Classify web content using LLMs — classify • cat.web

Wraps the Python catweb.classify() function. Accepts URLs (auto-fetched to text) or raw text strings. Injects web context (source domain, content type, metadata) into the classification prompt.

Usage

classify(
  categories,
  input_data = NULL,
  api_key = NULL,
  source_domain = NULL,
  content_type = NULL,
  web_metadata = NULL,
  description = "",
  filename = NULL,
  save_directory = NULL,
  timeout = 30L,
  user_model = "gpt-4o",
  mode = "image",
  creativity = NULL,
  safety = FALSE,
  chain_of_verification = FALSE,
  chain_of_thought = FALSE,
  step_back_prompt = FALSE,
  context_prompt = FALSE,
  thinking_budget = 0L,
  example1 = NULL,
  example2 = NULL,
  example3 = NULL,
  example4 = NULL,
  example5 = NULL,
  example6 = NULL,
  model_source = "auto",
  max_categories = 12L,
  categories_per_chunk = 10L,
  divisions = 10L,
  research_question = NULL,
  models = NULL,
  consensus_threshold = "unanimous",
  use_json_schema = TRUE,
  max_workers = NULL,
  fail_strategy = "partial",
  max_retries = 5L,
  batch_retries = 2L,
  retry_delay = 1,
  row_delay = 0,
  pdf_dpi = 150L,
  auto_download = FALSE,
  add_other = "prompt",
  check_verbosity = TRUE
)

Arguments

categories: A character vector of category names.
input_data: A character vector / list / data.frame column of URLs or text strings. Default NULL.
api_key: Character or NULL. API key for the LLM provider.
source_domain: Character or NULL. Source domain injected into the prompt as context (e.g. "nytimes.com").
content_type: Character or NULL. Content type (e.g. "news article", "blog post").
web_metadata: Named list or NULL. Additional metadata injected into the prompt.
description: Character. Context description. Default "".
filename: Character or NULL. Output CSV filename.
save_directory: Character or NULL. Output directory.
timeout: Integer. URL fetch timeout (seconds). Default 30L.
user_model: Character. Model name. Default "gpt-4o".
mode: Character. Processing mode. Default "image".
creativity: Numeric or NULL. Temperature. Default NULL.
safety: Logical. Default FALSE.
chain_of_verification: Logical. Default FALSE.
chain_of_thought: Logical. Default FALSE.
step_back_prompt: Logical. Default FALSE.
context_prompt: Logical. Default FALSE.
thinking_budget: Integer. Default 0L.
example1, example2, example3, example4, example5, example6: Optional few-shot examples.
model_source: Character. Default "auto".
max_categories: Integer. Default 12L.
categories_per_chunk: Integer. Default 10L.
divisions: Integer. Default 10L.
research_question: Character or NULL.
models: List of model specs for ensemble mode. Default NULL.
consensus_threshold: Character or numeric. Default "unanimous".
use_json_schema: Logical. Default TRUE.
max_workers: Integer or NULL. Default NULL.
fail_strategy: Character. Default "partial".
max_retries: Integer. Default 5L.
batch_retries: Integer. Default 2L.
retry_delay: Numeric. Default 1.0.
row_delay: Numeric. Default 0.0.
pdf_dpi: Integer. Default 150L.
auto_download: Logical. Default FALSE.
add_other: Logical or "prompt". Default "prompt".
check_verbosity: Logical. Default TRUE.

Value

A data.frame with classification results.

Examples

if (FALSE) { # \dontrun{
# Classify a list of URLs (auto-fetched to text)
results <- classify(
  categories    = c("News", "Opinion", "Tutorial"),
  input_data    = c("https://example.com/article-1",
                    "https://example.com/article-2"),
  source_domain = "example.com",
  content_type  = "blog post",
  api_key       = Sys.getenv("OPENAI_API_KEY"),
  user_model    = "gpt-4o-mini"
)

# Or classify raw text (no fetching)
results <- classify(
  categories = c("News", "Opinion", "Tutorial"),
  input_data = df$article_text,
  api_key    = Sys.getenv("OPENAI_API_KEY")
)
} # }