No description available.

No description available.

No description available.

京都大学 / 新制・課程博士 / 博士(人間健康科学) / 甲第25504号 / 人健博第124号 / 新制||人健||8(附属図書館) / 京都大学大学院医学研究科人間健康科学系専攻 / (主査)教授 黒木 裕士, 教授 中尾 恵, 教授 西浦 博 / 学位規則第4条第1項該当 / Doctor of Human Health Sciences / Kyoto University / DGAM

This research addresses the problem, can unsupervised learning generate a representation that improves on the commonly used term frequency-inverse document frequency (TF-IDF ) representation by capturing semantic relations? The analysis measures the quality of sentence classification using term TF-IDF representations, and finds a practical upper limit to precision and recall in a biomedical text classification task (F1-score of 0.85). Arguably, one could use ontologies to supplement TF-IDF, but ontologies are sparse in coverage and costly to create. This prompts a correlated question: can unsupervised learning capture semantic relations at least as well as existing ontologies, and thus supplement existing sparse ontologies? A shallow neural network implementing the Skip-Gram algorithm is used to generate semantic vectors using a corpus of approximately 2.4 billion words. The ability to capture meaning is assessed by comparing semantic vectors generated with MESH. Results indicate that semantic vectors trained by unsupervised methods capture comparable levels of semantic features in some cases, such as amino acid (92% of similarity represented in MESH), but perform substantially poorer in more expansive topics, such as pathogenic bacteria (37.8% similarity represented in MESH). Possible explanations for this difference in performance are proposed along with a method to combine manually curated ontologies with semantic vector spaces to produce a more comprehensive representation than either alone. Semantic vectors are also used as representations for paragraphs, which, when used for classification, achieve an F1-score of 0.92. The results of classification and analogical reasoning tasks are promising but a formal model of semantic vectors, subject to the constraints of known linguistic phenomenon, is needed. This research includes initial steps for developing a formal model of semantic vectors based on a combination of linear algebra and fuzzy set theory subject to the semantic molecularism linguistic model. This research is novel in its analysis of semantic vectors applied to the biomedical domain, analysis of different performance characteristics in biomedical analogical reasoning tasks, comparison semantic relations captured by between vectors and MESH, and the initial development of a formal model of semantic vectors. / Ph. D.

Yes / Online shopping is increasingly becoming people's first choice when shopping, as it is very convenient to choose products based on their reviews. Even for moderately popular products, there are thousands of reviews constantly being posted on e-commerce sites. Such a large volume of data constantly being generated can be considered as a big data challenge for both online businesses and consumers. That makes it difficult for buyers to go through all the reviews to make purchase decisions. In this research, we have developed models based on machine learning that can predict the helpfulness of the consumer reviews using several textual features such as polarity, subjectivity, entropy, and reading ease. The model will automatically assign helpfulness values to an initial review as soon as it is posted on the website so that the review gets a fair chance of being viewed by other buyers. The results of this study will help buyers to write better reviews and thereby assist other buyers in making their purchase decisions, as well as help businesses to improve their websites.

The demand for a creative workforce has never been higher, yet schools struggle to teach and assess creativity among students predictably and efficiently. Compositions are an effective way to incorporate creativity across the curriculum; however, essays are time consuming for teachers to score for objective quality or subjective creativity. In this study, I explored a) if high creativity scores are related to high quality and sophistication in academic writing, and b) if extant text-mining tools effectively identify quality, sophistication, and creativity in academic essays. I collected 230 essays written by Grade 11 students. Four human-raters analyzed these essays for quality, sophistication, and creativity, and I used text-mining tools designed to assess creativity to analyze the same. Using correlations - including the variables semantic distance (measured against the GloVe corpus), entropy (measured with Shannon's Entropy Index), and idea density (measured with CPIDR5.1) - I assessed both human-raters' and text-mining tools' proficiency at identifying quality, sophistication, and creativity in academic essays. Quality, sophistication, and creativity were also regressed on these same text-mining variables to assess which method - human or computer – and which of the text-mining tools - best predicts these dependent variables. Human-raters' creativity scores correlated with human-raters' quality scores (r = .418) and sophistication scores (r = .321), as well as the text-mining tools MeanSim (r = -.131), OCS Originality (r = .359), Idea Density (r = .368), and Entropy (r = .388). These findings suggest text-mining tools designed for creativity can capture quality and sophistication of student essays. A comparison of human-raters' creativity scores and text-mining models revealed text-mining models can capture quality (R2 = .445) and sophistication (R2 = .373) better than human raters can capture quality (R2 = .175) and sophistication (R2 = .103).

