Skip to content

Latest commit

 

History

History
100 lines (59 loc) · 6.46 KB

File metadata and controls

100 lines (59 loc) · 6.46 KB
graph LR
    Application_Entry_Point_Demo["Application Entry Point / Demo"]
    API_Inference_Orchestrator["API & Inference Orchestrator"]
    Data_Management_Utilities["Data Management & Utilities"]
    Model_Initialization_Configuration["Model Initialization & Configuration"]
    Word_Segmentation_WS_Component["Word Segmentation (WS) Component"]
    Part_of_Speech_POS_Component["Part-of-Speech (POS) Component"]
    Named_Entity_Recognition_NER_Component["Named Entity Recognition (NER) Component"]
    Application_Entry_Point_Demo -- "invokes" --> API_Inference_Orchestrator
    API_Inference_Orchestrator -- "triggers" --> Model_Initialization_Configuration
    Data_Management_Utilities -- "provides resources to" --> Model_Initialization_Configuration
    API_Inference_Orchestrator -- "sends text for segmentation to" --> Word_Segmentation_WS_Component
    Word_Segmentation_WS_Component -- "returns segmented words to" --> API_Inference_Orchestrator
    API_Inference_Orchestrator -- "sends segmented words for POS tagging to" --> Part_of_Speech_POS_Component
    Part_of_Speech_POS_Component -- "returns POS tags to" --> API_Inference_Orchestrator
    API_Inference_Orchestrator -- "sends segmented words for NER to" --> Named_Entity_Recognition_NER_Component
    Named_Entity_Recognition_NER_Component -- "returns named entities to" --> API_Inference_Orchestrator
    click API_Inference_Orchestrator href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/ckiptagger/API_Inference_Orchestrator.md" "Details"
    click Model_Initialization_Configuration href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/ckiptagger/Model_Initialization_Configuration.md" "Details"
    click Word_Segmentation_WS_Component href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/ckiptagger/Word_Segmentation_WS_Component.md" "Details"
    click Part_of_Speech_POS_Component href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/ckiptagger/Part_of_Speech_POS_Component.md" "Details"
    click Named_Entity_Recognition_NER_Component href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/ckiptagger/Named_Entity_Recognition_NER_Component.md" "Details"
Loading

CodeBoardingDemoContact

Details

The ckiptagger project is structured around a clear pipeline for Chinese Natural Language Processing. The Application Entry Point / Demo serves as the user interface, initiating the NLP process by interacting with the API & Inference Orchestrator. This orchestrator, primarily embodied by the CkipTagger class, manages the sequential execution of specialized NLP tasks: Word Segmentation, Part-of-Speech Tagging, and Named Entity Recognition. Before processing, Data Management & Utilities ensures that necessary external data resources are available. Each specialized NLP component (WS, POS, NER) is initialized and configured by the Model Initialization & Configuration module, which sets up their respective deep learning models. Data flows sequentially from raw text input through segmentation, then POS tagging, and finally NER, with intermediate results being passed back to the orchestrator for subsequent processing.

Application Entry Point / Demo

The user-facing interface and demonstration module for interacting with the ckiptagger library.

Related Classes/Methods:

API & Inference Orchestrator [Expand]

The central control unit managing the entire NLP pipeline, handling input/output, and orchestrating calls to individual models.

Related Classes/Methods:

Data Management & Utilities

Handles the acquisition and preparation of external data resources required by the NLP models.

Related Classes/Methods:

Model Initialization & Configuration [Expand]

Manages the setup, loading, and configuration of the deep learning models for each NLP task.

Related Classes/Methods:

Word Segmentation (WS) Component [Expand]

The specialized component responsible for tokenizing raw text into words.

Related Classes/Methods:

Part-of-Speech (POS) Component [Expand]

The specialized component for assigning grammatical tags to segmented words.

Related Classes/Methods:

Named Entity Recognition (NER) Component [Expand]

The specialized component for identifying and classifying named entities within text.

Related Classes/Methods: