of the user.
The multi-class sub-graph refers to a plurality of sub-graphs respectively belonging to different classes. Wherein at least a portion of the sub-patterns in the multiple classes of sub-patterns have the same information node. In some embodiments, the multiple classes of sub-patterns may include at least two classes of concept sub-patterns, company sub-patterns, stock sub-patterns, investment sub-patterns, product sub-patterns, event sub-patterns, industry sub-patterns.
According to embodiments of the present disclosure, each class of sub-spectrum may have, for example, at least two information nodes, and each class of sub-spectrum includes a plurality of map elements, each map element having a node attribute value for each information node.
In some embodiments, in each map element, for the same information node, it may have one or more node attribute values for that information node. The node attribute value may be content specific or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. Taking the map elements of the corporate sub-map as examples, the map elements can comprise node attribute values of 'hundred degrees' and node attribute values of 'Aiqi' and 'Aiqi' are sub-companies of 'hundred degrees', and based on the corresponding relation, the 'hundred degrees' are primary node attribute values and the 'Aiqi' is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
Fig. 2 shows a partial schematic diagram of a corporate sub-map 200 in accordance with an embodiment of the present disclosure.
The above-described information nodes and map elements are described in more detail below with reference to the corporate sub-map shown in fig. 2. The company sub-graph comprises three information nodes, namely a company name information node K C Name information node K H And industry name information node K S Further, the corporate sub-map includes, for example, map element E C1 Map element E C2 Wherein the map element E C1 Further comprising node attribute values for a plurality of information nodes. With further reference to FIG. 2, profile element E C1 The method comprises the steps of including node attribute values of company name information nodes such as "Aiqi skill", "Baidu", node attribute values of name information nodes such as "Li Yanhong", "Liu Ji", and node attribute values of industry name information nodes such as "Internet". Map element E C2 The method comprises the steps of "Dajiang" node attribute value of company name information node, "Wang Tao" node attribute value of person name information node and "unmanned aerial vehicle" node attribute value of industry name information node. Further, based on the correspondence between the node attribute values, under the company name information node, "hundred degrees" are primary node attribute values, and "Aiqi skill" is a secondary node attribute value.
Fig. 3 illustrates a flowchart of an exemplary method 300 of extracting a atlas element corresponding to input information from a multi-class sub-spectrum according to an embodiment of the disclosure.
First, in step S301, a node attribute value of a target information node is determined from input information.
According to embodiments of the present disclosure, the target information node may be, for example, a plurality of information nodes from a plurality of classes of sub-maps, or may be a plurality of information nodes from a certain class of sub-maps, or may also be one information node in a certain class of maps. Embodiments of the present disclosure are not limited by the type and number of target information nodes determined. For example, the target information node may be set to only the company name, or the target information node may be set to three of the company name, the person name, and the product service name.
In some embodiments, the target information node comprises at least one of: company name, unit name, person name, industry name, product name, stock code, stock name, concept name.
According to the embodiment of the disclosure, the node attribute value of the target information node may be determined based on a preset policy, for example, when all or part of the content of the input information is the node attribute value of the target information node, the attribute value of the target information node may be directly obtained; when the content in the input information does not include the node attribute value of the target information node or includes only a part of the node attribute value of the target information node, the input information may be preprocessed, and the processed data may be used as the node attribute value of the target information node. Embodiments of the present disclosure are not limited by the manner in which the node attribute values of the target nodes are determined. For example, setting a person name as a target information node, and taking 'Ma Yun' input by a user as a node attribute value of the target information node directly; the method can also develop a sharing bicycle market for hundred-degree hand-in-hand ali input by a user, extract and obtain node attribute values of company name information nodes of hundred degrees and ali based on processes of entity identification, relation extraction and the like, obtain the node attribute values of product name information nodes of the sharing bicycle, and further obtain node attribute values of name information nodes corresponding to the node attribute values of Li Yanhong and Ma Yun in a company sub-map based on the hundred degrees and ali.
Furthermore, for the same target information node, the node attribute value of the target information node determined based on the input information may be one or more, and the embodiments of the present disclosure are not limited by the number of node attribute values of the determined target information node.
After determining the node attribute value of the target information node according to the input information, further, in step S302, a map element corresponding to the determined node attribute value is extracted from each class of sub-maps.
In some embodiments, the map elements extracted from each class of sub-maps corresponding to the determined node attribute values may be plural, for example, when there is a name of a person, for the node attribute value "Ma Yun" of the target information node, there may be plural map elements corresponding thereto in the company sub-map, which respectively belong to node attribute values having different company names. In some embodiments, in some class sub-spectrums, a spectrum element corresponding to the node attribute value of the target information node cannot be extracted, for example, when the node attribute value of the target information node is "fast hand App", the spectrum element corresponding to the node attribute value of the target information node may not be included in the stock sub-spectrum. Embodiments of the present disclosure are not limited by the number of extracted map elements and their sources.
Fig. 4 shows a schematic diagram of extracting a map element corresponding to input information from a multi-class sub-spectrum according to the input information according to an embodiment of the present disclosure.
