Methods of obtaining intelligence information around specific targets

Automatically expand and sort intelligence information through the connection path generation method, solving the problem of insufficient efficiency and accuracy of traditional intelligence analysis methods in the virtual environment, realizing efficient intelligence information acquisition and sorting, and supporting decision-making support.

CN111967250BActive Publication Date: 2024-11-29DATAINSIGHT
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202010721842.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-07-24
Publication Date
2024-11-29
Estimated Expiration
2040-07-24

AI Technical Summary

Technical Problem

Traditional intelligence analysis methods are difficult to quickly extract useful information in virtual information and data environments, especially when there are complex relationships between intelligence, and efficiency and accuracy cannot meet the needs of modern intelligence analysis.

Method used

Using the connection path generation method, by setting the target entity and its seed relationship, using search strategies to expand the target set, extracting entities and entity relationships, building intelligence sets, and sorting them according to factors such as timeliness and relevance, and automatically collecting and sorting them. Intelligence information.

Benefits of technology

Convert a large amount of unknown quality low-association information into high-association and high-value visual intelligence to improve the integrity, accuracy and efficiency of intelligence acquisition, and support analysis, investigation and decision-making.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN111967250B_ABST
    Figure CN111967250B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for acquiring intelligence information around a specific target. The method only requires a user to set a target entity of interest and an initial seed relationship, and then the entity relationship and the target entity can be automatically expanded by using a search engine or existing index data, and intelligence associated with the target entity can be collected. Then, the intelligence information is comprehensively sorted according to factors such as freshness, relevance to the target entity of interest, and path length, so as to capture intelligence information, build an intelligence collection, and determine the relationship network of the target entity, thereby obtaining various types of intelligence associated with the target entity, and organizing and displaying the intelligence. The method can convert a large amount of information of unknown quality, low relevance, and low value into a small amount of information that is easy to understand, highly relevance, and high value, which is visualized and operable, so as to improve the completeness, accuracy, and efficiency of obtaining the required situation, and thus provide assistance for analysis, investigation, and decision-making.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligence analysis, and in particular to a method for acquiring intelligence information around a specific target. Background Art

[0002] Intelligence is a strategic resource for enterprises to maintain their competitive advantage in the knowledge economy era, and is the basis for analyzing and predicting industry development trends and making scientific strategic decisions. It is recognized as the "fourth core competitiveness" of enterprises in addition to capital, technology, and talent. At present, most information resources are based on the Internet, and have undergone a morphological upgrade from form to content. Digital resources account for an increasing proportion of all information resources. Digital information resources have the characteristics of large quantity, complex format, fast update speed, and strong timeliness. Traditional intelligence analysis methods are mainly carried out manually, and it is difficult to quickly extract useful information and effectively use it in the virtual information and data environment space environment. Especially when there are complex correlations between intelligence, manual analysis methods cannot meet the current intelligence analysis needs in terms of efficiency and accuracy. Summary of the invention

[0003] In view of the above problems, the present invention proposes a method for acquiring intelligence information around a specific target.

[0004] To achieve the purpose of the present invention, a method for generating a personal network path is provided, comprising the following steps:

[0005] S12, setting the target entity of interest and the seed relationship of the target entity, and setting the target network resource associated with the target entity;

[0006] S13, adding the target entity to the target set, and adding the seed relationship of the target entity to the relationship set;

[0007] S14, using a set search strategy to expand the target entity in the target set along the seed relationship to obtain an extended relationship set;

[0008] S15, identifying the relationship subjects in the extended relationship set, and acquiring the related web page set of each relationship subject in the target network resource;

[0009] S16, extracting entities and entity relationships from the set of related web pages, and expanding the target set and the relationship set according to the extracted entities and entity relationships, so as to update the target set and the relationship set;

[0010] S17, for the updated target set and relationship set, looping through steps S14 to S16 until the updated target set and relationship set meet the set analysis requirements;

[0011] S18, capturing each piece of intelligence information of the target set according to the updated target set and relationship set, and constructing an intelligence set according to each piece of intelligence information.

[0012] In one embodiment, the method for acquiring intelligence information around a specific target is characterized by further comprising:

[0013] The relationship network of the target entity is extracted according to the updated target set and relationship set.

[0014] In one embodiment, the set search strategy includes a depth-first strategy, a breadth-first strategy, or a credibility-first strategy.

