Group, Extract and Aggregate: Summarizing a Large Amount of Finance News for Forex Movement Prediction

Proceedings of the Second Workshop on Economics and Natural Language Processing, pages 41–50 Hong Kong, November 4. c⃝2019 Association for Computational Linguistics 41 Group, Extract and Aggregate: Summarizing a Large Amount of Finance News for Forex Movement Prediction Deli Chen1∗, Shuming Ma1, Keiko Harimoto2, Ruihan Bao2, Qi Su1, Xu Sun1 1MOE Key Lab of Computational Linguistics, School of EECS, Peking University 2Mizuho Securities Co., Ltd. {chendeli,shumingma,sukia,xusun}@pku.edu.cn, {keiko.harimoto,ruihan.bao}@mizuho-sc.com Abstract Incorporating related text information has proven successful in stock market prediction. However, it is a huge challenge to utilize texts in the enormous forex (foreign currency ex- change) market because the associated texts are too redundant. In this work, we propose a BERT-based Hierarchical Aggregation Model to summarize a large amount of finance news to predict forex movement. We firstly group news from different aspects: time, topic and category. Then we extract the most crucial news in each group by the SOTA extractive summarization method. Finally, we conduct interaction between the news and the trade data with attention to predict the forex move- ment. The experimental results show that the category based method performs best among three grouping methods and outperforms all the baselines. Besides, we study the influence of essential news attributes (category and re- gion) by statistical analysis and summarize the influence patterns for different currency pairs. 1 Introduction Deep learning and Natural Language Processing technologies have been widely applied in market prediction tasks (Strauß et al., 2018; Alostad and Davulcu, 2017; Li et al., 2015; Ni et al., 2019), and the market related finance news has proven very useful for the prediction (Ding et al., 2016; Xu and Cohen, 2018). However, the studies of predic- tion in forex market, which is the largest market in the world with the highest daily trading volume, is much less than that in the stock market. Fig- ure 1 shows the average numbers per hour of forex related news. There is a large amount of finance news related to forex trading with different influ- ence, so it is a huge challenge to extract the useful semantic information from news. Most of previ- ous works (Bakhach et al., 2016; Shen and Liang, ∗This work is done when Deli Chen is a intern at Mizuho Securities. 329 234 331 282 0 50 100 150 200 250 300 350 US_EU US_JP US_CN US_GB Figure 1: Average numbers per hour of forex related news from Reuters in 2013-2017. US EU represents news related to US, Europe or both of them. 2016; Pradeepkumar and Ravi, 2016; Contreras et al., 2018; Weeraddana et al., 2018) on forex pre- diction ignore related text totally and focus on the forex trade data only, which loses the important se- mantic information. Yet existing works (Seifollahi and Shajari, 2019; Nassirtoussi et al., 2015) apply- ing finance news in forex prediction mainly rely on manual rules to build feature vectors, which can hardly access the semantic information effectively. To make better use of finance news, we propose a novel neural model: Bert-based Hierarchical Ag- …

Content truncated to 500 words