The above process may be described in more detail with reference to fig. 4. For example, if the preset target information node is a person name and a company name, when the input information of the user is Liu Ji, searching the map element corresponding to Liu Ji through the company sub-map to obtain a map element E C1 The method comprises the steps of carrying out a first treatment on the surface of the Further, the map element E can be obtained C1 Company name information node K corresponding to node attribute value of Liu Ji C Is a node attribute value of (a). According to the embodiment of the disclosure, the company name information node K can be used for C As its node attribute value, the node attribute value "hundred" of the company name information node corresponding to the input information "Liu Ji" can be obtained.
After obtaining the node attribute value "Liu Ji" of the name information node and the node attribute value "hundred degrees" of the company name information node, further, extracting a map element corresponding to the determined node attribute value from each sub-map of the constructed multi-class sub-map. For example, when there are a company sub-graph, a product sub-graph and a stock sub-graph currently, the company sub-graph element E corresponding to the node attribute value "hundred degrees" of the company name information node and the node attribute value "Liu Ji" of the name information node can be extracted C1 Sub-graph element E of product P1 Stock subgraph element E G1 。
After extracting the plurality of map elements, in step S102, stitching the extracted map elements to generate an information map corresponding to the input information may include: and taking each determined node attribute value as a link point, and linking map elements extracted from the multi-class sub-maps and corresponding to the node attribute values.
Fig. 5 shows a schematic diagram of linking, with each determined node attribute value as a link point, a map element extracted from the multi-class sub-map and corresponding to the node attribute value.
The graph stitching process may be described in more detail with reference to fig. 4 and 5. Based on the process shown in fig. 4, after extracting a plurality of map elements, further, as shown in fig. 5, the company sub-map elements E may be spliced with the node attribute value "hundred" of the company name information node and the node attribute value "Liu Ji" of the person name information node as link points C1 Sub-graph element E of product P1 Stock subgraph element E G1 An information map having company name information nodes, person name information nodes, stock information nodes, product information nodes, industry name information nodes is generated.
By extracting node attribute values of the target information node based on the input information, map elements in sub-maps of a plurality of different categories are effectively associated, so that an information map of the target information node having a plurality of categories can be generated, and thus information of a plurality of categories associated with the input information can be returned.
In some embodiments, the multiple classes of sub-spectrums include at least event sub-spectrums, and the input information is event information, wherein determining a node attribute value of the target information node from the input information includes: for the input event information, determining a node attribute value of a target information node corresponding to the event information based on the event sub-map.
The input information is, for example, a specific event name or a description of a certain event, which may, for example, include only short names of related events, or may also include person, company, topic or concept names related to the event. The present disclosure is not limited by the specific content of the input event information. It may input, for example, "Wei Zexi event" or "Ma Huateng hand-in-hand Yang Zhenyu to initiate a scientific exploration prize".
For example, the event sub-map may have at least two types of information nodes therein, which may be, for example, entity class information nodes and topic class information nodes. The entity class information node includes, for example, an event name, a company name, a person name, a stock, etc., and the topic class information node includes, for example, a topic name, a concept name, a semantic tag, etc.
According to an embodiment of the present disclosure, an event sub-graph includes a plurality of graph elements, each graph element having a node attribute value for each information node.
In some embodiments, in a map element of an event sub-map, it may have one or more node attribute values for that information node for the same information node. The node attribute value may be specific content, or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements of the event sub-map.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. For example, the map elements of an event sub-map may include a node attribute value "mobile communication" and a node attribute value "5G communication" corresponding to a topic name information node, where "5G communication" is a sub-topic concept of "mobile communication", and based on the correspondence, then "mobile communication" is a primary node attribute value, and "5G communication" is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
Fig. 6 shows a schematic diagram of determining a node attribute value of a target information node corresponding to the event information based on the event sub-map.
Determination of the event sub-map based on the event sub-map and the same will be specifically described with reference to fig. 6And a process of the target information node corresponding to the event information. For the event information of the 'hundred-degree bar Wei Zexi event' input by the user, taking the event information as a target information node, searching a map element corresponding to the event 'hundred-degree bar Wei Zexi event' in an event sub-map to obtain a map element E V1 The method comprises the steps of carrying out a first treatment on the surface of the Further, the related event 'hundred degree end and field hospital cooperation' of the event can be obtained through the event sub-graph, and the related company is 'hundred degree online' and the related subject is 'medical dispute'. Thus, node attribute values of company name information nodes and topic information nodes related to the event are determined based on the event sub-map.
After determining the node attribute value of the company name information node and the node attribute value of the subject information node related to the input event, the determined node attribute value of the company name information node and/or the determined node attribute value of the subject information node may be continued to be used as the node attribute value of the target information node, and the corresponding icon element may be further searched in the company sub-map.
In some embodiments, the process of constructing the multi-class sub-spectrum is further included before extracting the spectrum elements corresponding to the input information from the multi-class sub-spectrum according to the input information.
Fig. 7 illustrates an exemplary flowchart of a method 700 of constructing a sub-map according to an embodiment of the present disclosure.
According to an embodiment of the present disclosure, constructing each class of sub-spectrum in the multi-class sub-spectrum includes: first, in step S701, at least two information nodes are set, and a preset correspondence relationship between the information nodes is set. Further, in step S702, a node attribute value of the information node is extracted from the external data. After the extraction of the node attribute values is completed, in step S703, the extracted node attribute values are correlated to form a map element. Further, in step S704, the obtained map elements are denoised, and a sub-map is obtained.
