Entity-oriented Investment Big Data Analysis Foundation for Evidence-Based Investment Support

2022 Fiscal Year Final Research Report

• Project/Area Number: 19H04116
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 60080:Database-related
• Research Institution: Kyoto University
• Principal Investigator: Ma Qiang 京都大学, 情報学研究科, 准教授 (30415856)
• Co-Investigator(Kenkyū-buntansha): 湯本 高行 兵庫県立大学, 社会情報科学部, 准教授 (20453152)
• Project Period (FY): 2019-04-01 – 2023-03-31
• Keywords: 投資情報学 / 意思決定支援 / 要因分析 / 時系列データ / 深層学習 / 金融工学 / ポートフォリオ

Outline of Final Research Achievements

In this study, we conducted research and development of the basic technology to support investment activities.　As the analysis foundation, we developed representation models and analysis methods that efficiently manage and process time series data.　For factors analysis, we introduced the trend change point of the benchmark price as an interference factor and developed the factor analysis method based on the dynamic state space model. We also developed a method to estimate the economic impact of news articles, which are an important source of information for investment. For investors analysis, we introduced portfolio theory and developed the feature analysis technology of traders, who are investors in social trading services.
We published four journal articles in English, six reviewed international conference papers, and 11 domestic conference papers.

Free Research Field

データ工学

Academic Significance and Societal Importance of the Research Achievements

本研究では，ニュース・報告書，マーケット情報やユーザの投資履歴など多様な投資ビッグデータを横断して分析し，「金融・投資商品に影響を及ぼす要因(事象)」や「投資のエキスパートがいつ・何を取引したか」を明らかにする．これにより投資に必要な知識やエビデンスを発見して意思決定のプロセスと結果の可読性を向上させることで，従来サービスにおける一般投資者の不安を緩和し，安心して投資できる仕組みの構築に貢献する．