MCPcopy Create free account
hub / github.com/917Dhj/DeepPaperNote / extract_dataset_candidates

Function extract_dataset_candidates

scripts/common.py:2205–2220  ·  view source on GitHub ↗
(text: str)

Source from the content-addressed store, hash-verified

2203
2204
2205def extract_dataset_candidates(text: str) -> list[str]:
2206 found: list[str] = []
2207 seen = set()
2208 for sentence in split_sentences(text):
2209 if not any(token in sentence.lower() for token in ["dataset", "benchmark", "corpus", "participants", "patients"]):
2210 continue
2211 candidates = re.findall(r"\b[A-Z][A-Za-z0-9+\-]{2,}(?:[ -][A-Z][A-Za-z0-9+\-]{2,})?\b", sentence)
2212 for candidate in candidates:
2213 norm = candidate.lower()
2214 if norm in seen:
2215 continue
2216 seen.add(norm)
2217 found.append(candidate)
2218 if len(found) >= 8:
2219 return found
2220 return found
2221
2222
2223def extract_metric_claims(text: str) -> list[str]:

Callers 1

mainFunction · 0.90

Calls 1

split_sentencesFunction · 0.85

Tested by

no test coverage detected