According to embodiments of the present disclosure, in a map element, for the same information node, it may have one or more node attribute values for that information node. The node attribute value may be content specific or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. Taking the map elements of the company sub-map as examples, the map elements can comprise node attribute values of 'hundred degrees' and node attribute values of 'Aiqi' which are sub-companies of 'hundred degrees', and based on the corresponding relation, the 'hundred degrees' are primary node attribute values, and the 'Aiqi' is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
According to the embodiment of the disclosure, denoising of the obtained map elements can be realized by calculating the node importance, filtering the noise node attribute value by calculating the importance, saving the storage space of the map and accelerating the query speed; on the other hand, when a plurality of candidate nodes are hit simultaneously when the query pattern is blurred, the node with high importance is preferentially selected; or can be realized by calculating the confidence coefficient of the node, for example, the confidence coefficient of the node is calculated by adopting the source times of the node and the confidence coefficient of each source; or the relation confidence coefficient can be calculated, namely the relation source number, the source confidence coefficient, the node confidence coefficient at two sides of the relation, and the node-out and node-in co-occurrence times heuristic type relation confidence coefficient is adopted. The present disclosure is not limited by the pattern element denoising approach.
Hereinafter, a process of constructing a sub-map will be specifically described by taking a construction company sub-map as an example. First, when constructing the corporate subgraph, as shown in step S701, for example, a corporate name, a person name, and an industry may be selected as preset information nodes, and further preset relationships thereof may be set as follows: for the attribute value of each company information node, the corresponding attribute value of the name information node is the individual supplied to the company, and the corresponding attribute value of the industry information node is the industry in which the company is located. Further, for each different attribute value in the company name information node, it may also be set that it includes a affiliation of the main company with the subsidiary company.
Further, in step S702, a node attribute value of the information node is extracted from the external data. For example, node attribute values for information nodes such as person names, company names, etc. may be extracted from unstructured text through a named entity recognition process.
Further, in step S703, the relationships between the extracted node attribute values may be obtained based on a relationship extraction process, which may be obtained through structured form data on a website, for example, or may be classified according to a predefined relationship class based on entity identification. Further, the node attribute values satisfying the correspondence among the extracted plurality of node attribute values are associated to form a map element. For example, the node attribute value "hundred degrees" of the information node of the company is correlated with the node attribute value "Li Yanhong" corresponding to the name information node of the individual person who supplies the company, and the node attribute value "internet" corresponding to the industry information node of the industry where the company is located, so as to obtain a map element.
The step of associating the node attribute values may be implemented by connecting the node attribute values through line segments, or may be placed in a corresponding form, which is not limited by the form in which the node attribute values are associated with each other. Further, when the association of the node attribute values is implemented by using line segments, the correlation between the node attribute values may also be noted on the line segments.
After generating a plurality of map elements, in step S704, the obtained map elements are denoised, and a sub-map is obtained. The denoising process may, for example, first calculate node importance, node confidence, relationship confidence, and then weight average the three to screen out noisy unstable data.
In some embodiments, the map element denoising further includes node attribute value disambiguation for the information node, which further may include company name disambiguation and person name disambiguation.
For company name disambiguation, a plurality of node attribute values conforming to a company profile correspondence are associated based on a pre-established company profile correspondence by for each node attribute value of a company name information node.
Fig. 8 illustrates an exemplary flow chart of a method 800 of corporate name disambiguation in an embodiment of the present disclosure.
Referring to fig. 8, first, in step S801, for each node attribute value in the company information node in the sub-map, the company name and the corresponding plurality of company abbreviations thereof are extracted from the external data. Further, in step S802, based on the obtained corporate full names and corresponding multiple corporate abbreviations, a full name and corresponding relationship of each corporate is obtained, and a corporate Jian Quan name sub-map is generated; finally, in step S803, a plurality of node attribute values conforming to the full scale and short term correspondence among them are associated based on the short full scale correspondence obtained in the sub-map of the company Jian Quanchen.
In some embodiments, after the sub-map of company Jian Quan is obtained in step S802, instead of implementing association of company names in full abbreviation in step S803, the full names of all companies and corresponding multiple abbreviations in the sub-map of company Jian Quan are input into the input end of the neural network algorithm, and the relationship model of the full names of the companies is generated through the neural network algorithm. Then, by inputting the names of the designated companies at the input end of the company simple-name relationship model, a plurality of corresponding short names can be obtained at the output end of the company simple-name relationship model, and the simple-name relationship of the designated companies can be obtained accordingly. Further, based on the simple full scale relationship of the designated company, a plurality of node attribute values conforming to the corresponding relationship of the full scale and the short scale can be associated.
For name disambiguation, name node ambiguity can be resolved by partitioning a plurality of attribute values of company name information nodes associated with the same name information node, and identifying the node attribute value of the name information node with the node attribute value of each company name information node.
Fig. 9 illustrates an exemplary flow chart of a method of name disambiguation of an embodiment of the present disclosure.
