日時: 6月20日(月)3限(13:30-15:00)

会場: L3

司会: 矢田 竣太郎
YAO LEAN FRANZL LIM D, 中間発表 ソーシャル・コンピューティング 荒牧 英治, 中村 哲, 若宮 翔子, 矢田 竣太郎
Title: Quantifying Medical Research Interest: Building a System to Support Research on Rare Diseases
Abstract: Research on rare diseases is difficult because the scarcity of information, data, and support. There is no precise definition of a rare disease, but the common agreement is that it affects a small proportion of the population. Although there are only a few cases for each rare disease, collectively they affect a significant portion of the population. We propose to support the research efforts on rare diseases by improving the utilization of available information on rare diseases and to look at trends in medical research to discover driving factors that can push institutions to pursue these types of research. In this project, we develop a system to visualize trends in medical research and use the output with external data to uncover underlying factors that drive these numbers. Considering that most rare diseases are genetic in origin, we utilize the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes in order to categorize each article according to each disease category (including genetic disorders). We apply a named-entity-recognition (NER) model (F1 = 0.7811) on case reports available on PubMed (N=298,303) in order to categorize each article into to see the distribution of interest a country has for the different disease categories. Our preliminary analysis showed that a country's allocation to research and development is a significant indicator in the amount of research on rare diseases and countries that have a lower level of globalization have significantly lower effects from increasing their allocation.
Language of the Presentation: English
前田 拓哉 M, 2回目発表 ソーシャル・コンピューティング 荒牧 英治, 中村 哲, 若宮 翔子, 矢田 竣太郎
Title: Investigating the Implicit Value Orientations of Language-Based Approaches to Online Hate and Abuse
Abstract: Although NLP research on online abuse is often motivated by humanistic reasons, the implicit theoretical orientations, assumptions, and values of researchers and developers can shape automated detection systems in ways that constrain non-technologists agency and lead to adverse offline consequences. It is therefore important to investigate how technological problems are formulated, how NLP research is framed and reported, and how publications deal with matters of transparency and accountability. In this project, I use a combination of qualitative methods like archival studies and discourse analysis to analyze 153 NLP papers on online abuse that were published between January 2019 and January 2022, focusing on how the authors conceptualize and contextualize the given problem and justify their methods. The discursive trends indicate that humanistic concerns do not necessarily translate into socially-oriented research methodologies, suggesting a need for greater methodological reflection.
Language of the Presentation: English
ZHANG ZHOUQING M, 1回目発表 ソーシャル・コンピューティング 荒牧 英治, 中村 哲, 若宮 翔子, 矢田 竣太郎
title: A study of multicultural contextual discussions about breast cancer topics based on social media platforms
abstract: Nowadays, with a high number of new breast cancer patients diagnosed every year, breast cancer has become a major topic in people's daily life. People from the same culture have different views on such a heavy topic, and people from different cultures also have different views, that may be fearful, pessimistic, positive, normal, etc. Our study explores and learns the perceptions and emotional tendencies of people from two different communities in China and Japan on the topic of breast cancer by extracting breast cancer related tweets from social media platforms (Twitter in Japan and Weibo in China). In order to systematically compare the similarities and differences between the Chinese and Japanese communities on the topic of breast cancer.
language of the presentation: English
清基 英則 M, 1回目発表 ソーシャル・コンピューティング 荒牧 英治, 和田 隆広, 若宮 翔子, 矢田 竣太郎, 松田 裕貴
title: Estimating tweet directivity using linguistic features
abstract: With the Covid-19 pandemic, governments and local authorities are often required to provide accurate and prompt information using social media. For the communication of such information, it is important for social media users themselves to know if they fall within the targeted demographics of these communications (age, gender, etc.), which we label here as ‘directivity’. Previous studies have mainly focused on examining the attributes of the information provider, but to our knowl- edge, there have not been any studies that examine the attributes of targeted users (receivers). In this study, we first assumed that tweets by magazine publishers are crafted for their targeted readership. We then collected tweets from the official Twitter accounts of these magazines and manually labeled the target age and gender of each magazine to create a dataset of tweet directivity. Using this dataset, we then classified the target user demographic (age group and gender) of these tweets through machine learning. We analyzed the results of this experiment and discussed the usefulness of our quantitative estimates of directivity.
language of the presentation: Japanese
発表題目: 言語的特徴を用いたツイートの指向性推定
発表概要: 新型コロナウイルスの拡大に伴い,政府や自治体などはソーシャルメディアを用いた正確かつ迅速な情報 発信を行うことが求められている.正確かつ迅速な情報発信のためには,特定の対象(年代や性別など)に向けて発 信された情報を,その対象が自分に向けて発信されていると理解できるかどうか,すなわち「指向性」が重要である. 情報発信者の属性を特定する研究は多いが,情報が対象とする受信者の属性を特定する研究は見受けられない.本研 究では,Twitter における雑誌の公式アカウントが発信するツイートは読者層向けに最適化されていると考え,雑誌の 公式アカウントによるツイートを収集し,各雑誌の対象年齢と性別をラベル付けした指向性ツイートデータセットを 構築した.このデータセットを用いて,対象ツイートがどの年齢,どの性別に向けられているものなのか機械学習モ デルで分類した.この実験結果を分析し,指向性の定量的測定がもたらす価値を考察した.

会場: L2

司会: 陳 娜
DOUHA N'GUESSAN YVES-ROLAND D, 中間発表 サイバーレジリエンス構成学 門林 雄基, 安本 慶一, 林 優一, 妙中 雄三
Title: Investigating Cost-Benefit of Cybersecurity Awareness Training for Smart-Home Users
Abstract: The human factor is still a crucial issue in the security chain. People who will live in a smart home could be exposed to many cyber threats due to the remaining lack of Internet of Things (IoT) device security. Cybersecurity awareness training could help households to become more resilient to face cyberattacks. However, the financial costs of training programs and the significant amount of time needed to notice security countermeasures could refrain many smart-home users from engaging in cybersecurity education. In recent years, game theory has successfully contributed to modeling and understanding human interaction and decision-making. Game theory provides the mathematical framework necessary to analyze the potential outcomes of given strategy sets. In this research, we propose a game-theoretic approach to analyze the security investment cost-benefit of households. The game model consists of three smart-home users and one attacker. We investigate the pure and mixed Nash equilibria of the proposed game. Our numerical results show that a rational smart-home user would invest in cybersecurity training under certain conditions to maximize his payoff. Moreover, the increase in the quality of services accessible in a smart home and the rewards for noticing security countermeasures compared to the potential impacts of cyberattacks would increase the payoffs of households and reinforce the security behaviors. This research study emphasizes the urgent need to address human security toward a more resilient smart home. Future research will investigate an evolutionary game approach to model more realistic attack scenarios involving different populations of agents related to the smart-home ecosystem.
Language of the presentation: English