Aller au contenu principal
Profil bibliographique

Tomoya Ichikawa

Informations fournies par OpenAlex. Research Africa ne déduit ni nationalité, ni poste, ni coordonnées personnelles.

10Publications signalées
77Citations signalées
1Affiliations récentes

Les institutions déclarées

Les domaines associés

Dementia and Cognitive Impairment ResearchTopic ModelingAlzheimer's disease research and treatmentsNatural Language Processing TechniquesStress Responses and Cortisol

Les publications récentes

Accès ouvert 2026 article OpenAlex

Serum Amyloid β Oligomer May Predict Treatment Response in Middle‐Aged and Late‐Life Patients With Depression

Kentaro Shimizu, Seita Yasuda, Hitoshi Maeshima, Tomoya Ichikawa et autres

AIM: Late-life patients with depression are reportedly less responsive to antidepressant treatment than younger patients. Additionally, patients who have depression comorbid with Alzheimer's disease (AD) and those with an amyloid β (Aβ) burden have shown a poor response to antidepressant treatment, suggesting …

jp (code pays fourni par la source)

1 citation Neuropsychopharmacology Reports
Accès ouvert 2023 article OpenAlex

Auxiliary Learning for Named Entity Recognition with Multiple Auxiliary Training Data

Taiki Watanabe, Tomoya Ichikawa, Akihiro Tamura, Tomoya Iwakura et autres

固有表現抽出 (Named Entity Recognition; NER) は,テキストからの知識獲得に用いられる要素技術の一つであり,たとえば,化学物質や医療の知識抽出に用いられている.NERの性能改善のため,対象タスクの教師データとは別の教師データを補助教師データとして用いる補助学習が提案されている.従来の補助学習では補助教師データとして1種類の教師データしか用いていない.そこで,本研究では,複数種類の教師データを補助教師データとして活用するNERの学習手法 (Multiple Utilization of NER Corpora Helpful for Auxiliary BLESsing; MUNCHABLES) を提案する.具体的には,補助教師データ毎の補助学習を順次行うことで,対象タスクのモデルを補助教師データの種類の数だけ再学習する方法と,全種類の教師データを一つの補助学習で用いる方法の2種類の学習手法を提案する.評価実験では,化学物質名抽出タスクにおいて,7種類の化学/科学技術分野の補助教師データを用いて提案手法で学習したモデルの評価を行った.その結果,提案手法によるモデルはマルチタスク学習や1種類の補助教師データを用いる補助学習手法によるモデルと比べて,7種類のデータセットにおける F1 値のマイクロ平均,マクロ平均ともに高い性能となることを確認した.また,s800のデータセットにおいて従来手法と比較をして最も高い F1 値を達成した.

jp, us (code pays fourni par la source)

0 citations Journal of Natural Language Processing
Accès ouvert 2022 conference-paper OpenAlex

Auxiliary Learning for Named Entity Recognition with Multiple Auxiliary Biomedical Training Data

Taiki Watanabe, Tomoya Ichikawa, Akihiro Tamura, Tomoya Iwakura et autres

Named entity recognition (NER) is one of the core technologies for knowledge acquisition from text and has been used for knowledge extraction of chemicals and medicine.As one of the NER improvement approaches, multitask learning that learns a model from multiple training data …

jp, us, cn (code pays fourni par la source)

4 citations
2019 article OpenAlex

Serum levels and mutual correlations of amyloid β in patients with depression

Seita Yasuda, Hajime Baba, Hitoshi Maeshima, Takahisa Shimano et autres

AIM: Epidemiological studies have shown that depression is a risk factor for Alzheimer's disease (AD). Although the biological mechanism underlying the link between depression and AD is unclear, altered amyloid β (Aβ) metabolism in patients with depression has been suggested as a …

jp (code pays fourni par la source)

12 citations Geriatrics and gerontology international/Geriatrics & gerontology international
Accès ouvert 2003 article OpenAlex

Behavior of Heat Embrittlement in Welded Pipe of Nuclear Plant and Probabilistic Estimation of Safety and Reliability.

Nagatoshi OKABE, Maho Hosogi, Akira Nishihara, Tomoya Ichikawa

Embrittlement behavior of the duplex stainless steel in the intermediate temperature range (275 to 500°C) is one of the very serious problems for applying reliability assessment to practical pipes in a nuclear power plant. In this work, it is shown by formularization …

0 citations Journal of the Society of Materials Science Japan

BNTIC News n’est pas le producteur de ces données. Les publications sont interrogées à la demande dans Crossref, OpenAIRE, DOAJ, Europe PMC, HAL, DataCite, AfricArXiv, ROR et la Banque mondiale, sans clé d’accès. OpenAlex reste optionnel. Aucun service payant n’est nécessaire et aucune donnée externe n’est enregistrée en base. Consulter les sources et leurs limites.