Client view acquisition method and device and storage medium

Abstract

The invention discloses a client view obtaining method and device and a storage medium, and the method comprises the steps: obtaining client information and business data of each subsidiary company, and constructing a target organization structure tree based on the client information and the business data; performing word segmentation on the data in the target organizational structure tree to generate an enterprise word library; external multi-source data are obtained, feature extraction is carried out on the multi-source data, and corresponding customer feature tags are generated; matching the enterprise lexicon with the customer feature tag, and importing an obtained matching result into a search engine; and obtaining a search request of a user, and outputting a corresponding client view based on the search request. According to the method, the corresponding customer view is output based on the search request, so that the business structure and hierarchical relationship of the customer can be obtained, the customer feature tag is mastered in real time, powerful data support is provided for subsequent precision marketing, customer relationship management and business decision, and the business processing efficiency and accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a method and device for obtaining customer view and storage medium. BACKGROUND

[0002] In the field of insurance services, the enterprise customer relationship management (CRM) system as the core business support platform, its performance will affect the quality of customer service and business success or failure. Among them, the enterprise customer relationship management can show the basic information of the customer and the historical communication data of the customer, so as to understand the customer relationship and communication history.

[0003] However, the enterprise customer relationship management system in the related art has single information display, which makes it difficult for business personnel to quickly and accurately locate target customers or their associated enterprises from a large number of customers, thereby increasing the complexity and time cost of information retrieval and reducing the business processing efficiency. SUMMARY

[0004] The present application provides a method and device for obtaining customer view to solve the above technical problems in the related art.

[0005] Therefore, the present application provides a method for obtaining customer view, which can match enterprise vocabulary with customer feature labels and import the matching results into a search engine, so that the corresponding customer view can be output based on the search request, thereby obtaining the business structure and hierarchical relationship of the customer and real-time grasping the customer feature labels, providing strong data support for subsequent precision marketing, customer relationship management and business decision-making, and improving the efficiency and accuracy of business processing.

[0006] Another object of the present application is to provide a device for obtaining customer view.

[0007] To achieve the above object, the present application provides a method for obtaining customer view, which comprises: obtaining customer information and business data of each subsidiary company, and constructing a target organizational structure tree based on the customer information and the business data; segmenting the data in the target organizational structure tree to generate an enterprise vocabulary; obtaining external multi-source data, and extracting features from the multi-source data to generate corresponding customer feature labels; matching the enterprise vocabulary with the customer feature labels, and importing the matching results into a search engine; obtaining a search request of a user, and outputting a corresponding customer view based on the search request.

[0008] The customer view acquisition method of the embodiment of the present application can further have the following additional technical features. In an embodiment of the present application, the customer information and business data of each subsidiary company are acquired, and a target organizational structure tree is constructed based on the customer information and the business data, comprising: An all-amount organizational structure tree is constructed based on the data of the offline database through HIVE; The customer information and business data of each subsidiary company are acquired, and the customer information and business data are extracted from the data of each subsidiary company through a multi-relation search (MRS) loading tool; The customer information and the business data are stored in a data cache queue; An incremental update is performed on the all-amount organizational structure tree based on real-time data streams in the data cache queue through Flink, a target organizational structure tree is obtained, and the data in the data cache queue is stored in the offline database.

[0009] In an embodiment of the present application, the data in the target organizational structure tree is segmented to generate an enterprise vocabulary, comprising: Basic terms are extracted from the target organizational structure tree; Term frequency, part of speech, and hierarchical relationship of the basic terms are determined; A dictionary file is generated based on the term frequency, part of speech, and hierarchical relationship of the basic terms; The dictionary file is loaded by a segmenter to generate an enterprise vocabulary.

[0010] In an embodiment of the present application, external multi-source data are acquired, and feature extraction is performed on the multi-source data to generate corresponding customer feature labels, comprising: External multi-source data are acquired, and the external multi-source data are classified and preprocessed to obtain preprocessed data; The preprocessed data are input into a natural language processing engine to perform hierarchical processing to obtain a customer feature view; Based on the customer feature view, customer feature labels are generated according to a preset label system and mapping rules.