Referring to fig. 9, first, in step S901, for a node attribute value of a person name information node in a map element, the map element is divided with the node attribute value as a center point, to obtain a plurality of map sub-elements independent from each other. Further, in step S902, for each sub-element in the plurality of independent map sub-elements, a plurality of company name information nodes directly associated with the node attribute value of the person name information node and the node attribute values thereof are extracted. Finally, in step S903, for each company name information node, the node attribute value of the person name information node is identified using the node attribute value of the company information node.
When the attribute value of the name information node is taken as a center point to divide the map elements, for example, the attribute value of the name information node in the current map elements can be removed, so that a plurality of independent map sub-elements can be obtained.
In some embodiments, extracting the plurality of company name information nodes and the node attribute values thereof directly associated with the node attribute values of the person name information node from each of the plurality of independent map sub-elements may include: when a plurality of node attribute values of company name information nodes directly related to the name information node attribute value exist in one map subelement, the relation confidence coefficient of the node attribute values of the plurality of company name information nodes and the node attribute values of the name information nodes can be calculated; and extracts the node attribute value of the company name information node having the greatest relation confidence therein.
By the method for identifying the name by correspondingly associating the name with the company name, false association during sub-graph association or sub-graph elements which introduce excessive noise and redundancy due to the fact that the person is renamed when the name information node is used as the target information node can be avoided.
Fig. 10 shows an exemplary block diagram of an information map construction apparatus 110 according to an embodiment of the present disclosure.
The information map construction apparatus 110 shown in fig. 10 includes: a profile element extraction module 120 and an information profile generation module 130.
Wherein the map element extraction module 120 is configured to extract a map element corresponding to the input information from the multi-class sub-spectrum according to the input information.
The information map generation module 130 is configured to splice the extracted map elements to generate an information map corresponding to the input information.
The input information may be query information directly input by the user, or may be information that the computer system queries itself in response to user input information or control information. The embodiment of the disclosure is not limited by the source of the input information and the input mode thereof. For example, the query information input by the user in the web search bar may be the query information of the user, or the information generated by preprocessing the query information of the user by the computer may be the query information of the user.
The multi-class sub-graph refers to a plurality of sub-graphs respectively belonging to different classes. Wherein at least a portion of the sub-patterns in the multiple classes of sub-patterns have the same information node. In some embodiments, the multiple classes of sub-patterns may include at least two classes of concept sub-patterns, company sub-patterns, stock sub-patterns, investment sub-patterns, product sub-patterns, event sub-patterns, industry sub-patterns.
In some embodiments, in a map element, for the same information node, it may have one or more node attribute values for that information node. The node attribute value may be content specific or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. Taking the map elements of the company sub-map as examples, the map elements can comprise node attribute values of 'hundred degrees' and node attribute values of 'Aiqi' which are sub-companies of 'hundred degrees', and based on the corresponding relation, the 'hundred degrees' are primary node attribute values, and the 'Aiqi' is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
Fig. 2 shows a partial schematic diagram of a corporate sub-map 200 in accordance with an embodiment of the present disclosure.
The above-described information nodes and map elements are described in more detail below with reference to the corporate sub-map shown in fig. 2. The company sub-graph comprises three information nodes, namely a company name information node K C Name information node K H And industry name information node K S . Further, the corporate sub-map includes, for example, map element E C1 Map element E C2 Wherein the map element E C1 Further comprising node attribute values for a plurality of information nodes. With further reference to FIG. 2, profile element E C1 The method comprises the steps of including node attribute values of company name information nodes such as "Aiqi skill", "Baidu", node attribute values of name information nodes such as "Li Yanhong", "Liu Ji", and node attribute values of industry name information nodes such as "Internet". Map element E C2 The method comprises the steps of "Dajiang" node attribute value of company name information node, "Wang Tao" node attribute value of person name information node and "unmanned aerial vehicle" node attribute value of industry name information node. Further, based on the correspondence between the node attribute values, the node attribute value "hundred degrees" is a primary node attribute value, and the node attribute value "aiqi skill" is a secondary node attribute value.
In the map element extraction module 120, a flow as shown in fig. 3 may be performed, where a map element corresponding to input information is extracted from multiple types of sub-spectrums according to the input information. It may further comprise: a target information node determination module 121 and a map element correspondence module 122.
The target information node determining module 121 is configured to perform the operation shown in step S301 in fig. 3, and determine the node attribute value of the target information node according to the input information.
According to embodiments of the present disclosure, the target information node may be, for example, a plurality of information nodes from a plurality of classes of sub-maps, or may be a plurality of information nodes from a certain class of sub-maps, or may also be one information node in a certain class of maps. Embodiments of the present disclosure are not limited by the type and number of target information nodes determined. For example, the target information node may be set to only the company name, or the target information node may be set to three of the company name, the person name, and the product service name.
In some embodiments, the target information node comprises at least one of: company name, unit name, person name, industry name, product name, stock code, stock name, concept name.
