Committee |
Date Time |
Place |
Paper Title / Authors |
Abstract |
Paper # |
NLC |
2023-09-06 13:20 |
Osaka |
Osaka Metropolitan University. Nakamozu Campus. (Primary: On-site, Secondary: Online) |
LoRA Tuning Conversational Japanese Large Language Models using Japanese Instruction Dataset Masahiro Suzuki, Masanori Hirano, Hiroki Sakaji (UTokyo) NLC2023-1 |
In this study, we performed LoRA tuning on large language models (LLM) based on both Japanese and English using Japanese... [more] |
NLC2023-1 pp.1-6 |
NLC |
2021-09-16 09:10 |
Online |
Online |
An Attempt to Extraction of Cause-Effect Expressions by Sequence-Labeling Hiroki Sakaji, Kiyoshi Izumi (The Univ. of Tokyo), Atsuo Kato (DIR), Shintaro Nagao (DAM) NLC2021-6 |
In this research, we propose a method for extracting cause-effect expressions using the general language model and clue ... [more] |
NLC2021-6 pp.1-4 |
NLC |
2021-09-16 14:35 |
Online |
Online |
Performance Validation of Pre-Trained BERT in the Financial Domain Masahiro Suzuki, Hiroki Sakaji, Masanori Hirano, Kiyoashi Izumi (The Univ. of Tokyo) NLC2021-12 |
Recently, general-purpose language models pre-trained on large corpora such as BERT have been widely used. In Japanese, ... [more] |
NLC2021-12 pp.26-29 |
AI |
2021-02-12 13:10 |
Online |
Online |
Calculating Similarities of Overviews on Data Exchange Platform Hiroki Sakaji, Teruaki Hayashi (U Tokyo), Yoshiaki Fukami (Keio Univ.), Kiyoshi Izumi (U Tokyo) AI2020-28 |
[more] |
AI2020-28 pp.33-36 |
AI |
2021-02-12 14:00 |
Online |
Online |
development of simulation model towards reproducing characteristics of data exchange markets Hiroyasu Matsushima (Shiga Univ.), Teruaki Hayashi, Hiroki Sakaji (Univ. Tokyo) AI2020-30 |
The institutional design of the data distribution market, where data is traded between different businesses, is consider... [more] |
AI2020-30 pp.41-43 |
NLC |
2020-09-10 17:30 |
Online |
Online |
Verification of Data Similarity on Data Distribution Platforms using Metadata Hiroki Sakaji, Teruaki Hayashi (UT) NLC2020-12 |
[more] |
NLC2020-12 pp.57-60 |
AI |
2019-02-23 14:40 |
Tokyo |
|
Estimation of Tags using Various Data for Online Video Sharing Sites Hiroki Sakaji (U-Tokyo), Akio Kobayashi (AIP), Masaki Kohana (Ibaraki Univ.), Yasunao Takano (Aoyama Gakuin Univ.), Kiyoshi Izumi (U-Tokyo) AI2018-50 |
[more] |
AI2018-50 pp.71-75 |
AI, JSAI-KBS, JSAI-DOCMAS, JSAI-SAI, IPSJ-ICS |
2018-03-02 16:00 |
Hokkaido |
|
A Text Analysis Platform for Extracting Domain-Specific Logical Relations Hiroki Sakaji, Kiyoshi Izumi (Tokyo Univ.) AI2017-46 |
In recent years, it is expected to apply artificial intelligence field methods and technologies to various scenes in the... [more] |
AI2017-46 pp.21-24 |
NLC |
2017-09-08 10:00 |
Tokyo |
Seikei University |
An attempt to extract bill lists for collecting bills of local councils Takuma Tanaka (TUT), Hiroki Sakaji (The Univ of Tokyo), Akio Kobayashi (AIP), Yasutomo Kimura (Otaru Univ of Commerce), Shigeru Masuyama (TUT) NLC2017-23 |
Abstract In councils of local governments (local council for short), bills are submitted by its head, committees,and mem... [more] |
NLC2017-23 pp.41-46 |
NLC |
2017-09-08 17:10 |
Tokyo |
Seikei University |
Determination of unexpectedness of cause-result expressions extracted from summary of financial statements Risa Murono, Hiroyuki Sakai (Seikei Univ.), Hiroki Sakaji (The Univ. of Tokyo), Jason Bennett (SMAM) NLC2017-34 |
In this research, we propose a method to automatically determine whether cause-result expressions extracted from summary... [more] |
