Semantic feature representation to capture news impact

Boyi Xie, Dingquan Wang, Rebecca J. Passonneau

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

This paper presents a study where semantic frames are used to mine financial news so as to quantify the impact of news on the stock market. We represent news documents in a novel semantic tree structure and use tree kernel support vector machines to predict the change of stock price. We achieve an efficient computation through linearization of tree kernels. In addition to two binary classification tasks, we rank news items according to their probability to affect change of price using two ranking methods that require vector space features. We evaluate our rankings based on receiver operating characteristic curves and analyze the predictive power of our semantic features. For both learning tasks, the proposed semantic features provide superior results.

Original languageEnglish (US)
Pages231-236
Number of pages6
StatePublished - Jan 1 2014
Event27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014 - Pensacola, United States
Duration: May 21 2014May 23 2014

Other

Other27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014
CountryUnited States
CityPensacola
Period5/21/145/23/14

Fingerprint

Semantics
Trees (mathematics)
Vector spaces
Linearization
Support vector machines

All Science Journal Classification (ASJC) codes

  • Computer Science Applications

Cite this

Xie, B., Wang, D., & Passonneau, R. J. (2014). Semantic feature representation to capture news impact. 231-236. Paper presented at 27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014, Pensacola, United States.
Xie, Boyi ; Wang, Dingquan ; Passonneau, Rebecca J. / Semantic feature representation to capture news impact. Paper presented at 27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014, Pensacola, United States.6 p.
@conference{9d9f5f8322a24cceba4354bf97c2d1f8,
title = "Semantic feature representation to capture news impact",
abstract = "This paper presents a study where semantic frames are used to mine financial news so as to quantify the impact of news on the stock market. We represent news documents in a novel semantic tree structure and use tree kernel support vector machines to predict the change of stock price. We achieve an efficient computation through linearization of tree kernels. In addition to two binary classification tasks, we rank news items according to their probability to affect change of price using two ranking methods that require vector space features. We evaluate our rankings based on receiver operating characteristic curves and analyze the predictive power of our semantic features. For both learning tasks, the proposed semantic features provide superior results.",
author = "Boyi Xie and Dingquan Wang and Passonneau, {Rebecca J.}",
year = "2014",
month = "1",
day = "1",
language = "English (US)",
pages = "231--236",
note = "27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014 ; Conference date: 21-05-2014 Through 23-05-2014",

}

Xie, B, Wang, D & Passonneau, RJ 2014, 'Semantic feature representation to capture news impact' Paper presented at 27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014, Pensacola, United States, 5/21/14 - 5/23/14, pp. 231-236.

Semantic feature representation to capture news impact. / Xie, Boyi; Wang, Dingquan; Passonneau, Rebecca J.

2014. 231-236 Paper presented at 27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014, Pensacola, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Semantic feature representation to capture news impact

AU - Xie, Boyi

AU - Wang, Dingquan

AU - Passonneau, Rebecca J.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - This paper presents a study where semantic frames are used to mine financial news so as to quantify the impact of news on the stock market. We represent news documents in a novel semantic tree structure and use tree kernel support vector machines to predict the change of stock price. We achieve an efficient computation through linearization of tree kernels. In addition to two binary classification tasks, we rank news items according to their probability to affect change of price using two ranking methods that require vector space features. We evaluate our rankings based on receiver operating characteristic curves and analyze the predictive power of our semantic features. For both learning tasks, the proposed semantic features provide superior results.

AB - This paper presents a study where semantic frames are used to mine financial news so as to quantify the impact of news on the stock market. We represent news documents in a novel semantic tree structure and use tree kernel support vector machines to predict the change of stock price. We achieve an efficient computation through linearization of tree kernels. In addition to two binary classification tasks, we rank news items according to their probability to affect change of price using two ranking methods that require vector space features. We evaluate our rankings based on receiver operating characteristic curves and analyze the predictive power of our semantic features. For both learning tasks, the proposed semantic features provide superior results.

UR - http://www.scopus.com/inward/record.url?scp=84923916087&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923916087&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84923916087

SP - 231

EP - 236

ER -

Xie B, Wang D, Passonneau RJ. Semantic feature representation to capture news impact. 2014. Paper presented at 27th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2014, Pensacola, United States.