[0011] In an embodiment of the present application, the natural language processing engine comprises a basic analysis layer, a deep semantic understanding layer, and a feature correlation aggregation layer; the preprocessed data are input into the natural language processing engine to perform hierarchical processing to obtain a customer feature view, comprising: The preprocessed data are segmented, part of speech is annotated, entities are recognized, and key information is extracted through the basic analysis layer to obtain basic information elements; The basic information elements are analyzed from a plurality of predefined customer feature dimensions through the deep semantic understanding layer to obtain customer features. The customer feature view is obtained by associating and aggregating the customer features using the customer identifier as the key through the feature association aggregation layer.

[0012] In one embodiment of the present invention, the method further includes: Retrieve annotation data for user-submitted anomaly tags; The customer feature tags are corrected and updated based on the labeled data.

[0013] In one embodiment of the present invention, the method further includes: The customer feature tags of the customer are periodically acquired and stored, and a snapshot of the customer profile is formed based on the customer feature tags; Based on the customer profile snapshots arranged in chronological order, construct the customer's historical trajectory data; Obtain the user's historical analysis requests for the target customer, and obtain the target customer's historical trajectory data based on the historical analysis requests; Based on the historical trajectory data of the target customer, the evolution process of the target customer's feature tags within a preset time range is output in the form of a visual chart.

[0014] To achieve the above objectives, another aspect of the present invention provides a customer view acquisition device, the device comprising: The module is used to acquire customer information and business data of each subsidiary, and to construct the target organizational structure tree based on the customer information and business data; The first generation module is used to segment the data in the target organizational structure tree into words and generate an enterprise thesaurus. The second generation module is used to acquire external multi-source data, extract features from the multi-source data, and generate corresponding customer feature tags. The matching module is used to match the enterprise thesaurus with the customer feature tags and import the matching results into the search engine; The output module is used to obtain the user's search request and output the corresponding customer view based on the search request.

[0015] Another object of the present invention is to provide an electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in any one of the preceding aspects.

[0016] Another object of the present invention is to provide a computer storage medium storing computer-executable instructions; the computer-executable instructions, when executed by a processor, cause the computer to perform the method described in any one of the preceding aspects.

[0017] This invention discloses a method, apparatus, and storage medium for acquiring customer views. The method involves acquiring customer information and business data from various subsidiaries, constructing a target organizational structure tree based on this information, and segmenting the data in the target organizational structure tree to generate an enterprise thesaurus. It also involves acquiring external multi-source data, extracting features from the multi-source data, and generating corresponding customer feature tags. The enterprise thesaurus is then matched with the customer feature tags, and the matching results are imported into a search engine. Finally, the invention acquires user search requests and outputs corresponding customer views based on these requests. Therefore, this invention can match an enterprise thesaurus with customer feature tags and import the matching results into a search engine, enabling the output of corresponding customer views based on search requests. This allows for the acquisition of customer business structures and hierarchical relationships, and real-time monitoring of customer feature tags, providing strong data support for subsequent precision marketing, customer relationship management, and business decision-making, thereby improving the efficiency and accuracy of business processing.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0019] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a method for obtaining a customer view according to an embodiment of the present invention; Figure 2 This is a structural diagram of a customer view acquisition device according to an embodiment of the present invention. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] In related technologies, enterprise customer relationship management systems mainly focus on the storage and management of static information, such as company files and offline data. While the aforementioned static information is important for long-term preservation, in a rapidly changing market environment, information with extremely high timeliness, such as news updates and market intelligence, also has immeasurable value and urgently needs to be captured and processed in real time.

[0023] Furthermore, in one embodiment of the present invention, the enterprise customer relationship management system in the related art has shortcomings in the acquisition and processing of public opinion information. Not only is its timeliness low, but it also relies on manual tagging, thus limiting the timeliness and accuracy of the information. Based on this, the present invention proposes a method for acquiring a customer view, providing strong data support for subsequent precision marketing, customer relationship management, and business decision-making, thereby improving the efficiency and accuracy of business processing.

[0024] The method and apparatus for obtaining a customer view according to embodiments of the present invention are described below with reference to the accompanying drawings.

[0025] Figure 1 This is a flowchart of a method for obtaining a customer view according to an embodiment of the present invention.

[0026] like Figure 1 As shown, the method includes: S1: Obtain customer information and business data from each subsidiary, and construct the target organizational structure tree based on the customer information and business data.

[0027] In one embodiment of the present invention, the method for obtaining customer information and business data of each subsidiary and constructing a target organizational structure tree based on the customer information and business data may include the following steps: S11 uses HIVE data from an offline database to build a full organizational structure tree.

[0028] In one embodiment of the present invention, the data in the offline database is existing and saved data.

[0029] Furthermore, in one embodiment of the present invention, a full organizational structure tree can be constructed using HIVE based on data from an offline database.

[0030] S12: Obtain customer information and business data from each subsidiary, and extract customer information and business data from the data of each subsidiary through the loading tool of the multi-relationship search MRS.

[0031] In one embodiment of the present invention, after periodically acquiring customer information and business data of each subsidiary, customer information and business data can be extracted from the data of each subsidiary through the loading tool of the Multi-Relationship Search MRS.

[0032] In one embodiment of the present invention, customer information and business data can be extracted from the data of each subsidiary through the Loader tool of the Multi-Relationship Search MRS.

[0033] In one practical example of the present invention, the above-mentioned customer information may include basic customer information and change information.

[0034] S13 stores customer information and business data in a data cache queue.

[0035] In one embodiment of the present invention, after obtaining customer information and business data through the above steps, the customer information and business data can be stored in a data cache queue to ensure the stability and fault tolerance of the data loading process.

[0036] In one embodiment of the present invention, by introducing a data cache queue, a "peak shaving and valley filling" strategy can be used to achieve smooth data loading.

[0037] S14 uses Flink to incrementally update the full organizational structure tree based on the real-time data stream in the data cache queue, obtains the target organizational structure tree, and stores the data in the data cache queue to the offline database.

[0038] In one embodiment of the present invention, real-time data streams from the data storage queue can be periodically acquired to incrementally update the full organizational structure tree and obtain the target organizational structure tree. In one embodiment of the present invention, the period can be set as needed, such as one day. Furthermore, in one embodiment of the present invention, if there is no updated real-time data stream in the data storage queue, then there is no need to incrementally update the full organizational structure tree.

[0039] In one embodiment of the present invention, the full organizational structure tree is incrementally updated by Flink based on the real-time data stream in the data cache queue. After obtaining the target organizational structure tree, the data in the data cache queue can be stored in an offline database.

[0040] In one embodiment of the present invention, if the offline database is migrated or backed up, the full organizational structure tree can be reconstructed using HIVE based on the data in the offline database to ensure data consistency.

[0041] S2 segments the data in the target organizational structure tree into words to generate an enterprise thesaurus.

[0042] In one embodiment of the present invention, after determining the target organizational structure tree through the above steps, the data in the target organizational structure tree can be segmented into words to generate an enterprise thesaurus, so as to provide a data foundation for subsequent text analysis and information retrieval.

[0043] Specifically, in one embodiment of the present invention, the method for segmenting data in the target organizational structure tree to generate an enterprise thesaurus may include the following steps: S21, Extract basic terms from the target organizational structure tree.

[0044] In one embodiment of the present invention, each level is split from the full path in the target organizational structure tree to obtain basic terms.

[0045] For example, in one embodiment of the present invention, the basic terms may include: department name: Group / XX District / Technology Center; position / job title: job title; employee name: extract the corresponding employee name.

[0046] S22, determine the word frequency, part of speech, and hierarchical relationship of the basic terms.

[0047] In one embodiment of the present invention, the part-of-speech tagging results of the basic terms are obtained, for example, nouns.

[0048] In one embodiment of the present invention, the word frequency of each basic term can be determined by statistics.

[0049] In one embodiment of the present invention, the hierarchical relationship of basic terms can be obtained through the target organizational structure tree or annotation.

[0050] S23, Generate a dictionary file based on the word frequency, part of speech, and hierarchical relationship of the basic entries.

[0051] In one embodiment of the present invention, the word frequency, part-of-speech and hierarchical relationship of the basic entries obtained through the above steps are merged, and a dictionary file is generated according to the format of the word segmenter.

[0052] S24: Load the dictionary file using the word segmenter to generate the enterprise thesaurus.

[0053] In one embodiment of the present invention, after obtaining the dictionary file through the above steps, a word segmenter can be used to load the dictionary file and generate an enterprise thesaurus. In another embodiment of the present invention, an NLP word segmenter can be used to load the dictionary file and generate an enterprise thesaurus.

