Financial News Dataset from Bloomberg and Reuters

450,341 news from Bloomberg and 109,110 news from Reuters.

License

Very sorry to announce that those datasets are no longer available online for research purposes (NLP...) due to copyright issues.

If you have a request about it, send it to me at [email protected] and put the words "bloomberg dataset" in your email body.

Examples

Reuters (109,110 news)

-- Pluspetrol says losing $2.4 mln/day in Peru protest
-- 
-- Sat Oct 21, 2006 8:11pm EDT
-- http://www.reuters.com/article/2006/10/22/businesspro-oil-peru-pluspetrol-dc-idUSN2127888220061022

 

 LIMA, Peru  (Reuters) - Argentine oil company Pluspetrol 
said on Saturday it was losing about $2.4 million a day in 
revenue after suspending operations this week because hundreds 
of indigenous protesters occupied its oil wells. 

  Pluspetrol has said some of its workers were being held 
hostage at its oil fields in the Amazon region of northern Peru 
[...] by the protests.

Bloomberg (450,341 news)

-- Baoshan's 3rd-Quarter Profit Gains on Steel Demand
-- Janet Ong
-- 2006-10-27T13:21:05Z
-- http://www.bloomberg.com/news/2006-10-27/baoshan-s-3rd-quarter-profit-gains-on-steel-demand-update4-.html

          
          
             Baoshan Iron & Steel Co., China's
biggest steelmaker, said its third-quarter net income rose 42
percent and reversed three straight quarters of declines as
demand recovered.  
 Net income rose to 4.7 billion  yuan  ($595.7 million) in the
quarter ended Sept. [...] will be effective from April 1, 2007.  
 To contact the reporter for this story:
Helen Yuan in  Shanghai  at 
 [email protected]   
 To contact the editor responsible for this story:
Keith Gosman at 
 [email protected]  

License

This dataset was compiled and first used in Ding et al. (2014).

• [Ding et al., 2014] Xiao Ding, Yue Zhang, Ting Liu, and Junwen Duan. Using structured events to predict stock price movement: An empirical investigation. In Proc. of EMNLP, pages 1415–1425, Doha, Qatar, October 2014. Association for Computational Linguistics.

Other papers cite this dataset:

• Ding et al. (2015)

Citation

@misc{BloombergReutersDataset2015,
  author = {Philippe Remy, Xiao Ding},
  title = {Financial News Dataset from Bloomberg and Reuters},
  year = {2015},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/philipperemy/financial-news-dataset}},
}