Mithilss/finance-data

--- dataset_info: features: - name: 'Unnamed: 0' dtype: string - name: Date dtype: string - name: Article_title dtype: string - name: Stock_symbol dtype: string - name: Url dtype: string - name: Publisher dtype: string - name: Author dtype: string - name: Article dtype: string - name: Lsa_summary dtype: string - name: Luhn_summary dtype: string - name: Textrank_summary dtype: string - name: Lexrank_summary dtype: string splits: - name: train num_bytes: 24086212809 num_examples: 15549299 download_size: 10485927609 dataset_size: 24086212809 configs: - config_name: default data_files: - split: train path: data/train-* ---This dataset consists of news articles and their associated information such as publish date, article title, stock symbol, source URL, publisher, author, article content, and four different text summaries. It is suitable for text summarization, natural language processing, and other research fields. The training set contains approximately 15.5 million examples.

搜集汇总

背景与挑战

背景概述

该数据集是一个金融新闻数据集，包含约275万条记录，涵盖股票新闻文章及其多种自动生成的摘要（如LSA、Luhn、TextRank、LexRank摘要），适用于金融文本分析和摘要研究。