Search Results for author: Makoto Onizuka

Found 20 papers, 12 papers with code

Language-agnostic Representation from Multilingual Sentence Encoders for Cross-lingual Similarity Estimation

1 code implementation EMNLP 2021 Nattapong Tiyajamorn, Tomoyuki Kajiwara, Yuki Arase, Makoto Onizuka

Experimental results on both quality estimation of machine translation and cross-lingual semantic textual similarity tasks reveal that our method consistently outperforms the strong baselines using the original multilingual embedding.

Cross-Lingual Semantic Textual Similarity Machine Translation +3

Large Language Models as Urban Residents: An LLM Agent Framework for Personal Mobility Generation

no code implementations22 Feb 2024 Jiawei Wang, Renhe Jiang, Chuang Yang, Zengqing Wu, Makoto Onizuka, Ryosuke Shibasaki, Chuan Xiao

The key technical contribution is a novel LLM agent framework that accounts for individual activity patterns and motivations, including a self-consistency approach to align LLMs with real-world activity data and a retrieval-augmented strategy for interpretable activity generation.


Shall We Talk: Exploring Spontaneous Collaborations of Competing LLM Agents

1 code implementation19 Feb 2024 Zengqing Wu, Shuyuan Zheng, Qianying Liu, Xu Han, Brian Inhyuk Kwon, Makoto Onizuka, Shaojie Tang, Run Peng, Chuan Xiao

Recent advancements have shown that agents powered by large language models (LLMs) possess capabilities to simulate human behaviors and societal dynamics.

BClean: A Bayesian Data Cleaning System

1 code implementation11 Nov 2023 Jianbin Qin, Sifan Huang, Yaoshu Wang, Jing Zhu, Yifan Zhang, Yukai Miao, Rui Mao, Makoto Onizuka, Chuan Xiao

By evaluating on both real-world and synthetic datasets, we demonstrate that BClean is capable of achieving an F-measure of up to 0. 9 in data cleaning, outperforming existing Bayesian methods by 2% and other data cleaning methods by 15%.

Bayesian Inference graph partitioning

A Simple and Scalable Graph Neural Network for Large Directed Graphs

1 code implementation14 Jun 2023 Seiji Maekawa, Yuya Sasaki, Makoto Onizuka

In response, we propose a simple yet holistic classification method A2DUG which leverages all combinations of node representations in directed and undirected graphs.

Classification Node Classification

GuP: Fast Subgraph Matching by Guard-based Pruning

1 code implementation11 Jun 2023 Junya Arai, Yasuhiro Fujiwara, Makoto Onizuka

Subgraph matching, which finds subgraphs isomorphic to a query, is the key to information retrieval from data represented as a graph.

Information Retrieval Retrieval

Learned spatial data partitioning

1 code implementation8 Jun 2023 Keizo Hori, Yuya Sasaki, Daichi Amagata, Yuki Murosaki, Makoto Onizuka

Due to the significant increase in the size of spatial data, it is essential to use distributed parallel processing systems to efficiently analyze spatial data.


Scardina: Scalable Join Cardinality Estimation by Multiple Density Estimators

1 code implementation31 Mar 2023 Ryuichi Ito, Yuya Sasaki, Chuan Xiao, Makoto Onizuka

In recent years, machine learning-based cardinality estimation methods are replacing traditional methods.

GNN Transformation Framework for Improving Efficiency and Scalability

1 code implementation25 Jul 2022 Seiji Maekawa, Yuya Sasaki, George Fletcher, Makoto Onizuka

We propose a framework that automatically transforms non-scalable GNNs into precomputation-based GNNs which are efficient and scalable for large-scale graphs.

Scaling Private Deep Learning with Low-Rank and Sparse Gradients

no code implementations6 Jul 2022 Ryuichi Ito, Seng Pei Liew, Tsubasa Takahashi, Yuya Sasaki, Makoto Onizuka

Applying Differentially Private Stochastic Gradient Descent (DPSGD) to training modern, large-scale neural networks such as transformer-based models is a challenging task, as the magnitude of noise added to the gradients at each iteration scales with model dimension, hindering the learning capability significantly.

An Empirical Study of Personalized Federated Learning

1 code implementation27 Jun 2022 Koji Matsuda, Yuya Sasaki, Chuan Xiao, Makoto Onizuka

Federated learning is a distributed machine learning approach in which a single server and multiple clients collaboratively build machine learning models without sharing datasets on clients.

BIG-bench Machine Learning Personalized Federated Learning

Predicting Parking Lot Availability by Graph-to-Sequence Model: A Case Study with SmartSantander

1 code implementation21 Jun 2022 Yuya Sasaki, Junya Takayama, Juan Ramón Santana, Shohei Yamasaki, Tomoya Okuno, Makoto Onizuka

Nowadays, so as to improve services and urban areas livability, multiple smart city initiatives are being carried out throughout the world.


Similarity Search on Computational Notebooks

no code implementations30 Jan 2022 Misato Horiuchi, Yuya Sasaki, Chuan Xiao, Makoto Onizuka

In this paper, we propose a similarity search on computational notebooks and develop a new framework for the similarity search.

FedMe: Federated Learning via Model Exchange

no code implementations15 Oct 2021 Koji Matsuda, Yuya Sasaki, Chuan Xiao, Makoto Onizuka

First, to optimize the model architectures for local data, clients tune their own personalized models by comparing to exchanged models and picking the one that yields the best performance.

BIG-bench Machine Learning Federated Learning

AIREX: Neural Network-based Approach for Air Quality Inference in Unmonitored Cities

no code implementations16 Aug 2021 Yuya Sasaki, Kei Harada, Shohei Yamasaki, Makoto Onizuka

Since existing methods aim to infer air quality of locations only in monitored cities, they do not assume inferring air quality in unmonitored cities.

Air Quality Inference

Edit Distance Based Curriculum Learning for Paraphrase Generation

no code implementations ACL 2021 Sora Kadotani, Tomoyuki Kajiwara, Yuki Arase, Makoto Onizuka

Curriculum learning has improved the quality of neural machine translation, where only source-side features are considered in the metrics to determine the difficulty of translation.

Machine Translation Paraphrase Generation +1

Fast Subgraph Matching by Exploiting Search Failures

no code implementations28 Dec 2020 Junya Arai, Makoto Onizuka, Yasuhiro Fujiwara, Sotetsu Iwamura

That is, our algorithm generates failure patterns when a partial embedding is found unable to become an isomorphic embedding.


Monotonic Cardinality Estimation of Similarity Selection: A Deep Learning Approach

no code implementations15 Feb 2020 Yaoshu Wang, Chuan Xiao, Jianbin Qin, Xin Cao, Yifang Sun, Wei Wang, Makoto Onizuka

The feature extraction model transforms original data and threshold to a Hamming space, in which a deep learning-based regression model is utilized to exploit the incremental property of cardinality w. r. t.

Management regression

Cannot find the paper you are looking for? You can Submit a new open access paper.