受賞区分：国内学会・会議・シンポジウム等の賞 

担当区分：筆頭著者,　最終著者,　責任著者   掲載種別：研究論文（学術雑誌）   出版者・発行元：Decision Analytics Journal  

In behavioral economics, sentiments influence decision-making processes, with positive sentiments tending to underestimate risks and negative sentiments overestimate them. At the individual level, these sentiments shape economic behavior in ways that collectively influence broader economic dynamics. With the proliferation of textual data and advancements in natural language processing techniques, the analysis of economic sentiments has garnered growing attention, with large language models showing superior analytical capabilities. However, unlike many machine learning tasks where true labels are available through human annotation, sentiment analysis encounters challenges in obtaining true labels due to the psychological biases and inconsistencies inherent in human assessments. To address this issue, this study introduced a data filtering methodology to enhance data reliability and developed a robust sentiment analysis model tailored to the Japanese economy. The findings revealed that our model not only outperforms existing models in terms of generalization capability across diverse datasets — achieving RMSE values of 0.09–0.11 and classification accuracies of 0.83–0.88 — but also effectively captures fluctuations in other quantitative economic indicators, as evidenced by Euclidean distances of up to 1.56, which is smaller than the records 4.33 and 4.24 of the existing models. Moreover, a statistically significant correlation between qualitative and quantitative economic indicators was identified, highlighting the potential of qualitative indicators in predicting economic conditions.

DOI： 10.1016/j.dajour.2025.100629

Open Access

Scopus

researchmap

担当区分：筆頭著者,　最終著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：Digital Scholarship in the Humanities  

Recognizing vocabulary as the essential language foundation and determining its richness continues to be a challenge in quantitative linguistics. This issue has also been widely applied in various practical contexts, including language acquisition, language change, psychology, and cognitive science. However, vocabulary is complex, and lexical richness has been typically measured using three distinct indicators: lexical diversity, density, and sophistication. Among them, lexical diversity has been widely studied, yet reliable metrics independent of text length remain lacking. The characteristics of lexical density and sophistication have not been dealt with in depth, and studies focusing on the Japanese language are limited. This study aimed to delineate the distinctive features of lexical diversity, density, and sophistication, and to explore their interrelationships within Japanese. A corpus of formal policy speeches delivered by Japanese prime ministers was utilized. Numerical experiments were conducted in two stages: first, lexical diversity, density, and sophistication were analyzed separately; subsequently, the impact of text length on these indicators was examined. The findings indicated that text length did not affect lexical density and sophistication, and no consistent relationship was observed among the three indicators.

DOI： 10.1093/llc/fqaf023

Open Access

Web of Science

Scopus

researchmap

担当区分：筆頭著者,　最終著者,　責任著者   掲載種別：研究論文（学術雑誌）   出版者・発行元：Decision Analytics Journal  

Machine learning algorithms are typically evaluated using benchmark datasets under the assumption that these datasets are clean. However, recent studies have revealed the presence of label noise in many benchmark datasets, indicating a biased evaluation to date. Confident learning (CL), an emerging method for noise detection, has been regarded a higher priority than the development of new learning algorithms. Although CL is promoted as applicable to various types of data, existing research has largely concentrated on its application to large-scale datasets. Given that many domains handle datasets of more modest size, this study proposed a shadow-based framework for label noise detection called ShadowN, and conducted a comprehensive comparison with CL using six smaller datasets. Four key aspects were examined: the number of detected noises, the distribution of assigned noise scores, the improvement in classification accuracy, and the accuracy of noise detection with artificial noise injection. The results indicated that ShadowN achieved the highest overall classification accuracy and demonstrated superior precision and F-score across all noise levels. While the current implementation of ShadowN is limited to binary classification, our findings underscore its practical value and demonstrate its potential for enhancing data quality in real-world machine learning workflows.

DOI： 10.1016/j.dajour.2025.100588

Open Access

Scopus

researchmap

担当区分：筆頭著者,　最終著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：SN Computer Science  

Feature selection plays a critical role in machine learning, yet existing methods primarily focus on generating “feature rankings" or “valid feature sets” without providing insights into the distinct roles of different features. This study proposes a feature selection method that categorizes features into four categories: relevant, interaction, redundant, and irrelevant. By incorporating feature subcategorization, the method simultaneously satisfies the requirements for both minimal-optimal features and all-relevant features, which are crucial for real-world applications. The experiments used synthetic data with predefined feature categories and real-world data. Evaluation was conducted based on two metrics: the first measured the percentage of relevant features selected, and the second assessed whether the importance level of each feature was accurately determined. Based on the experimental findings, the proposed method demonstrated greater precision in assessing the feature ranks and facilitated the creation of a more accurate learning model. Moreover, it enhanced the interpretability and utility of the feature selection results.

DOI： 10.1007/s42979-025-03982-7

Scopus

researchmap

その他リンク： https://link.springer.com/article/10.1007/s42979-025-03982-7/fulltext.html

担当区分：筆頭著者,　最終著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Proceedings 2025 5th Asia Conference on Information Engineering Acie 2025  

Phuket, Thailand, January 10-12

DOI： 10.1109/ACIE64499.2025.00028

Scopus

researchmap

総ページ数：368p   記述言語：日本語

CiNii Research

researchmap

開催年月日： 2025年9月

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2025年7月

記述言語：英語   会議種別：ポスター発表  

researchmap

開催年月日： 2025年3月

記述言語：英語   会議種別：口頭発表（一般）  

開催地：Bergamo, Italy   国名：イタリア共和国  

researchmap

開催年月日： 2024年9月

記述言語：日本語  

researchmap

開催年月日： 2024年8月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

種別：査読等 

種別：査読等 

種別：査読等 

種別：査読等 

種別：査読等