The node attribute value of the target information node can be determined based on a preset strategy, for example, when all or part of the content of the input information is the node attribute value of the target information node, the attribute value of the target information node can be directly obtained; when the content in the input information does not include the node attribute value of the target information node or includes only a part of the node attribute value of the target information node, the input information may be preprocessed, and the processed data may be used as the node attribute value of the target information node. Embodiments of the present disclosure are not limited by the manner in which the node attribute values of the target nodes are determined. For example, setting a person name as a target information node, and taking 'Ma Yun' input by a user as a node attribute value of the target information node directly; the method can also develop a sharing bicycle market for hundred-degree hand-in-hand ali input by a user, extract and obtain node attribute values of company name information nodes of hundred degrees and ali based on processes of entity identification, relation extraction and the like, obtain the node attribute values of product name information nodes of the sharing bicycle, and further obtain node attribute values of name information nodes corresponding to the node attribute values of Li Yanhong and Ma Yun in a company sub-map based on the hundred degrees and ali.
Furthermore, for the same target information node, the node attribute value of the target information node determined based on the input information may be one or more, and the embodiments of the present disclosure are not limited by the number of node attribute values of the determined target information node.
The map element correspondence module 122 is configured to perform the operation shown as step S302 in fig. 3, extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
In some embodiments, the map elements extracted from each class of sub-maps corresponding to the determined node attribute values may be plural, for example, when there is a name of a person, for the node attribute value "Ma Yun" of the target information node, there may be plural map elements corresponding thereto in the company sub-map, which respectively belong to node attribute values having different company names. In some embodiments, in some class sub-spectrums, a spectrum element corresponding to the node attribute value of the target information node cannot be extracted, for example, when the node attribute value of the target information node is "fast hand App", the spectrum element corresponding to the node attribute value of the target information node may not be included in the stock sub-spectrum. Embodiments of the present disclosure are not limited by the number of extracted map elements and their sources.
In some embodiments, further, the information map generating module 130 may include a map element linking module 131, in which a process as shown in fig. 5 may be performed in the map element linking module 131, and each determined node attribute value is used as a linking point to link a map element extracted from the multi-class sub-map and corresponding to the node attribute value.
The graph stitching process may be described in more detail with reference to fig. 4 and 5. Based on the process shown in fig. 4, after extracting a plurality of map elements, further, as shown in fig. 5, the company sub-map elements E may be spliced with the node attribute value "hundred" of the company name information node and the node attribute value "Liu Ji" of the person name information node as link points C1 Sub-graph element E of product P1 Stock subgraph element E G1 An information map having company name information nodes, person name information nodes, stock information nodes, and product name information nodes is generated.
In some embodiments, the multiple classes of sub-spectra include at least event sub-spectra, and the input information is event information. The target information node determining module 121 may include an event information correspondence module 121', and in the event information correspondence module 121', for the input event information, a node attribute value of a target information node corresponding to the event information may be determined based on the event sub-map.
The input information is, for example, a specific event name or a description of a certain event, which may, for example, include only short names of related events, or may also include person, company, topic or concept names related to the event. The present disclosure is not limited by the specific content of the input event information. It may input, for example, "Wei Zexi event" or "Ma Huateng hand-in-hand Yang Zhenyu to initiate a scientific exploration prize".
For example, the event sub-map may have at least two types of information nodes therein, which may be, for example, entity class information nodes and topic class information nodes. The entity class information node includes, for example, an event name, a company name, a person name, a stock, etc., and the topic class information node includes, for example, a topic name, a concept name, a semantic tag, etc.
According to an embodiment of the present disclosure, an event sub-graph includes a plurality of graph elements, each graph element having a node attribute value for each information node.
In some embodiments, in a map element of an event sub-map, it may have one or more node attribute values for that information node for the same information node. The node attribute value may be specific content, or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements of the event sub-map.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. For example, the map elements of an event sub-map may include a node attribute value "mobile communication" and a node attribute value "5G communication" corresponding to a topic name information node, where "5G communication" is a sub-topic concept of "mobile communication", and based on the correspondence, then "mobile communication" is a primary node attribute value, and "5G communication" is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
Fig. 11 shows an exemplary block diagram of an information graph construction apparatus 950 according to an embodiment of the disclosure.
The information-graph construction device 950 shown in fig. 11 may be implemented as one or more special-purpose or general-purpose computer system modules or components, such as a personal computer, notebook computer, tablet computer, cell phone, personal digital assistant (personal digital assistance, PDA), and any smart portable device. The information profile construction device 950 may include at least one processor 960 and a memory 970, among other things.
Wherein the at least one processor is configured to execute program instructions. The memory 970 may be present in the information-graph construction device 950 in different forms of program storage units as well as data storage units, such as hard disk, read-only memory (ROM), random Access Memory (RAM), which can be used to store various data files used by the processor in processing and/or executing the information-graph construction process, as well as possible program instructions executed by the processor. Although not shown in the figures, the information-graph construction device 950 may also include an input/output component that supports input/output data flow between the information-graph construction device 950 and other components (e.g., the image-capturing device 980). The information profile construction device 950 may also send and receive information and data from a network through a communication port.
In some embodiments, the set of instructions stored by the memory 970, when executed by the processor 960, causes the information-graph construction device 950 to perform operations comprising: extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and stitching the extracted map elements to generate an information map corresponding to the input information.
The input information may be query information directly input by the user, or may be information that the computer system queries itself in response to user input information or control information. The embodiment of the disclosure is not limited by the source of the input information and the input mode thereof. For example, the query information input by the user in the web search bar may be the query information of the user, or the information generated by preprocessing the query information of the user by the computer may be the query information of the user.
