Frequently Asked Questions (FAQ)

Taxonic is an automated semantic classification engine built for academic publishers, libraries, bookstores, and EdTech platforms. It converts raw, unstructured text—such as book titles, academic course names, syllabus excerpts, and digital learning assets—into official taxonomy subject codes (such as THEMA, LCSH, and HECoS) in under 0.5 seconds per item.

Manual tagging traditionally takes catalogers and librarians 15 to 45 minutes per record, costing organizations $10 to $25 per title. Taxonic eliminates this bottleneck by automating subject classification at high speed with over 99% cost reduction.

Taxonic is designed for anyone managing bibliographic metadata or educational content catalogs:

• Librarians & Technical Services Leads: Clear cataloging backlogs and generate accurate call numbers and subject headings.
• Publishers & Backlist Managers: Enrich backlist catalogs with modern THEMA/BISAC subject codes to prevent retail feed rejections and boost e-commerce discoverability.
• EdTech & Learning Platforms: Automatically tag digital courseware, syllabi, and assets against higher-education frameworks like HECoS.
• Bookstores & Distributors: Normalize incoming publisher feeds directly into storefront category nodes.

Generic keyword search fails when titles use natural language instead of taxonomy terms (e.g., matching "Pottery Techniques" to "Ceramic Artefacts"). Generic LLMs, on the other hand, frequently hallucinate invalid, out-of-date, or non-existent taxonomy codes.

Taxonic combines local dense vector embeddings (BAAI/bge-m3) with deterministic, ontology-validated taxonomy schemas. It understands the deep contextual meaning of text while guaranteeing that every returned subject code strictly exists within the official taxonomy standard.

Taxonic can classify:

• Book titles and subtitles
• Back-cover blurbs, abstracts, and summaries
• Academic course names and degree module descriptions
• Syllabus excerpts and learning object summaries
• Messy, unformatted metadata strings from legacy CSV/XML feeds

Taxonic operates directly inside Google Sheets via a custom spreadsheet formula, requiring zero software installation or technical training.

1. Open your spreadsheet containing titles and metadata.
2. Enter your assigned API Key in the sidebar or formula settings.
3. Type the custom formula in any empty cell: =TAXONIC(A2, B2, "THEMA") (where A2 is the title and B2 is an optional description).
4. Press Enter. Taxonic populates the official code and definition instantly.

The standard formula syntax is:

Code
=TAXONIC(title, [description], [standard], [code_prefix])

• title (Required): Cell reference containing the book or asset title (e.g., A2).
• description (Optional): Cell reference containing summary or syllabus text (e.g., B2).
• standard (Optional): Taxonomy standard key such as "THEMA", "HECOS", or "LCSH".
• code_prefix (Optional): Restricts results to codes starting with a specific prefix (e.g., "10").

Yes. You can drag the =TAXONIC() formula down across thousands of rows. The Google Sheets add-on handles request batching sequentially under the hood, allowing non-technical catalogers to re-index whole backlists without needing developer resources.

No. Because Taxonic operates directly within standard Google Sheets using familiar spreadsheet syntax, editorial staff, catalogers, and backlist managers can start using it immediately without training.

Taxonic provides a RESTful API with sub-50ms latency. Applications send an HTTPS POST request to /v1/classify with a JSON payload containing item metadata.

Code
POST /v1/classify HTTP/1.1
Host: api.taxonic.ai
Content-Type: application/json
X-Taxonic-API-Key: YOUR_API_KEY

Example Request Payload:

Code
{
  "title": "50 Algorithms of Complex Computations",
  "description": "For Bachelors and Technical Degrees",
  "standard": "hecos",
  "code_prefix": "101",
  "max_results": 2,
  "min_score": 0.5
}

Yes. Taxonic natively supports structured function calling schemas (OpenAI Tools / Claude Tools / LangChain). Autonomous agents can declare taxonic_classify as an external tool to fetch verified, hallucination-free classification codes before producing their final output.

JSON Schema Tool Definition Example:

Code
{
  "type": "function",
  "function": {
    "name": "taxonic_classify",
    "description": "Classifies books, research papers, and course syllabi into standardized subject codes using Taxonic.",
    "parameters": {
      "type": "object",
      "properties": {
        "title": { "type": "string" },
        "description": { "type": "string" },
        "standard": {
          "type": "string",
          "enum": ["hecos", "thema", "lcsh", "bisac", "dewey"]
        }
      },
      "required": ["title", "standard"]
    }
  }
}

Taxonic indexes major global standards across publishing, libraries, and higher education:

• exact Match: Returned when the input text matches an official category title or subject code directly. It receives a confidence score of 1.0.
• semantic Match: Returned when Taxonic uses dense vector indexing to calculate the contextual similarity between input text and category scope definitions. The score reflects Cosine Similarity distance (0.0 to 1.0).

No. The confidence score is a normalized mathematical cosine similarity vector metric between 0.0 and 1.0. A score of 0.60+ indicates high semantic relevance in most taxonomy standards.

Taxonic maintains pre-indexed vector stores for standard releases. When a standard body publishes an update, Taxonic updates backend models on servers seamlessly. Your spreadsheet functions and API calls automatically benefit from updated definitions without requiring codebase changes or manual software updates.

You can obtain an API Key in two ways:

1. Directly from Us: Request an enterprise or pilot key directly through the Taxonic Developer Portal.
2. API Marketplaces: Subscribe through API marketplaces such as RapidAPI, which provides instant key generation, usage tracking, and flexible monthly subscription tiers.

Yes. Taxonic provides a free trial quota allowing you to test classification live in your own Google Sheets or via cURL/Python API calls without any upfront commitment or credit card requirement.

Taxonic is billed based on classification query volume (per item classified). Tiered usage packages (e.g., 10,000 or 100,000 classification calls) are available for both API integration and Google Sheets usage. Enterprise plans offer high-volume bulk rates and dedicated SLA support.

Taxonic returns standard HTTP error codes formatted according to the RFC 9457 Problem Details specification (application/problem+json).

Example RFC 9457 Error Response (404 Not Found):

Code
{
  "type": "urn:taxonic:error:standard_not_found",
  "title": "Standard Not Found",
  "status": 404,
  "error_code": "STANDARD_NOT_FOUND",
  "detail": "Standard 'INVALID_STD' is not indexed. Available standards: HECOS, THEMA, LCSH, BISAC, DEWEY.",
  "instance": "/v1/classify",
  "request_id": "18f0d878b65f4929ac5a96d1294e72d2"
}

When given very short titles (e.g., "Algebra"), Taxonic returns candidates with lower confidence scores across broad sub-disciplines. Passing an optional description, syllabus excerpt, or using code_prefix to constrain the search space significantly increases precision.

For technical support, custom taxonomy indexing requests, or enterprise SLA inquiries, contact our engineering team directly at support@taxonic.ai or reach out through your RapidAPI developer dashboard.