The purpose of this dissertation was to use the technology of text mining and topic modeling to explore unobserved themes of medical case reports that involve medical imaging. Case reports have a valuable place in medical research because they provide educational benefits, offer evidence, and encourage discussions. Their form has evolved throughout the years, but they have remained a key staple in providing important information to the medical communities around the world with educational context and illuminating visuals. Examining medical case reports that have been published throughout the years on multiple medical subjects can be challenging, therefore text mining and topic modeling methods were used to analyze a large set of abstracts from medical case reports involving radiology. The total number of abstracts used for the data analysis was 68,845 that were published between the years 1975 to 2022. The findings indicate that text mining and topic modeling can offer a unique and reproducible approach to examine a large quantity of abstracts for theme analysis.

This research examines the relationship between quality management (QM) practices and performance in the healthcare industry via the conduct of three studies. The results of this research contribute both to advancing QM theory as well as in developing a unique text mining method that is illustrated by examining QM in the healthcare industry. Essay 1 explains the relationship between operational performance and QM practices in the healthcare industry. This study analyzed the findings from the literature using meta-analysis. We applied confirmatory semantic analysis (CSA) to examine the Baldrige winners' applications. Essay 2 examines the benefits associated with an effective QM program in the healthcare industry. This study addressed the research question about how effective QM practice results in improved hospital performance. This study compares the performance of Baldrige Award-winning hospitals with matching hospitals, state average, and national average. The results show that the Baldrige Award can lead to an increase in patient satisfaction in certain periods. Essay 3 discusses the contribution of an online clinic appointment system (OCAS) to QM practices. An enhanced trust model was built on understanding the mechanism of patients' trust formation in the OCAS. Understanding the determinants related to patients' trust and willingness to use OCAS can provide valuable guidance for medical institutions to establish health information technology-based services in the quality service improvement programs. This research has three significant contributions. First, this research analyzes the role of QM practices in the healthcare industry. Second, this research attempts to develop a unique text mining method. Third, this research provides a validated trust model and contributes to the body of research on the trust of healthcare information technology.

Vor dem Hintergrund des steigenden Interesses von computergestützter Textanalyse in Forschung und Praxis entwickelt dieser Beitrag auf Basis aktueller Literatur ein generisches Vorgehensmodell für Text-Mining-Prozesse. Das Ziel des Beitrags ist, die dabei anfallenden, umfangreichen Aktivitäten zu strukturieren und dadurch die Komplexität von Text-Mining-Vorhaben zu reduzieren. Das Forschungsziel stützt sich auf die Tatsache, dass im Rahmen einer im Vorfeld durchgeführten, systematischen Literatur-Review keine detaillierten, anwendungsneutralen Vorgehensmodelle für Text Mining identifiziert werden konnten. Aufbauend auf den Erkenntnissen der Literatur-Review enthält das resultierende Modell daher sowohl induktiv begründete Komponenten aus spezifischen Ansätzen als auch aus literaturbasierten Anforderungen deduktiv abgeleitete Bestandteile. Die Evaluation des Artefakts belegt die Nützlichkeit des Vorgehensmodells im Vergleich mit dem bisherigen Forschungsstand.