The multi-class sub-graph refers to a plurality of sub-graphs respectively belonging to different classes. Wherein at least a portion of the sub-patterns in the multiple classes of sub-patterns have the same information node. In some embodiments, the multiple classes of sub-patterns may include at least two classes of concept sub-patterns, company sub-patterns, stock sub-patterns, investment sub-patterns, product sub-patterns, event sub-patterns, industry sub-patterns.
According to embodiments of the present disclosure, each class of sub-spectrum may have, for example, at least two information nodes, and each class of sub-spectrum includes a plurality of map elements, each map element having a node attribute value for each information node.
In some embodiments, in a map element, for the same information node, it may have one or more node attribute values for that information node. The node attribute value may be content specific or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. Taking the map elements of the company sub-map as examples, the map elements can comprise node attribute values of 'hundred degrees' and node attribute values of 'Aiqi' which are sub-companies of 'hundred degrees', and based on the corresponding relation, the 'hundred degrees' are primary node attribute values, and the 'Aiqi' is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
In some embodiments, extracting, from the multi-class sub-spectrum, a spectrum element corresponding to the input information according to the input information includes: determining a node attribute value of a target information node according to the input information; and extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
According to embodiments of the present disclosure, the target information node may be, for example, a plurality of information nodes from a plurality of classes of sub-maps, or may be a plurality of information nodes from a certain class of sub-maps, or may also be one information node in a certain class of maps. Embodiments of the present disclosure are not limited by the type and number of target information nodes determined. For example, the target information node may be set to only the company name, or the target information node may be set to three of the company name, the person name, and the product service name.
In some embodiments, the target information node comprises at least one of: company name, unit name, person name, industry name, product name, stock code, stock name, concept name.
Furthermore, for the same target information node, the node attribute value of the target information node determined based on the input information may be one or more, and the embodiments of the present disclosure are not limited by the number of node attribute values of the determined target information node.
In some embodiments, the map elements extracted from each class of sub-maps corresponding to the determined node attribute values may be plural, for example, when there is a name of a person, for the node attribute value "Ma Yun" of the target information node, there may be plural map elements corresponding thereto in the company sub-map, which respectively belong to node attribute values having different company names. In some embodiments, in some class sub-spectrums, a spectrum element corresponding to the node attribute value of the target information node cannot be extracted, for example, when the node attribute value of the target information node is "fast hand App", the spectrum element corresponding to the node attribute value of the target information node may not be included in the stock sub-spectrum. Embodiments of the present disclosure are not limited by the number of extracted map elements and their sources.
In some embodiments, the multiple classes of sub-spectrums include at least event sub-spectrums, and the input information is event information, wherein determining a node attribute value of the target information node from the input information includes: for the input event information, determining a node attribute value of a target information node corresponding to the event information based on the event sub-map.
The input information is, for example, a specific event name or a description of a certain event, which may, for example, include only short names of related events, or may also include person, company, topic or concept names related to the event. The present disclosure is not limited by the specific content of the input event information. It may input, for example, "Wei Zexi event" or "Ma Huateng hand-in-hand Yang Zhenyu to initiate a scientific exploration prize".
For example, the event sub-map may have at least two types of information nodes therein, which may be, for example, entity class information nodes and topic class information nodes. The entity class information node includes, for example, an event name, a company name, a person name, a stock, etc., and the topic class information node includes, for example, a topic name, a concept name, a semantic tag, etc.
According to an embodiment of the present disclosure, an event sub-graph includes a plurality of graph elements, each graph element having a node attribute value for each information node.
In some embodiments, in a map element of an event sub-map, it may have one or more node attribute values for that information node for the same information node. The node attribute value may be specific content, or may be "default" or "null". The method is not limited by the number and specific content of the node attribute values of the information nodes in the map elements of the event sub-map.
In some embodiments, the node attribute values may be further classified, for example, into primary node attribute values and secondary node attribute values. For example, the map elements of an event sub-map may include a node attribute value "mobile communication" and a node attribute value "5G communication" corresponding to a topic name information node, where "5G communication" is a sub-topic concept of "mobile communication", and based on the correspondence, then "mobile communication" is a primary node attribute value, and "5G communication" is a secondary node attribute value. The method is not limited by different levels of node attribute values in the map elements and the corresponding relation between the levels.
In some embodiments, the information-map construction apparatus 950 may receive input information from an input device external to the information-map construction apparatus 950, perform the information-map construction method described above, implement the functions of the information-map construction device described above, based on the input information.
In some embodiments, the process of constructing the multi-class sub-spectrum is further included before extracting the spectrum elements corresponding to the input information from the multi-class sub-spectrum according to the input information.
Although in fig. 10, the processor 960 and the memory 970 are presented as separate modules, it will be appreciated by those skilled in the art that the above-described device modules may be implemented as separate hardware devices or may be integrated as one or more hardware devices. The specific implementation of the different hardware devices should not be taken as a factor limiting the scope of protection of the present disclosure, as long as the principles described in this disclosure can be implemented.
This application uses specific words to describe embodiments of the application. Reference to "a first/second embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic is associated with at least one embodiment of the present application. Thus, it should be emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various positions in this specification are not necessarily referring to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of the present application may be combined as suitable.