[0015] In one embodiment, constructing an intelligence set according to various intelligence information includes:

[0016] The web pages associated with each piece of intelligence information are sorted according to their freshness, relevance to the target entity, or path length, and an intelligence set of the target entity is constructed according to the sorting result.

[0017] In one embodiment, the target entity includes a word or phrase with descriptive meaning, content with specific meaning in a specific field, and / or a domain-specific vocabulary;

[0018] The target entities have attribute relationships, causal relationships, temporal relationships, inversion relationships, conditional relationships and / or superior-subordinate event relationships.

[0019] In one embodiment, the seed relationship includes an entity relationship associated with the target entity given by a corresponding domain intelligence expert, so as to initialize the corresponding algorithm.

[0020] In one embodiment, before step S12, the method further includes:

[0021] S11, collect intelligence corpus to build corpus;

[0022] The extracting entities and entity relationships from the set of related web pages includes:

[0023] Based on the corpus, entities and entity relationships in a set of related web pages are extracted by using a set of rule templates and a machine learning method.

[0024] Specifically, the expanding the target set and the relationship set according to the extracted entities and entity relationships respectively includes:

[0025] According to the freshness of the web pages associated with the extracted entities or the strength of the relationship with the target entity, the extracted entities and entity relationships are sorted and selected, and the target set and the relationship set are expanded according to the results of the sorting and selection.

[0026] In one embodiment, the set analysis requirements include the amount of intelligence, the search depth, or the analysis granularity.

[0027] The above-mentioned intelligence information acquisition method around a specific target only requires the user to set the target entity of interest and the initial seed relationship, and then use the search engine or existing index data to automatically expand the entity relationship and the target entity, and collect the intelligence associated with the target entity; then sort it according to factors such as timeliness, relevance to the target entity of interest, and path length to capture intelligence information, build an intelligence collection, and determine the relationship network of the target entity, so as to obtain various types of intelligence associated with the target entity, organize and display the intelligence. It can convert a large amount of information of unknown quality, low relevance, and low value into a small amount of easy-to-understand, highly relevant, high-value visual and actionable intelligence, so as to improve the completeness, accuracy, and efficiency of obtaining the required situation, thereby providing assistance for analysis, investigation, and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of a method for acquiring intelligence information around a specific target according to an embodiment;

[0029] Figure 2 It is a flow chart of a method for acquiring intelligence information around a specific target according to another embodiment. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0031] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0032] In one embodiment, reference Figure 1 As shown, Figure 1 The flowchart of the method for obtaining intelligence information around a specific target of this embodiment includes the following steps:

[0033] S12, setting the target entity of interest and the seed relationship of the target entity, and setting the target network resource associated with the target entity.

[0034] In the above steps, users can set the target entity and the seed relationship of the target entity according to the entity corresponding to the intelligence (intelligence information) they need to obtain. The above target network resources can include online scopes, and users can also set various types of web pages associated with the target entity (online scopes) and the subject types related to the target entity to set the target network resources associated with the target entity. The above online scope refers to which network resources are obtained, such as Sina News webpages, Sina Weibo, or some forums, etc.; the subject type refers to the subject of concern, for example, the target of concern is a well-known person Zhang San, and the subject is attending a meeting or participating in an event, etc.

[0035] S13, adding the target entity to the target set, and adding the seed relationship of the target entity to the relationship set.

[0036] In one embodiment, the target entity includes a word or phrase with descriptive meaning, content with specific meaning in a specific field, and / or a domain-specific vocabulary;

[0037] The target entities have attribute relationships, causal relationships, temporal relationships, inversion relationships, conditional relationships and / or superior-subordinate event relationships.

[0038] The target entity concerned in this embodiment refers to a word or phrase with descriptive meaning, which can be a person's name, a place name, an organization name, a product name, or content with a certain meaning in a certain field, such as the name of a disease, a drug, an organism in the medical field, or a special vocabulary involved in the field, etc. The relationship between target entities (such as entity relationship) includes but is not limited to: attribute relationship, causal relationship, temporal relationship, inversion relationship, conditional relationship, and superior-subordinate event relationship, etc.

[0039] In one embodiment, the seed relationship includes an entity relationship associated with the target entity given by a corresponding domain intelligence expert, so as to initialize the corresponding algorithm. For example, when the target entity is a certain enterprise, the seed relationship may be "Chairman".