NLC2017-34 pp.93-98 |
NLC |
2016-09-08 13:00 |
Tokyo |
Convention Room AP Shibuya-Dogenzaka (Tokyo) |
Extraction of New Relevant Companies by Using Extracted Common Elements from Relevant Companies Sets Miryu Tanaka, Hiroyuki Sakai, Hiroki Sakaji (Seikei Univ.) NLC2016-16 |
In this paper, we propose a method that extracts new relevant companies by using common elements estimated from relevant... [more] |
NLC2016-16 pp.19-24 |
NLC |
2016-02-04 14:40 |
Kyoto |
Campus Plaza Kyoto |
Extraction of Characteristic Comments from Nico Nico Douga Shinsaku Yoshioka, Hiroki Sakaji, Hiroyuki Sakai (Seikei Univ.) NLC2015-44 |
(Advance abstract in Japanese is available) [more] |
NLC2015-44 pp.1-5 |
NLC |
2015-09-10 14:45 |
Tokyo |
Convention Room AP Shibuya-Dogenzaka (Tokyo) |
Extraction of Positive and Negative Comments from Nico Nico Douga Hiroki Sakaji, Hiroyuki Sakai (Seikei Univ.) NLC2015-22 |
(Advance abstract in Japanese is available) [more] |
NLC2015-22 pp.31-35 |
NLC |
2015-09-10 15:10 |
Tokyo |
Convention Room AP Shibuya-Dogenzaka (Tokyo) |
Extraction of Laboratory Front Pages from University Web Sites Atsuya Miyazaki, Hiroyuki Sakai, Hiroki Sakaji (Seikei Univ.) NLC2015-23 |
In this paper, we propose a method that extracts a laboratory front pages from university web sites. Our method extracts... [more] |
NLC2015-23 pp.37-41 |
NLC |
2014-09-12 09:30 |
Tokyo |
Yahoo Japan Inc., 11F Seminar Room |
Extraction of Positive Comments from Nico Nico Douga Junya Ishibuchi, Hiroki Sakaji, Hiroyuki Sakai (Seikei Univ.) NLC2014-21 |
In this paper, we propose a method that extracts positive comments (e.g. "もっと評価されるべき"(motto hyouka sarerubeki: should be... [more] |
NLC2014-21 pp.17-21 |
NLC |
2014-09-12 13:25 |
Tokyo |
Yahoo Japan Inc., 11F Seminar Room |
Estimation of Tags Relevant to a Search Query for the Company Search System Hiroyuki Sakai, Hiroki Sakaji (Seikei Univ.) NLC2014-26 |
In this paper, we propose a method that annotates tags corresponding to search queries in our company search system.
Th... [more] |
NLC2014-26 pp.41-45 |
NLC |
2014-09-12 13:50 |
Tokyo |
Yahoo Japan Inc., 11F Seminar Room |
Classification of comments on Nico Nico Douga for annotation based on referred contents Akihito Ikeda, Akio Kobayashi (Toyohashi Univ. of Tech.), Hiroki Sakaji (Seikei Univ.), Shigeru Masuyama (Toyohashi Univ. of Tech.) NLC2014-27 |
For Nico Nico Douga, a video sharing website in Japan, one of the largest feature is posted comments. The comments are o... [more] |
NLC2014-27 pp.47-52 |
NLC |
2014-09-12 17:10 |
Tokyo |
Yahoo Japan Inc., 11F Seminar Room |
Automatic Associating the Nikkei Press-Releases with Nico Nico Douga Keita Shibahara, Hiroki Sakaji, Hiroyuki Sakai (Seikei Univ.) NLC2014-33 |
In recent years, animation sites, such as animation share services and animation distribution services, have spread by d... [more] |
NLC2014-33 pp.75-79 |
NLC |
2014-02-06 13:45 |
Kyoto |
Campus Plaza Kyoto |
Extraction of Character's Impression Expressions from Twitter Yukina Tahara, Hiroki Sakaji, Hiroyuki Sakai (Seikei Univ.) NLC2013-48 |
In this paper, we introduce a method of extracting character’s impression expressions from Twitter. We focus Yuru-chara ... [more] |
NLC2013-48 pp.5-10 |
NLC |
2011-07-07 13:35 |
Tokyo |
IBM Japan, Ltd. |
An Extraction Method of Causal Knowledge from Newspaper Corpus Hiroki Sakaji, Shigeru Masuyama (TUT) NLC2011-2 |
This paper proposes a method that extracts causal knowledge from news paper articles via clue expressions.
Our method d... [more] |
NLC2011-2 pp.7-10 |