Furthermore, those skilled in the art will appreciate that the various aspects of the invention are illustrated and described in the context of a number of patentable categories or circumstances, including any novel and useful procedures, machines, products, or materials, or any novel and useful modifications thereof. Accordingly, aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.) or by a combination of hardware and software. The above hardware or software may be referred to as a "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of the present application may take the form of a computer product, comprising computer-readable program code, embodied in one or more computer-readable media.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The foregoing is illustrative of the present invention and is not to be construed as limiting thereof. Although a few exemplary embodiments of this invention have been described, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the novel teachings and advantages of this invention. Accordingly, all such modifications are intended to be included within the scope of this invention as defined in the following claims. It is to be understood that the foregoing is illustrative of the present invention and is not to be construed as limited to the specific embodiments disclosed, and that modifications to the disclosed embodiments, as well as other embodiments, are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.
Claims (14)
1. An information map construction method, comprising:
extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and
stitching the extracted map elements to generate an information map corresponding to the input information;
wherein at least a part of sub-spectrums in the multi-class sub-spectrums have the same information node, and extracting the spectrum element corresponding to the input information from the multi-class sub-spectrums according to the input information comprises:
Determining a node attribute value of a target information node according to the input information; and
and extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
2. The information graph construction method of claim 1, wherein each class of sub-graph has at least two information nodes, and each class of sub-graph includes a plurality of graph elements; for each information node, each map element has a node attribute value for that information node.
3. The information map construction method of claim 2, wherein stitching the extracted map elements to generate the information map corresponding to the input information comprises:
and taking each determined node attribute value as a link point, and linking map elements extracted from the multi-class sub-maps and corresponding to the node attribute values.
4. The information graph construction method of claim 1, wherein the multi-class sub-graph includes at least two classes of concept sub-graph, company sub-graph, stock sub-graph, investment sub-graph, product sub-graph, event sub-graph, industry sub-graph.
5. The information graph construction method of claim 2, wherein the target information node includes at least one of: company name, unit name, person name, industry name, product name, stock code, stock name, concept name.
6. The information-graph construction method of claim 2, wherein the multi-class sub-graph includes at least an event sub-graph, and the input information is event information,
wherein determining the node attribute value of the target information node according to the input information comprises:
for the input event information, determining a node attribute value of a target information node corresponding to the event information based on the event sub-map.
7. The information graph construction method of claim 1, wherein before extracting the graph element corresponding to the input information from the multiple types of sub-graphs according to the input information, constructing the multiple types of sub-graphs, wherein constructing each type of sub-graph in the multiple types of sub-graphs includes:
setting at least two information nodes and setting the corresponding relation between the information nodes;
extracting node attribute values of the information nodes from external data;
correlating the extracted node attribute values to form a map element;
denoising the obtained map elements to obtain a sub-map.
8. The graph construction method of claim 7, wherein the at least two information nodes include company name information nodes, and denoising the resulting graph elements includes:
For each node attribute value of the company name information node, a plurality of node attribute values conforming to the company simple full scale corresponding relation are associated based on the pre-established company simple full scale corresponding relation.
9. The graph construction method of claim 7, wherein the at least two information nodes include a person name information node, and denoising the resulting graph element includes:
dividing a plurality of attribute values of company name information nodes associated with the same person name information node, and identifying the node attribute value of the person name information node by utilizing the node attribute value of each company name information node, thereby eliminating ambiguity of the person name information node.
10. An information map construction apparatus comprising:
a map element extraction module configured to extract a map element corresponding to the input information from the multi-class sub-graph according to the input information;
an information map generation module configured to splice the extracted map elements to generate an information map corresponding to the input information.
11. The information graph construction apparatus of claim 10, wherein each class of sub-graph has at least two information nodes, and each class of sub-graph includes a plurality of graph elements; for each information node, each map element has a node attribute value for that information node;
Wherein, the atlas element extraction module includes:
the target information node determining module is configured to determine a node attribute value of the target information node according to the input information; and
and a map element correspondence module configured to extract, from each class of sub-maps, a map element corresponding to the determined node attribute value.
12. The information map construction apparatus according to claim 10, wherein the information map generation module includes:
and the map element linking module is configured to link the map elements extracted from the multi-class sub-maps and corresponding to the node attribute values by taking each determined node attribute value as a linking point.
13. An information-graph construction apparatus, wherein the apparatus comprises a processor and a memory containing a set of instructions that, when executed by the processor, cause the information-graph construction apparatus to perform operations comprising:
extracting map elements corresponding to the input information from the multi-class sub-map according to the input information; and
stitching the extracted map elements to generate an information map corresponding to the input information;
wherein at least a part of sub-spectrums in the multi-class sub-spectrums have the same information node, and extracting the spectrum element corresponding to the input information from the multi-class sub-spectrums according to the input information comprises:
Determining a node attribute value of a target information node according to the input information; and
and extracting a map element corresponding to the determined node attribute value from each class of sub-maps.
