information extraction for survey papers

1. Obtain survey papers included in the Web of Science core collection database by limiting search topics, document types, dates, and other methods. The search formula for the topic is: (TS=(DATA MINING) OR TS=(LINE LEARNING)) AND DT=("VIEW") AND PY=(2015-2023) AND (WC=Computer Science, Artistic Intelligence OR WC=Computer Science, Theory&Methods OR WC=Computer Science, Information Systems).2. The supervised learning training dataset includes 390 segments of text, corresponding to 2454 question and answer samples; The test dataset includes 379 segments of text, corresponding to 2337 question and answer samples.3. Reference paper on annotation element types and relationship types: Research on Semantic Information Extraction Methods for Survey Papers