[0040] The collection of web pages or index data related to the target entity is used to define the scope and data basis for searching when expanding using the seed relationship.

[0041] S14, using a set search strategy to expand the target entities in the target set along the seed relationship to obtain an extended relationship set.

[0042] In one embodiment, the set search strategy includes a depth-first strategy, a breadth-first strategy, or a credibility-first strategy.

[0043] In this embodiment, the target entity in the target set is expanded along the seed relationship based on the index data of the search engine or the existing document set, and is realized by entity extraction and entity relationship extraction. The corresponding expansion method includes: taking the entity as a "node" and the entity relationship as an "edge", reaching the next node along the edge from the current node, and adding the next node to the target set. These nodes and edges constitute a "relationship network". The search strategy can be determined as needed, such as breadth first or depth first. Taking a well-known figure Zhang San as an example, Zhang San's classmate Li XX can be extracted in the corresponding expansion process, and Li XX is also placed in the target entity, that is, according to the classmate relationship, the original entity is expanded from the original {"Zhang San"} to {"Zhang San", "Li XX"}.

[0044] S15, identifying the relationship subjects in the extended relationship set, and acquiring the related web page set of each relationship subject in the target network resource.

[0045] The above steps can be based on the index data of the search engine or the existing document collection to obtain the relevant web page set of each target entity (such as the web pages included in the target network resource).

[0046] S16, extracting entities and entity relationships from the relevant web page set, and expanding the target set and the relationship set according to the extracted entities and entity relationships, so as to update the target set and the relationship set.

[0047] S17, for the updated target set and relationship set, loop through steps S14 to S16 until the updated target set and relationship set meet the set analysis requirements.

[0048] In one embodiment, the set analysis requirements include the amount of intelligence, the search depth, or the analysis granularity.

[0049] In this embodiment, analysis requirements may be preset, such as setting conditions such as target entities reaching a certain number, web page collections reaching a certain number, or search depth reaching a certain number as analysis requirements.

[0050] S18, capturing each piece of intelligence information of the target set according to the updated target set and relationship set, and constructing an intelligence set according to each piece of intelligence information.

[0051] In one embodiment, constructing an intelligence set according to various intelligence information includes:

[0052] The web pages associated with each piece of intelligence information are sorted according to their freshness, relevance to the target entity, or path length, and an intelligence set of the target entity is constructed according to the sorting result.

[0053] This embodiment can identify various types of intelligence information corresponding to the target entity from the updated target set and relationship set, and sort all the obtained intelligence information according to factors such as the freshness of the relevant web pages, the relevance to the target entity of concern, and the path length, so as to construct a complete and ordered intelligence set.

[0054] The above-mentioned freshness refers to the immediacy of information. For example, yesterday's weather forecast is invalid today. Only the latest one is valuable. The above-mentioned relationship network can include the relationship between the target entity and related entities. For example, A is related to B, B is related to C, C is related to D, etc. At this time, A, B, C, D and the relationship between them form a relationship network.

[0055] The above-mentioned intelligence information acquisition method around a specific target only requires the user to set the target entity of interest and the initial seed relationship, and then use the search engine or existing index data to automatically expand the entity relationship and the target entity, and collect the intelligence associated with the target entity; then sort it according to factors such as timeliness, relevance to the target entity of interest, and path length to capture intelligence information, build an intelligence collection, and determine the relationship network of the target entity, so as to obtain various types of intelligence associated with the target entity, organize and display the intelligence, which can convert a large amount of information of unknown quality, low relevance, and low value into a small amount of easy-to-understand, high-relevance, and high-value visual information.

[0056] digitized and actionable intelligence to improve the completeness, accuracy and efficiency of obtaining the required information, thereby assisting analysis, investigation and decision-making.

[0057] In one embodiment, the above-mentioned method for acquiring intelligence information around a specific target further includes:

[0058] The relationship network of the target entity is extracted according to the updated target set and relationship set.

[0059] This embodiment can also display the obtained relationship network to the user as part of the result. The above relationship network and intelligence set can complement and update each other. For example, the relationship network corresponding to the target entity can be updated based on the information included in the target entity intelligence set, and the corresponding intelligence set can be updated based on the relationship network of the target entity, etc.

[0060] In one embodiment, before step S12, the method further includes:

[0061] S11, collect intelligence corpus to build corpus;

[0062] The extracting entities and entity relationships from the set of related web pages includes:

[0063] Based on the corpus, entities and entity relationships in a set of related web pages are extracted by using a set of rule templates and a machine learning method.

[0064] The sources of the intelligence corpus in the above corpus can be internal information, information from the Internet, and third-party information.

[0065] When performing entity relationship extraction in this embodiment, entity relationship extraction can be regarded as a classification task, and entity relationship extraction can be performed based on a corpus, combined with domain intelligence expert rule templates and machine learning methods. Entity relationships can be marked with attribute values ​​such as relationship strength and probability.

[0066] Specifically, the expanding the target set and the relationship set according to the extracted entities and entity relationships respectively includes:

[0067] According to the freshness of the web pages associated with the extracted entities or the strength of the relationship with the target entity, the extracted entities and entity relationships are sorted and selected, and the target set and the relationship set are expanded according to the results of the sorting and selection.

[0068] In the above steps S14 and S16, the newly identified entities and entity relationships may be sorted and selected according to certain rules, such as freshness and strength of relationship with the target of interest, and the target set and / or relationship set may be expanded.

[0069] Specifically, for each piece of intelligence identified from the updated target set and relationship set, a path from the target of interest along the relationship network to another entity included in the intelligence can be given, as well as the label of each edge of the path.

[0070] In one embodiment, the above-mentioned intelligence information acquisition method around a specific target can also refer to Figure 2 As shown, it includes the following steps:

[0071] Step 1: Collect and organize intelligence data;

[0072] For example, if the focus is on technology companies, then internal information of technology companies, information from the Internet, and third-party information are collected to form a corpus for subsequent entity extraction and relationship extraction.

[0073] Step 2: Define the target entity of interest and the seed relationship of the target entity of interest; define the online scope and topic type related to the target entity;

[0074] For example, follow "XX Network Technology Co., Ltd." and define it as the target entity of interest; define seed relationships including "President", "Senior Vice President" and "Branch" etc.; specify the relevant web page set as the list of web pages returned from the website of XX Network Company plus a third-party search engine query for "XX Network".

[0075] Step 3: Add the target entity of interest to the target set, and add the seed relationship of the target entity of interest to the relationship set;

[0076] Add "XX Network Technology Co., Ltd." to the target set, and add "President", "Senior Vice President" and "Branch" to the relationship set;

[0077] Step 4: Based on a certain search strategy, expand the target entities in the current target set along the relationship;

[0078] When expanding, we take entities as "nodes" and entity relationships as "edges", follow the edges from the current node to the next node, and add the next node to the target set. These nodes and edges form a "relationship network". The search strategy can be determined as needed, such as breadth first or depth first.

[0079] For example, the method for expanding "XX Network Technology Co., Ltd." along the entity relationship "President" is: search and extract the relationship pattern of "President of XX Network Technology Co., Ltd." in the web page collection, obtain the entity "Chen Moumou", and add it to the target collection. When expanding, the most up-to-date data should be selected.

[0080] Similarly, the method for expanding "XX Network Technology Co., Ltd." along the relationship "branch" is: search and extract the relationship pattern of "XX Network Technology Co., Ltd.'s branch structure" in the web page collection, obtain the entities "XX Network Technology Co., Ltd. Sichuan Branch", "XX Network Technology Co., Ltd. Mudanjiang Branch", etc., and add them to the target collection. When expanding, the data with the highest relevance to the target of interest should be selected.

[0081] Step 5: Based on the index data of the search engine or the existing document collection, obtain the relevant web page set for each target entity;

[0082] Get the web page collection related to "Chen XX", "XX Network Technology Co., Ltd. Sichuan Branch", "XX Network Technology Co., Ltd. Mudanjiang Branch", etc.

[0083] Step 6: Extract entities and entity relationships from the web page set obtained in step 5, and use them to expand the target set and relationship set;

[0084] For example, in the web page related to "Chen Moumou", which contains the information "XX Network CEO / President Chen Moumou attended the Software Celebrity Forum and delivered a keynote speech on "Cloud Services Activate Enterprise Digitalization", entity extraction and entity relationship extraction are performed to obtain the entity "Software Celebrity Forum" and the entity relationship "Attendance"; the entity "Software Celebrity Forum" is added to the target set, and the entity relationship "Attendance" is added to the relationship set.