14. The information-graph construction apparatus of claim 13, wherein each class of sub-graph has at least two information nodes, and each class of sub-graph includes a plurality of graph elements; for each information node, each map element has a node attribute value for that information node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910114989.1A CN110162637B (en) | 2019-02-14 | 2019-02-14 | Information map construction method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910114989.1A CN110162637B (en) | 2019-02-14 | 2019-02-14 | Information map construction method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110162637A CN110162637A (en) | 2019-08-23 |
| CN110162637B true CN110162637B (en) | 2023-06-20 |
Family
ID=67644878
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910114989.1A Active CN110162637B (en) | 2019-02-14 | 2019-02-14 | Information map construction method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110162637B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110781317B (en) * | 2019-10-29 | 2022-03-01 | 北京明略软件系统有限公司 | Method and device for constructing event map and electronic equipment |
| CN111930906A (en) * | 2020-07-29 | 2020-11-13 | 北京北大软件工程股份有限公司 | Knowledge graph question-answering method and device based on semantic block |
| CN112612899A (en) * | 2020-11-24 | 2021-04-06 | 中国传媒大学 | Knowledge graph construction method and device, storage medium and electronic equipment |
| CN113708980B (en) * | 2021-10-27 | 2022-04-19 | 中国光大银行股份有限公司 | Topological graph generation method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9026524B1 (en) * | 2013-01-10 | 2015-05-05 | Relationship Science LLC | Completing queries using transitive closures on a social graph |
| CN106777331A (en) * | 2017-01-11 | 2017-05-31 | 北京航空航天大学 | Knowledge mapping generation method and device |
| CN108509420A (en) * | 2018-03-29 | 2018-09-07 | 赵维平 | Gu spectrum and ancient culture knowledge mapping natural language processing method |
| CN108763555A (en) * | 2018-06-01 | 2018-11-06 | 北京奇虎科技有限公司 | Representation data acquisition methods and device based on demand word |
| CN109033223A (en) * | 2018-06-29 | 2018-12-18 | 北京百度网讯科技有限公司 | For method, apparatus, equipment and computer readable storage medium across type session |
| CN109145122A (en) * | 2018-08-02 | 2019-01-04 | 北京仿真中心 | A kind of product know-how map construction and querying method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10726022B2 (en) * | 2016-08-26 | 2020-07-28 | Facebook, Inc. | Classifying search queries on online social networks |
-
2019
- 2019-02-14 CN CN201910114989.1A patent/CN110162637B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9026524B1 (en) * | 2013-01-10 | 2015-05-05 | Relationship Science LLC | Completing queries using transitive closures on a social graph |
| CN106777331A (en) * | 2017-01-11 | 2017-05-31 | 北京航空航天大学 | Knowledge mapping generation method and device |
| CN108509420A (en) * | 2018-03-29 | 2018-09-07 | 赵维平 | Gu spectrum and ancient culture knowledge mapping natural language processing method |
| CN108763555A (en) * | 2018-06-01 | 2018-11-06 | 北京奇虎科技有限公司 | Representation data acquisition methods and device based on demand word |
| CN109033223A (en) * | 2018-06-29 | 2018-12-18 | 北京百度网讯科技有限公司 | For method, apparatus, equipment and computer readable storage medium across type session |
| CN109145122A (en) * | 2018-08-02 | 2019-01-04 | 北京仿真中心 | A kind of product know-how map construction and querying method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110162637A (en) | 2019-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110162637B (en) | Information map construction method, device and equipment | |
| CN108804641B (en) | Text similarity calculation method, device, equipment and storage medium | |
| CN105893349B (en) | Classification tag match mapping method and device | |
| US10102191B2 (en) | Propagation of changes in master content to variant content | |
| CN108268600B (en) | AI-based unstructured data management method and device | |
| WO2017211051A1 (en) | Mining method and server for social network account of target subject, and storage medium | |
| CN104866478B (en) | Malicious text detection and identification method and device | |
| CN113590776B (en) | Knowledge graph-based text processing method and device, electronic equipment and medium | |
| CN111680498B (en) | Entity disambiguation method, device, storage medium and computer equipment | |
| CN112650858B (en) | Emergency assistance information acquisition method and device, computer equipment and medium | |
| CN112559631B (en) | Data processing method and device of distributed graph database and electronic equipment | |
| CN111400504A (en) | Method and device for identifying enterprise key people | |
| CN109492027B (en) | Cross-community potential character relation analysis method based on weak credible data | |
| CN111522968A (en) | Knowledge graph fusion method and device | |
| CN113033194A (en) | Training method, device, equipment and storage medium of semantic representation graph model | |
| CN111984797A (en) | Customer identity recognition device and method | |
| CN107704538A (en) | A kind of rubbish text processing method, device, equipment and storage medium | |
| CN115186738B (en) | Model training method, device and storage medium | |
| US20220335070A1 (en) | Method and apparatus for querying writing material, and storage medium | |
| CN110597982A (en) | Short text topic clustering algorithm based on word co-occurrence network | |
| CN113011152B (en) | Text processing method, device and equipment and computer readable storage medium | |
| CN113392220B (en) | Knowledge graph generation method and device, computer equipment and storage medium | |
| CN114398903A (en) | Intention recognition method and device, electronic equipment and storage medium | |
| CN112860811A (en) | Method and device for determining data blood relationship, electronic equipment and storage medium | |
| CN111476037B (en) | Text processing method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |