Towards principled user-side recommender systems
R Sato - Proceedings of the 31st ACM International Conference …, 2022 - dl.acm.org
Traditionally, recommendation algorithms have been designed for service developers.
However, recently, a new paradigm called user-side recommender systems has been …
However, recently, a new paradigm called user-side recommender systems has been …
Clear: A fully user-side image search system
R Sato - Proceedings of the 31st ACM International Conference …, 2022 - dl.acm.org
We use many search engines on the Internet in our daily lives. However, they are not
perfect. Their scoring function may not model our intent or they may accept only text queries …
perfect. Their scoring function may not model our intent or they may accept only text queries …
Active learning from the web
R Sato - Proceedings of the ACM Web Conference 2023, 2023 - dl.acm.org
Labeling data is one of the most costly processes in machine learning pipelines. Active
learning is a standard approach to alleviating this problem. Pool-based active learning first …
learning is a standard approach to alleviating this problem. Pool-based active learning first …
Making Translators Privacy-aware on the User's Side
R Sato - arxiv preprint arxiv:2312.04068, 2023 - arxiv.org
We propose PRISM to enable users of machine translation systems to preserve the privacy
of data on their own initiative. There is a growing demand to apply machine translation …
of data on their own initiative. There is a growing demand to apply machine translation …
Overhead-free User-side Recommender Systems
R Sato - arxiv preprint arxiv:2411.07589, 2024 - arxiv.org
Traditionally, recommendation algorithms have been designed for service developers. But
recently, a new paradigm called user-side recommender systems has been proposed. User …
recently, a new paradigm called user-side recommender systems has been proposed. User …
User-Side Realization
R Sato - arxiv preprint arxiv:2403.15757, 2024 - arxiv.org
Users are dissatisfied with services. Since the service is not tailor-made for a user, it is
natural for dissatisfaction to arise. The problem is, that even if users are dissatisfied, they …
natural for dissatisfaction to arise. The problem is, that even if users are dissatisfied, they …
Word Tour: One-dimensional Word Embeddings via the Traveling Salesman Problem
佐藤竜馬 - 自然言語処理, 2022 - jstage.jst.go.jp
第一の問題はデータ容量です. 例えば 40 万語彙 300 次元の埋め込みは約 1 GB
のメモリを消費します. 計算サーバー上で処理を行うのであればこれは許容範囲内ですが …
のメモリを消費します. 計算サーバー上で処理を行うのであればこれは許容範囲内ですが …
ウェブから能動学習の観点で有益なデータを取得する
佐藤竜馬 - 人工知能学会全国大会論文集 第 37 回 (2023), 2023 - jstage.jst.go.jp
抄録 能動学習はラベルの付いていないデータのプールの中からラベルを付けるべきデータを
繰り返し選択することでラベル付けのコストを削減することを目指す技術である …
繰り返し選択することでラベル付けのコストを削減することを目指す技術である …