[0054] S3 acquires external multi-source data, extracts features from the multi-source data, and generates corresponding customer feature tags.

[0055] In one embodiment of the present invention, the method for acquiring external multi-source data, extracting features from the multi-source data, and generating corresponding customer feature tags may include the following steps: S31. Acquire external multi-source data, classify and preprocess the external multi-source data to obtain preprocessed data.

[0056] In one embodiment of the present invention, the aforementioned external multi-source data may include publicly available data such as industry research reports, regional research reports, business registration information, corporate public opinion, and listed company annual reports.

[0057] In one embodiment of the present invention, after acquiring external multi-source data, data cleaning and alignment can be performed on the external multi-source data to obtain a structured text dataset. Specifically, in one embodiment of the present invention, the method for obtaining a structured text dataset by cleaning and aligning the external multi-source data may include: removing special symbols (such as emoticons and HTML tags), using Unified Encoding (UTF-8), and completing short sentences from the external multi-source data; and unifying customer identifiers and aligning missing values ​​to obtain a structured text dataset.

[0058] In one embodiment of the present invention, customer data from different channels are mapped to customer IDs in the target organizational structure tree.

[0059] S32 inputs the preprocessed data into the natural language processing engine to perform hierarchical processing and obtain a customer feature view.

[0060] In one embodiment of the present invention, after obtaining the preprocessed data through the above steps, the preprocessed data can be input into a natural language processing engine to perform hierarchical processing and obtain a customer feature view.

[0061] In one embodiment of the present invention, the natural language processing engine may include a basic parsing layer, a deep semantic understanding layer, and a feature association aggregation layer. Furthermore, in one embodiment of the present invention, the method of inputting preprocessed data into the natural language processing engine for hierarchical processing to obtain a customer feature view may include the following steps: S321, the preprocessed data is segmented, part-of-speech tagging is performed, entity recognition is performed and key information is extracted through the basic parsing layer to obtain basic information elements.

[0062] In one embodiment of the present invention, the aforementioned basic parsing layer may include a word segmenter, an NER model, and a fine-tuned BERT model.

[0063] In one embodiment of the present invention, a word segmenter can be used to segment and tag the preprocessed data to obtain a segmented sequence.

[0064] In one embodiment of the present invention, entity recognition can be performed on the segmented sequence using the NER model to obtain entity recognition results.

[0065] In one embodiment of the present invention, key information can be extracted from the entity recognition results by fine-tuning the BERT model to obtain basic information elements.

[0066] S322 uses a deep semantic understanding layer to parse basic information elements from multiple predefined customer feature dimensions to obtain customer features.

[0067] In one embodiment of the present invention, the deep semantic understanding layer can use NLP to parse basic information elements from multiple predefined customer feature dimensions to obtain customer features.

[0068] In one embodiment of the present invention, the aforementioned customer characteristics may include the customer's basic characteristics, customer association information, customer's ability to fulfill obligations, customer's behavioral preferences, and customer's credit history.

[0069] S323, through the feature association aggregation layer, customer features are associated and aggregated using customer identifier as the key to obtain a customer feature view.

[0070] In one embodiment of the present invention, the feature association aggregation layer can associate and aggregate customer features through customer identifiers to obtain a customer feature view.

[0071] Specifically, in one embodiment of the present invention, the customer ID is used as the customer identifier, and the customer's basic characteristics, performance ability, credit history, related information and behavioral preferences are integrated to obtain a customer feature view.

[0072] S33 generates customer feature tags based on the customer feature view and according to the preset tag system and mapping rules.

[0073] In one embodiment of the present invention, after obtaining the customer feature view through the above steps, customer feature tags can be generated according to a preset tag system and mapping rules.

[0074] In one embodiment of the present invention, the aforementioned preset tag system and mapping rules can be set according to user needs.

[0075] In one embodiment of the present invention, Table 1 is a relational table of a preset label system.

[0076] Table 1

[0077] Furthermore, in one embodiment of the present invention, customer feature tags can be generated according to mapping rules. Also, in one embodiment of the present invention, the mapping rules can be set based on experience or statistical data.

[0078] For example, assuming that customer A's associated information in the customer feature view includes: subsidiary B and supplier C, then the corresponding relationship network tags for customer A include [subsidiary B] and [supplier C]; and that customer A's credit history includes: AAA-rated credit enterprise, then the corresponding credit rating tag for customer A includes [AAA-rated credit].

[0079] S4 matches the enterprise thesaurus with customer feature tags and imports the matching results into the search engine.

[0080] In one embodiment of the present invention, after determining the enterprise thesaurus and customer feature tags through the above steps, the enterprise thesaurus and customer feature tags can be matched, and the matching results can be imported into the search engine.

[0081] In one embodiment of the present invention, matching can be performed based on the enterprise thesaurus and the customer ID in the customer feature tags to obtain the corresponding matching results. Furthermore, in one embodiment of the present invention, after obtaining the matching results through the above steps, the obtained matching results can be imported into a search engine. In one embodiment of the present invention, the search engine can be Elasticsearch.

[0082] S5 retrieves the user's search request and outputs the corresponding customer view based on the search request.

[0083] Furthermore, in one embodiment of the present invention, after importing the matching results obtained through the above steps into the search engine, the user's search request can be obtained, and the corresponding customer view can be output based on the search request.

[0084] In one embodiment of the present invention, a fuzzy search function can be supported, which can accurately locate target customers by using keywords such as customer names or company names entered by users, and comprehensively display the hierarchical relationship of target customers, including head offices, branches and subsidiaries.

[0085] Furthermore, in one embodiment of the present invention, advanced search functions with multiple conditions and multiple keywords can be supported, allowing users to set multiple search conditions as needed, such as company name, industry type, registered capital, etc., to filter out target customers that meet the conditions.

[0086] Furthermore, in one embodiment of the present invention, the technical components of Vue and ElementUI can be used to display customer information, underwriting, claims and modular business performance information, and related unified view information of subordinate companies at all levels at the target customer level.

[0087] Furthermore, in one embodiment of the present invention, the front end uses the Vue.js framework and the Element UI component library to build the user interface, realizing the dynamic display and interaction of the tree structure, which can display the hierarchical relationship of enterprise customers, including the head office, branches, subsidiaries, etc.

[0088] Furthermore, in one embodiment of the present invention, users can be divided into internal users and external users, and different users are subject to corresponding data access control. Specifically, in one embodiment of the present invention, for internal users, data access control can be implemented based on information such as the user's company and organization. For example, a customer of company A can view group customer information related to business generated by company A, and the business data is restricted to company A; for external users, they can query business information related to specific group customers.

[0089] Furthermore, in one embodiment of the present invention, a real-time API service can also be provided to support querying the corresponding relationship of the organizational structure tree of group enterprise customers through ES search. It can realize the search of specific enterprise customers through multiple conditions and multiple keywords, and support the rapid location and information display of enterprise customers at all levels. For example, it can support querying the list of enterprise customers at all levels under the group customer name, and also support tracing back to the group to which the subordinate enterprise belongs.

[0090] Furthermore, in one embodiment of the present invention, the above method may further include: obtaining annotation data of abnormal tags submitted by users, and correcting and updating customer feature tags based on the annotation data.

[0091] In one embodiment of the present invention, if a user discovers an anomaly in a customer feature tag during business application, the tag can be annotated, and the annotated data can be synchronized to a cache queue and periodically fed back to the natural language processing engine for retraining, thereby realizing online tagging and dynamic correction of tags, improving the accuracy and reliability of tags.

[0092] Furthermore, in one embodiment of the present invention, the above method may further include the following steps: Step 1: Periodically acquire and store customer feature tags, and form customer profile snapshots based on customer feature tags; Step 2: Construct historical trajectory data of customers based on customer profile snapshots arranged in chronological order; Step 3: Obtain the user's historical analysis requests for the target customer, and obtain the target customer's historical trajectory data based on the historical analysis requests; Step 4: Based on the historical trajectory data of the target customer, output the evolution process of the target customer's feature tags within a preset time range in the form of a visual chart.

[0093] In one embodiment of the present invention, the aforementioned historical analysis request may include customer information and the time range to be analyzed.

[0094] The customer view acquisition method proposed in this invention involves acquiring customer information and business data from each subsidiary, and constructing a target organizational structure tree based on the customer information and business data; segmenting the data in the target organizational structure tree to generate an enterprise thesaurus; acquiring external multi-source data and extracting features from the multi-source data to generate corresponding customer feature tags; matching the enterprise thesaurus with the customer feature tags and importing the matching results into a search engine; acquiring user search requests and outputting the corresponding customer view based on the search requests. Therefore, this invention can match the enterprise thesaurus with customer feature tags and import the matching results into a search engine, enabling the output of the corresponding customer view based on the search request. This allows for the acquisition of the customer's business structure and hierarchical relationships, and real-time monitoring of customer feature tags, providing strong data support for subsequent precision marketing, customer relationship management, and business decision-making, thereby improving the efficiency and accuracy of business processing.

[0095] To achieve the above embodiments, such as Figure 2 As shown, this embodiment also provides a customer view acquisition device 10, which includes: Module 201 is used to acquire customer information and business data of each subsidiary and build the target organizational structure tree based on the customer information and business data; The first generation module 202 is used to segment the data in the target organizational structure tree into words and generate an enterprise thesaurus. The second generation module 203 is used to acquire external multi-source data, extract features from the multi-source data, and generate corresponding customer feature tags. The matching module 204 is used to match the enterprise thesaurus with customer feature tags and import the matching results into the search engine; Output module 205 is used to obtain the user's search request and output the corresponding customer view based on the search request.

[0096] In one embodiment of the present invention, the above-mentioned construction module 201 is specifically used for: Build a full organizational structure tree using data from an offline database via HIVE; Obtain customer information and business data from each subsidiary, and extract customer information and business data from the data of each subsidiary through the multi-relationship search MRS loading tool; Store customer information and business data in a data cache queue; Flink uses real-time data streams from a data cache queue to incrementally update the full organizational structure tree, obtaining the target organizational structure tree, and then stores the data from the data cache queue into an offline database.

[0097] Furthermore, in one embodiment of the present invention, the first generation module 202 is specifically used for: Extract basic terms from the target organizational structure tree; Determine the word frequency, part of speech, and hierarchical relationship of basic terms; Dictionary files are generated based on the word frequency, part-of-speech, and hierarchical relationships of basic entries; Use a word segmenter to load dictionary files and generate an enterprise thesaurus.

[0098] Furthermore, in one embodiment of the present invention, the second generation module 203 is specifically used for: Acquire external multi-source data, classify and preprocess the external multi-source data to obtain preprocessed data; The preprocessed data is input into the natural language processing engine to perform hierarchical processing and obtain a customer feature view. Based on the customer feature view, customer feature tags are generated according to the preset tag system and mapping rules.

[0099] Furthermore, in one embodiment of the present invention, the natural language processing engine includes a basic parsing layer, a deep semantic understanding layer, and a feature association aggregation layer; the second generation module 203 is further configured to: The basic parsing layer performs word segmentation, part-of-speech tagging, entity recognition, and key information extraction on the preprocessed data to obtain basic information elements. Customer features are obtained by parsing basic information elements from multiple predefined customer feature dimensions through a deep semantic understanding layer; The customer feature view is obtained by associating and aggregating customer features using the customer identifier as the key through the feature association aggregation layer.

[0100] Furthermore, in one embodiment of the present invention, the above-described device is also used for: Retrieve annotation data for user-submitted anomaly tags; Customer feature tags are corrected and updated based on labeled data.

[0101] Furthermore, in one embodiment of the present invention, the above-described device is also used for: Periodically acquire and store customer feature tags, and form customer profile snapshots based on customer feature tags; Based on customer profile snapshots arranged in chronological order, construct historical trajectory data of customers; Obtain users' historical analysis requests for target customers, and obtain historical trajectory data of target customers based on historical analysis requests; Based on the historical trajectory data of target customers, the evolution of the target customer's feature tags within a preset time range is output in the form of visual charts.

[0102] The customer view acquisition device proposed in this invention acquires customer information and business data of each subsidiary, and constructs a target organizational structure tree based on the customer information and business data; it segments the data in the target organizational structure tree to generate an enterprise thesaurus; it acquires external multi-source data and extracts features from the multi-source data to generate corresponding customer feature tags; it matches the enterprise thesaurus with the customer feature tags and imports the matching results into a search engine; it acquires user search requests and outputs corresponding customer views based on the search requests. Therefore, this invention can match the enterprise thesaurus with customer feature tags and import the matching results into a search engine, enabling the output of corresponding customer views based on search requests. This allows for the acquisition of the customer's business structure and hierarchical relationships, and real-time monitoring of customer feature tags, providing strong data support for subsequent precision marketing, customer relationship management, and business decision-making, thereby improving the efficiency and accuracy of business processing.

