Theses

This page lists bachelor, master, and doctoral theses submitted in the Content Environment Design Laboratory (Ushiama Lab). Thesis titles are translated into English. For the original Japanese titles, please see the Japanese version of this page.

Doctoral Theses

A Study on Content Presentation Methods Using Smartphones to Support Activities in the Real World

A Study on Content Browsing Support Based on Interest Estimation from User Behavior

Master Theses

A Study on Recommendation Methods Considering Novelty to Support the Discovery of New Preferences
Design of an LLM Agent and Dialogue Process Control Framework to Support Deep Art Appreciation Based on Stepwise Art Criticism

Extracting and Visualizing Latent Relationships between News Articles Using SNS Comments
A Study on Repeated Music Recommendation Considering Human Cognitive Characteristics
Automatic Generation of Images Expressing Impressions of Music from Lyrics and Audio Using Generative AI
POI Recommendation with Explanations Based on User Intent Using BERT
Personalized Prediction of Offensive Comments Using Models of Poster and Reader Characteristics in Social Media
Estimating Electromyograms from Multi-exercise Training Videos for Strength Training Support

Extracting Sentences Describing Product Features from Reviews for Product Selection Support

Extracting Sentences Describing Product Features from Reviews for Product Selection Support
A Study on Follow Recommendation Methods in Social Networking Services

Situation Estimation Using Live Tweets for Sports Spectating Support
Automatic Generation of Infographics for Cooking Recipe Comprehension Support
A Study on News Comprehension Support Using Context Information Extracted from SNS

An Interactive Cooking Recipe System that Cooperates with Cooking Appliances to Support Users' Cooking Activities
Automatic Generation of Illustrated Maps Based on the Content of Geotagged Photos

A Study on Item Recommendation Methods Based on Interaction Histories

Supporting Item Narrowing and Selection Based on Feedback to Reviews
A Study on Neutral Understanding Support for News Distributed via SNS
A Study on Dynamic Web Page Optimization According to User Goals
A Study on Composition Support Using Machine Learning

A Study on Extracting Regional Characteristics and Estimating Regional Similarity Using Photo-sharing SNS

A Study on SNS to Foster Social Capital in Local Communities
Tweet Aggregation Methods for Effectively Presenting Excitement in Social Viewing
Review Ranking Methods for Item Selection Support Based on Comparison

A Study on Efficient Audio Browsing of SNS Streams
A Study on Tourist Spot Recommendation Considering Time of Visit

An Information Recommendation Method Using Twitter to Help Users Discover New Interests
Development of a “Browsing” Interface for E-book Novels Using an Interest-evoking Map

Personalization of Chart-style Rankings on User-generated Content Sites
A Recommender System for Museum Exhibits

A Study on Real-world Content Browsing Support on Camera Phones
A Study on an Efficient Content Browsing Interface on Mobile Phones Based on the Magnifying Glass Metaphor
Automatic Composition of News Programs Considering Early Termination in Micro-time Viewing

Automatic Composition of Digital Photo Slideshows Based on Semantic Relations

Bachelor Theses

Control of Tempo Expressions in Animation Editing Using Prosodic Features of Speech
Generation of Personalized Educational Videos via Automatic Editing with LLMs and Learner Feedback
An Investigation of the Opacity of Meta-information Processing in Machine Translation for Entertainment Content
Subjective Search Considering Empathy with Reviewers
Development of a “Non-answering AI” that Supports Improvement of Learners' Abilities

Fundamental Study on Building an Agent that Can Talk about Impressions of Novels
Creation of “Uta-e”: Experiencing the World of Waka Poetry Using Generative AI
Development of a Personal Training System Providing Continuous Support Using Generative AI
Review Retrieval with Interactive Feedback Using RAG for Reflecting on Consumed Content

Automatic Generation of High-quality Questions in Conversational Learning Systems
Review Writing Support Using Large Language Models for Movie Appreciation
Review Recommendation Based on Explainability of Rating Prediction Models
A Study on Automatic Generation of Novel Illustrations by AI

Extraction of “Striking” Recommendation Phrases Based on Explainability of “Like” Prediction for Book Reviews
A Method for Browsing News Comments Based on Reactions to News on SNS
Automatic Extraction of Implicit Replies in SNS for News Comprehension Support

Predicting the Number of “Likes” for Influencer Recommendation in SNS Advertising
Automatic Generation of Catchphrases for Tourist Spots Based on Online Reviews
A Study on Automatic Banner Ad Generation Using Generative Adversarial Networks
Estimating Training Effects from Exercise Videos for Training Support

Detection of Local “Soft News” Using SNS
An Interface for Browsing Live Tweets to Reflect on TV Drama Episodes
Proposing a Visitor Limiting Method that Balances Overtourism Control and Tourist Satisfaction

A Recommendation Method Considering Bias in News Distributed via Twitter
A Tweet Recommendation Method with High Serendipity Considering Users' “Like” Criteria
Analyzing Live Tweets for “Best Scene” Sports Viewing

Extracting Communities with Shared Interests Based on Conversations on Twitter

A Study on Tweet Filtering Based on Estimated Reader Impressions
A Study on Scene Feature Extraction in TV Dramas Using Live Tweets
A Serious Game Enabling the Experience of Collaboration with Others Sharing a Living Space
A Serendipitous User Recommendation Method on Twitter
A Study on Extracting Fan Characteristics of Music Artists Using Twitter

Analysis of the Relationship between Features of TV Dramas and Viewership Ratings Using Twitter
Fashion Information Recommendation Based on Users' Attitudes toward Trends on SNS
Discovering News Articles with Opposing Stances Using Twitter Users' Reactions
A Study on Responsive Web Page Optimization According to Users' Browsing Intent
Optimization of Book Recommendation Using Feedback to Reviews in a Book Review SNS

Recommendation of TV Anime Programs Using Viewing Behaviors in Social Viewing
Visualization of Culture and Climate Using SNS
A Study on Recommendation According to Users' Internal Context
Automatic Construction of Interactive Dance Tutorials Using Dancer Motion Data

Automatic Classification of Joke Tweets Using User Reactions
Content Selection Support by Presenting Analogies that Clarify Differences
Empathic Viewing: Supporting TV Viewing with Like-minded People Using SNS
Visualization for Efficient Selection on Online Recipe Sites

Efficient Browsing of Social Streams Using the Cocktail Party Effect in a 3D Audio Space
Constructing Ad-hoc Follow Networks on Twitter for Real-world Event Information Gathering
A Context-aware Interface Considering Skipping Behavior in Social Stream Reading
Recommendation Based on Influence Relationships among Music Artists

Book Recommendation Based on Readers' Impressions Extracted from Reviews
Supporting the Prevention of Rumor Propagation on Twitter
Geographic Information Ranking for Mobile Search Using GPS Logs

Personalized Information Recommendation Using a Micro-blog Interface
Effective Information Access by Overlaying Web Pages
Extracting Interesting Sentences from Reviews for “Browsing” Support of E-books

Ranking Methods Considering Quality of User-generated Content
News Browsing Methods Suitable for Skimming on Mobile Phones
Automatic Construction of Personalized Museum Tour Routes

A Content Browsing Interface for Small Mobile Devices Using Real-world Maps
Content Browsing on Mobile Phones Based on the Magnifying Glass Metaphor
Classification of Sports Articles Reflecting Differences in Users' Standpoints

These titles are English translations created for reference. Please refer to the Japanese page for the original wording.