CCTI Manual(Eng)

Educational Model for Classical Interpretation in the AI Era
CCTI
Classical Chinese Text Interpreter

Program Description and User Manual
This manual serves as a guide to understanding and utilizing CCTI as an "AI-collaborative system for Classical Chinese interpretation education."

Hyeon Kim Digital Curator

Ⅰ. System Introduction

1. System Overview

CCTI (Classical Chinese Text Interpreter) is an educational and research system designed to provide step-by-step assistance in the interpretation process of Classical Chinese texts utilizing Artificial Intelligence (LLM APIs).

Rather than acting as a mere automated translation tool for Classical Chinese source texts, CCTI is engineered to let learners collaborate with AI across the entire trajectory: from source text, punctuation, and glossary annotations to translation and comprehensive commentary.

Through this framework, users do not passively accept the interpretations proposed by AI. Instead, they review, emend, and refine the outputs of each stage to arrive at a more precise and evidence-based interpretation.

CCTI is characterized by the following features:

• Consistent data management centered around a text unit (`text_unit`).
• A step-by-step structure that clearly distinguishes AI-generated outputs from human-reviewed results.
• An XML-based structured data exchange methodology.
• Integration with MediaWiki to facilitate the sharing of interpretation results and collaborative learning.

CCTI can be applied not only to Classical Chinese, classics, and East Asian studies at the university and higher education levels, but also to collaborative interpretation projects among researchers and digital humanities practicum environments.

2. Design Intent and Scope of Purpose

1) Design Intent

The core guiding philosophy behind CCTI is to utilize AI as an interpretive instrument while firmly positioning the "AI as a collaborative assistant rather than the primary subject of interpretation." To achieve this, the system incorporates the following design principles:

• The AI is responsible for generating punctuated texts, drafting glossaries, and producing initial translation drafts.
• The learner is responsible for reviewing AI outputs, correcting errors, evaluating interpretive validity, and finalizing the ultimate entries.

In short, the objective of CCTI is not to provide immediate answers rapidly, but to train the cognitive thought process required to achieve an accurate interpretation.

2) Scope of Purpose

CCTI can be implemented for the following objectives:

• A step-by-step training tool in Classical Chinese interpretation education.
• A platform for sharing and comparing interpretation results among students, instructors, and researchers.
• A repository of structured interpretation data for subsequent research (serving as foundational data for advanced humanistic inquiries).
• Production of benchmark datasets for training domain-specific AI models for Eastern Classics (Vertical Models).
• An infrastructural model for establishing AI-driven digital humanities practicum ecosystems.

3. Anticipated Pedagogical Outcomes

Transcending its role as a simple technological utility, CCTI aims to expand the pedagogy of classical text interpretation itself. The anticipated educational outcomes are as follows:

1) Establishment of Process-Oriented Learning

Students are trained to focus on the 'validity of punctuation', 'precise semantic definition of terminology', and 'contextual variances in interpretation' rather than just looking at the final translated product.

☞ Shifting focus from the "product" of interpretation to the "logical process" of interpretation.

2) Enhancing Critical AI Literacy

Learners develop a scholarly disposition to interrogate AI-generated content—reviewing, modifying, and demanding structural evidence rather than accepting machine outputs unconditionally.

☞ Cultivating "Critical AI Literacy," a paramount competence in the artificial intelligence era, through organic praxis.

3) Cultivating Collaborative Learning Environments

Through its integration with a Wiki platform, this system enables individual interpretation results to be shared online, allowing peers to review and mutually advance their interpretations.

☞ Expanding the traditional physical "reading room (강독실)" environment into a digital collaborative workspace.

4) Advancing Digital Humanities Competencies

Classical Chinese education using this system empowers learners to simultaneously cultivate a complex matrix of competencies:

• Classical text interpretation proficiency.
• Comprehension of structured data formats (XML).
• AI prompt and utilization capabilities.
• Knowledge sharing and collaborative experiences.

☞ Forging the bedrock of Digital Humanities capabilities, which are becoming increasingly vital in contemporary humanistic research.

CCTI is an AI-driven educational support tool designed to shift Classical Chinese interpretation from 'product-centered learning' to 'process-oriented, critical, and collaborative learning.'

Ⅱ. System Configuration and Workflow

CCTI isolates the entire workflow of classical interpretation into step-by-step modules. While each module executes an independent function, they are organically intertwined within a single unified workflow.