[0103] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0104] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for obtaining a customer view, characterized in that, The method includes: Obtain customer information and business data from each subsidiary, and construct the target organizational structure tree based on the customer information and business data; The data in the target organizational structure tree is segmented into words to generate an enterprise thesaurus. Acquire external multi-source data, extract features from the multi-source data, and generate corresponding customer feature tags; The enterprise thesaurus is matched with the customer feature tags, and the matching results are imported into the search engine; Obtain the user's search request and output the corresponding customer view based on the search request.

2. The method according to claim 1, characterized in that, The process of acquiring customer information and business data from each subsidiary, and constructing a target organizational structure tree based on the customer information and business data, includes: Build a full organizational structure tree using data from an offline database via HIVE; Obtain customer information and business data from each subsidiary, and extract customer information and business data from the data of each subsidiary through the multi-relationship search MRS loading tool; The customer information and the business data are stored in a data cache queue; Flink incrementally updates the full organizational structure tree based on the real-time data stream in the data cache queue to obtain the target organizational structure tree, and stores the data in the data cache queue into the offline database.

3. The method according to claim 1, characterized in that, The step of segmenting the data in the target organizational structure tree to generate an enterprise thesaurus includes: Extract basic terms from the target organizational structure tree; Determine the word frequency, part of speech, and hierarchical relationship of the basic terms; A dictionary file is generated based on the word frequency, part-of-speech, and hierarchical relationship of the basic terms; The dictionary file is loaded using a word segmenter to generate an enterprise thesaurus.

4. The method according to claim 1, characterized in that, The process of acquiring external multi-source data and extracting features from the multi-source data to generate corresponding customer feature tags includes: Acquire external multi-source data, and classify and preprocess the external multi-source data to obtain preprocessed data; The preprocessed data is input into a natural language processing engine to perform hierarchical processing and obtain a customer feature view. Based on the customer feature view, customer feature tags are generated according to a preset tag system and mapping rules.

5. The method according to claim 4, characterized in that, The natural language processing engine includes a basic parsing layer, a deep semantic understanding layer, and a feature association and aggregation layer; The step of inputting the preprocessed data into a natural language processing engine for hierarchical processing to obtain a customer feature view includes: The preprocessed data is segmented, part-of-speech tagging is performed, entity recognition is performed, and key information is extracted by the basic parsing layer to obtain basic information elements. The deep semantic understanding layer parses the basic information elements from multiple predefined customer feature dimensions to obtain customer features; The customer feature view is obtained by associating and aggregating the customer features using the customer identifier as the key through the feature association aggregation layer.

6. The method according to claim 1, characterized in that, The method further includes: Retrieve annotation data for user-submitted anomaly tags; The customer feature tags are corrected and updated based on the labeled data.

7. The method according to claim 1, characterized in that, The method further includes: The customer feature tags of the customer are periodically acquired and stored, and a snapshot of the customer profile is formed based on the customer feature tags; Based on the customer profile snapshots arranged in chronological order, construct the customer's historical trajectory data; Obtain the user's historical analysis requests for the target customer, and obtain the target customer's historical trajectory data based on the historical analysis requests; Based on the historical trajectory data of the target customer, the evolution process of the target customer's feature tags within a preset time range is output in the form of a visual chart.

8. A device for acquiring a customer view, characterized in that, The device includes: The module is used to acquire customer information and business data of each subsidiary, and to construct the target organizational structure tree based on the customer information and business data; The first generation module is used to segment the data in the target organizational structure tree into words and generate an enterprise thesaurus. The second generation module is used to acquire external multi-source data, extract features from the multi-source data, and generate corresponding customer feature tags. The matching module is used to match the enterprise thesaurus with the customer feature tags and import the matching results into the search engine; The output module is used to obtain the user's search request and output the corresponding customer view based on the search request.

9. An electronic device, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A computer storage medium, wherein, The computer storage medium stores computer-executable instructions; when executed by a processor, the computer-executable instructions can implement the method described in any one of claims 1-7.

Images