IEICE Technical Report

Natural Language Understanding and Models of Communication

Workshop Date : 2026-09-14 - 2026-09-16 / Issue Date :

NLC2026-1
A Response Strategy in Dialogue Systems for Improving User Satisfaction
Nakamura Hikaru, Suzuki Yu (Gifu Univ)
pp. 1 - 6

NLC2026-2
Supporting Meeting Minutes Creation Using Whisper and a Local LLM -- Improving Proper Noun Recognition Accuracy in Whisper through Fine-Tuning --
Hayato Atarashi, Osamu Shirakawa, Yoshitaka Hayashi (Canon IT Solutions)
pp. 7 - 12

NLC2026-3
Design and Evaluation of a Shared Task for Question Answering over Agricultural Management Indicators
Akio Kobayashi, Kenta Baba, Masahiro Ohtomo, Junichi Ishihara, Tetsuo Katsuragi (NARO), Hiroki Sakaji (Hokudai), Yasutomo Kimura (OUC)
pp. 13 - 18

NLC2026-4
Collection of Region-Specific Agricultural Technical Information and Its Utilization through Generative AI -- Three Years of the BRIDGE AI Social Implementation in Agriculture Project --
Tetsuo Katsuragi, Akio Kobayashi, Junichi Ishihara, Masahiro Ohtomo, Kenta Baba (NARO), Hiroki Sakaji (Hokudai), Yasutomo Kimura (OUC)
pp. 19 - 22

NLC2026-5
Extraction of Expressions Indicating Corporate Maturity from Securities Reports
Aozora Takeshita (Seikei Univ.), Kaito Takano (Nomura Asset Management), Kei Nishina, Hiroyuki Sakai (Seikei Univ.)
pp. 23 - 28

NLC2026-6
Construction of a Dataset and Proposal of a Method for Identifying Corresponding Paragraphs in Annual Securities Reports Across Years
Mizuki Okada, Masahiro Suzuki (Amova Asset Management)
pp. 29 - 34

NLC2026-7
Performance Evaluation of Text Classification Models under Economic Shocks Using the Economy Watchers Survey -- A Comparison between Recent Training Data and Data from Past Shocks --
Masahiro Suzuki (Amova Asset)
pp. 35 - 40

NLC2026-8
The Residual Stream Is Enough: Vocabulary-Space Trajectories for Hallucination Detection
Seongmin Lee, Naoki Yoshinaga (UTokyo/NII LLMC)
pp. 41 - 46

NLC2026-9
Causal Classification of LLMs via Hidden State Injection
Hiroki Sakaji (Hokudai)
pp. 47 - 52

NLC2026-10
Assessing LLMs' Metaphor Comprehension through Impression Rating Prediction
Kato Masaki, Kanako Komiya (TUAT), Kato Sachi (HU), Uno Ryoko (TUAT), Asahara Masayuki (NINJAL)
pp. 53 - 58

NLC2026-11
Comparing Lexical Commonalities Reflected in Document Embedding Spaces: Toward Selecting Document Organization Methods for Different Analytical Purposes
Ryosei Kawaguchi, Shin Hirayama, Haruhiro Katayose (Kwansei Gakuin Univ.)
pp. 59 - 64

Note: Each article is a technical report without peer review, and its polished version will be published elsewhere.

The Institute of Electronics, Information and Communication Engineers (IEICE), Japan