[0085] In step 4 and step 6, entity extraction and entity relationship extraction are performed based on the corpus established in step 1, combined with domain intelligence expert rule templates and machine learning methods. Entity relationship extraction is regarded as a classification task, and entity relationships can be marked with attribute values ​​such as relationship strength and probability.

[0086] Step 7: Repeat steps 4-6 until the analysis requirements are met. The analysis requirements can be the amount of intelligence, the depth of search, the granularity of intelligence analysis, etc.

[0087] Step 8: Sort all the intelligence obtained according to the freshness, relevance to the target entity of interest, path length and other factors, and display the relationship network to the user as part of the result. Each piece of intelligence generated can give the path from the target of interest along the relationship network to another entity included in the intelligence, as well as the label of each edge of the path. This method can be used to obtain the relationship mining of intelligent intelligence analysis results. For example, the path from the target of interest "XX Network Technology Co., Ltd." along the relationship network to "Chen Moumou" has a path length of 1 and the edge is labeled "Director, President". This method can be used to obtain intelligence classification of intelligent intelligence analysis results, including "character dynamics", "competitive intelligence" and "corporate risks". Among them, "character dynamics" includes "leadership visits", "leadership activities", "position changes" and "public speeches".

[0088] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0089] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0090] The terms "including" and "having" and any variations thereof in the embodiments of the present application are intended to cover non-exclusive inclusions. For example, a process, method, device, product, or equipment that includes a series of steps or modules is not limited to the listed steps or modules, but may optionally include steps or modules that are not listed, or may optionally include other steps or modules that are inherent to these processes, methods, products, or equipment.

[0091] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for acquiring intelligence information around a specific target, characterized in that: The steps include: S12, setting the target entity of interest and the seed relationship of the target entity, and setting the target network resource associated with the target entity; S13, adding the target entity to the target set, and adding the seed relationship of the target entity to the relationship set; S14, using a set search strategy to expand the target entity in the target set along the seed relationship to obtain an extended relationship set; S15, identifying the relationship subjects in the extended relationship set, and acquiring the related web page set of each relationship subject in the target network resource; S16, extracting entities and entity relationships from the set of related web pages, and expanding the target set and the relationship set according to the extracted entities and entity relationships, so as to update the target set and the relationship set; S17, for the updated target set and relationship set, looping through steps S14 to S16 until the updated target set and relationship set both meet the set analysis requirements; S18, capturing each intelligence information of the target set according to the updated target set and relationship set, and constructing an intelligence set according to each intelligence information; Among them, building an intelligence collection based on various intelligence information includes: Sort the web pages associated with each intelligence information according to their freshness, relevance to the target entity, or path length, and construct an intelligence set of the target entity based on the sorting results; The target entities include words or phrases with descriptive meanings, content with specific meanings in a specific field, and / or specialized vocabulary in a specific field; The target entities have attribute relationships, causal relationships, temporal relationships, reversal relationships, conditional relationships and / or superior-subordinate event relationships; The seed relationship includes an entity relationship associated with the target entity given by an intelligence expert in the corresponding field, so as to be used for initializing the corresponding algorithm.

2. The method for acquiring intelligence information around a specific target according to claim 1, characterized in that: Also includes: The relationship network of the target entity is extracted according to the updated target set and relationship set.

3. The method for acquiring intelligence information around a specific target according to claim 1, characterized in that: The set search strategy includes a depth-first strategy, a breadth-first strategy, or a credibility-first strategy.

4. The method for acquiring intelligence information around a specific target according to claim 1, characterized in that: Before step S12, the method further includes: S11, collect intelligence corpus to build corpus; The extracting entities and entity relationships from the set of related web pages includes: Based on the corpus, entities and entity relationships in a set of related web pages are extracted by using a set of rule templates and a machine learning method.

5. The method for acquiring intelligence information around a specific target according to claim 4, characterized in that: The expanding the target set and the relationship set respectively according to the extracted entities and entity relationships comprises: According to the freshness of the web pages associated with the extracted entities or the strength of the relationship with the target entity, the extracted entities and entity relationships are sorted and selected, and the target set and the relationship set are expanded according to the results of the sorting and selection.

6. The method for acquiring intelligence information around a specific target according to claim 1, characterized in that: The set analysis requirements include the amount of intelligence, search depth or analysis granularity.

Citation Information

Patent Citations

  • Methods and systems for providing search capabilities for instant messaging tools

    CN102298627A

  • User attention based network sensitive information monitoring system and method

    CN103853720A