Stock price prediction using financial news articles

Abstract

Stock price prediction is one of the most important issues to be investigated in academic and financial researches. Data mining techniques are frequently involved in the studies aimed to achieve this problem. In this paper we investigate predicting stock prices using financial news articles. A prediction model, finding and analyzing correlation between contents of news articles and stock prices and then making predictions for future prices, was developed. We retrieve financial news articles published in last year, and we get stock prices for same period. All articles are labeled positive or negative according to their effects on stock price. So we use price changes to label the articles. While analyzing textual data, we use word couples consisting of a noun and a verb as features instead of using single words. Afterwards, support vector machines classifier is trained with labeled train articles. Finally, classes of test articles are predicted with using the model resulted from train phase. We achieve serious success rates that prove predictive power of our system.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 48 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.