no code implementations • 17 Oct 2023 • Mitsuki Morita, Masato Kikuchi, Tadachika Ozono
Our results suggest that further analyses of the features contributing to lyric-lyricist classification performance on the lowest lyricist-singer entropy group may improve the feature extraction task for lyricists.
no code implementations • 28 Oct 2022 • Masato Kikuchi, Tadachika Ozono
In this paper, we integrated the estimator with the UNB.
no code implementations • 28 Oct 2022 • Masato Kikuchi, Yuhi Kusakabe, Tadachika Ozono
A naive likelihood ratio (LR) estimation using the observed frequencies of events can overestimate LRs for infrequent data.
no code implementations • 13 Jul 2022 • Shogo Anda, Masato Kikuchi, Tadachika Ozono
We determined proper labels based for the words identified through pattern matching from product reviews to create the training data.
no code implementations • 17 Dec 2021 • Weijian Li, Masato Kikuchi, Tadachika Ozono
Large-scale e-commerce sites can collect and analyze a large number of user preferences and behaviors, and thus can recommend highly trusted products to users.
no code implementations • 5 Nov 2021 • Masato Kikuchi, Mitsuo Yoshida, Kyoji Umemura, Tadachika Ozono
However, because this method deals with a large number of discrete values, the running time and memory usage for estimation are problematic.
no code implementations • 8 Oct 2021 • Kento Kaku, Masato Kikuchi, Tadachika Ozono, Toramatsu Shintani
The results show that our evaluation model can identify top-conference titles more effectively than intermediate English and beginner students.