@article{oai:ipsj.ixsq.nii.ac.jp:02006752,
 author = {西村,龍之介 and 江木,啓訓 and Ryunosuke Nishimura and Hironori Egi},
 issue = {1},
 journal = {情報処理学会論文誌},
 month = {Jan},
 note = {本研究では，発言に含まれる単語の抽象度を用いて，協調的議論における知識収束を定量化する手法を提案する．協調的議論では，個々の参加者が持つ具体的な意見を，全員が合意できる普遍的な意見へと抽象化する知識収束と呼ばれるプロセスが生じる．しかし，既存の知識収束の分析手法は質的な分析に基づいており，具体化や抽象化の程度が十分に定量化されていなかった．そこで，本研究では，発言に含まれる単語の抽象度を定量化する手法を提案した．評価実験では，6グループを対象として協調的議論を実施し，知識収束の過程と抽象度の変化とを比較した．その結果，知識収束の過程で生じる具体化や抽象化は，抽象度の移動平均の変化として現れることが分かった．本研究は，知識収束に関する質的研究と量的研究を架橋するものであり，抽象度を軸とした新たな知識収束の分析手法への展開が期待される．, This study proposes a method for quantifying knowledge convergence in collaborative discussions using the abstractness of words included in participants' utterances. In collaborative discussions, a process known as knowledge convergence occurs, in which participants' concrete individual opinions are abstracted into more universal knowledge that can be accepted by the group as a whole. However, existing approaches to analyzing knowledge convergence have primarily relied on qualitative methods, and the degree of concretization or abstraction has not been sufficiently quantified. To address this gap, we propose a method that quantifies the abstractness of words used in utterances. In an evaluation experiment involving six groups, collaborative discussions were conducted, and the process of knowledge convergence was compared with changes in abstractness. The results showed that the processes of concretization and abstraction during knowledge convergence were reflected as changes in the moving average of abstractness. This study serves as a bridge between qualitative and quantitative research on knowledge convergence, and it offers the potential for developing a new analytical framework based on abstractness.},
 pages = {38--48},
 title = {発言に含まれる単語の抽象度を用いた協調的議論における知識収束の定量化および評価},
 volume = {67},
 year = {2026}
}