1. Structural Hierarchy of the System

The CCTI system is organized into the following three layers:

1) Authentication and Session Management Layer

• User login.
• Database / Project selection.

2) Interpretation Process Management Layer

• Text unit list management.
• Visualization of interpretation stage progress states.
• Navigation between workflow phases.

3) Interpretation Workspace Layer

• Source text input and metadata editing.
• AI-driven generation of punctuation and glossaries.
• Review and emendation interface for learners.
• Translation generation reflecting human contributions.
• Result sharing and publishing (Wiki).

2. Functional Module Configuration

Each phase of the interpretation workflow within the CCTI system is modularized into independent program units. The configuration of modules by function is as follows:

1) Login

• Authentication module for securing system access.
• Operates via database access accounts (account/password).

2) Interpretation Manager

• The central dashboard of the system.
• Displays the registry of registered text units.
• Visualizes the operational progress state of each text unit.
• Directs navigation to source text, punctuation, glossary, translation, and Wiki screens.

3) Insert New Text

• Initial registration module for the classical text unit slated for interpretation.
• Unique ID verification.
• Automated generation of base XML templates.
• Saves the raw source text (`source_text`) and corresponding metadata.

4) Update Source Text

• Modifies the source text and metadata of previously registered units.
• Refines the Classical Chinese source while preserving the structural integrity of the XML.
• Manages the baseline data for subsequent AI processing.

5) Edit Punctuated Hanmun

• Generates standard punctuated Classical Chinese text using AI.
• Provides an interface to review and emend the punctuated text version (`hanmun_punctuated`).
• Interlocks directly with the preliminary glossary draft generation.

6) Edit Glossary

• Populates an AI-generated preliminary glossary draft (`glossary_draft`).
• Facilitates the composition of the human-reviewed version (`glossary_reviewed`).
• Organizes interpretive evidence including personal names, geographic locations, core concepts, and grammatical structures.

7) Edit Translation

• Generates an AI translation synthesized from the punctuated text and the reviewed glossary.
• Interface for reviewing the initial translation draft (`translation_draft`).
• Finalizes and commits the verified translation product (`translation_reviewed`).

8) Result Sharing (Wiki)

• Publishes interpretation outcomes as Wiki articles.
• Consolidates source text, glossaries, translations, commentaries, and contextual relation networks into a singular document.
• Supports collaborative analysis and open-access learning.

3. Execution Steps for Interpretation Tasks

1) System Access and Initialization

• Navigate to the CCTI access URL.
• Verify the target server, database, and project parameters on the login screen.
• Authenticate utilizing curator credentials (account/password).

2) Text Registration

• Access the "Insert New Text" module from the Interpretation Manager.
• Input the unique Text Unit ID and execute a duplication check.
• Populate the source XML and commit to save.

3) Sequential Workflow Execution

• Complement the raw text via the Source Text Editor.
• Execute and verify AI punctuation within the Punctuation Editor.
• Validate terms, concepts, and grammar structures via the Glossary Editor.
• Trigger AI translation compilation within the Translation Generator.
• Critically review and edit the draft to establish the final verified translation.

4) Sharing and Collaborative Review

• Sync and publish the finalized interpretation outputs onto the Wiki.
• Compare, contrast, and debate interpretation variants with peer learners and instructors.
• Maintain the flexibility to revert to any specific preceding stage for re-editing if necessary.

※ Core Operational Principles

• AI provides the initial structural draft.
• The learner remains the sovereign agent of judgment and emendation.
• The entirety of the interpretive process is preserved as structured data.

CCTI is an AI-era pedagogical utility engineered to systematically record, track, and educate users across the entire lifecycle of a single Classical Chinese text unit: spanning from ingestion and AI drafting to human critique and public publishing.

Ⅲ. Data Structure and Formats

CCTI records all data ingested, generated, and edited during the classical text interpretation workflow into a relational database.

The underlying data schema is engineered around a single classical Text Unit, progressively accumulating and mapping both AI-generated data sets and human-reviewed data entries within a unified tabular architecture.

The repository of classical interpretation data amassed through CCTI operations will serve as vital historical datasets for the future compilation of specialized Classical Chinese dictionaries and usage lexicons, advanced computational text analysis, and the optimization of domain-specific AI models for Eastern Classics (Vertical Models).

CCTI records all data ingested, generated, and edited during the classical text interpretation workflow into a relational database.
This structural data will actively contribute to the future engineering of domain-specific AI models for Eastern Classics (Vertical Models).

☞ Detailed Specifications of CCTI Data Structure and Formats