Digital handwriting is increasingly used in educational settings; however, many classroom styluses omit pen pressure sensitivity due to cost and procurement constraints. While prior work has shown that pen pressure supports written arithmetic, its role in other learning activities remains insufficiently understood. In this study, we investigate how pen pressure–based stroke modulation affects problem-solving performance in digital handwriting, using geometry problems as a case study. We conducted a between-subjects experiment comparing pressure-sensitive and non-pressure-sensitive conditions in tasks requiring spatial reasoning and iterative diagram construction. Although overall accuracy did not differ significantly between conditions, detailed analyses revealed that the absence of pen pressure disproportionately affected lower-performing participants and reduced accuracy in unfamiliar or cognitively demanding problems. In solid geometry tasks, the non-pressure-sensitive condition also resulted in longer completion times. Qualitative analyses further showed that pen pressure enabled effective visual organization and depth representation, whereas its absence led participants to adopt compensatory diagramming strategies, such as relocating annotations outside figures. These findings indicate that omitting pen pressure is not a cognitively neutral design decision. Pen pressure functions as a representational resource that supports exploratory reasoning and visual organization in digital handwriting, with important implications for the design of educational input devices and learning environments.

In crowdsourced user experiments that collect performance data from graphical user interface (GUI) interactions, some participants ignore instructions or act carelessly, threatening the validity of performance models. We investigate a pre-task screening method that requires simple GUI operations analogous to the main task and uses the resulting error as a continuous quality signal. Our pre-task is a brief image-resizing task in which workers match an on-screen card to a physical card; workers whose resizing error exceeds a threshold are excluded from the main experiment. The main task is a standardized pointing experiment with well-established models of movement time and error rate. Across mouse- and smartphone-based crowdsourced experiments, we show that reducing the proportion of workers exhibiting unexpected behavior and tightening the pre-task threshold systematically improve the goodness of fit and predictive accuracy of GUI performance models, demonstrating that brief pre-task screening can enhance data quality.

Zoos and aquariums often house animals in groups, making it difficult for visitors to recognize individuals. Penguins exemplify this challenge, even though identifying individuals can foster empathy and engagement. We addressed the following research questions: (RQ1) How do zoos and aquariums currently provide individual identification and what challenges remain? (RQ2) How does a sketch-based identification system affect visitor behavior and awareness in a real-world setting? To answer these questions, we (1) surveyed 25 facilities and clarified the limitations of existing practices, (2) refined a drawing-based retrieval algorithm to handle partial observations, (3) implemented a practical interactive system, and (4) validated it in a field study with 167 visitor groups (270 individuals). Results showed that the system encouraged name-based conversations, enhanced recognition of individual penguins, and increased dwell time. These findings suggest that sketch-based retrieval can enrich visitor experiences by promoting active observation and deeper engagement.

Numerous studies have investigated the relationship between text format and memory performance, revealing that less readable text formats can sometimes enhance memory retention. This counterintuitive phenomenon is often referred to as the disfluency effect, which suggests that perceptual difficulty may lead to deeper cognitive processing. The objective of this study was to explore whether a similar disfluency effect would occur in reading comprehension. We therefore conducted two experiments: a paper-based experiment and a screen-based experiment. In both experiments, participants read passages presented in two font styles and two types of handwritten text, followed by reading comprehension questions. In the paper-based experiment, participants tended to perform slightly better when the text was more readable, although the difference was not statistically significant. In the screen-based experiment, where response times were recorded, participants spent less time answering when the text was less readable, and their accuracy was also slightly lower. These findings suggest that, in the case of long passages such as those used in this study, the cognitive load imposed by less readable text may interfere with comprehension, potentially offsetting the benefits predicted by the disfluency effect.

With the digitization of education, opportunities for handwritten input using tablets and other digital devices have increased. However, the impact of digital handwriting on problem-solving performance and its contributing factors remains unclear. This study focuses on pen pressure, a fundamental element of handwriting, and investigates how pressure sensitivity affects arithmetic problem-solving. We hypothesized that allowing users to vary stroke darkness through pen pressure would enhance accuracy. To test this, we conducted an experiment comparing a pressure-sensitive condition, where pen pressure affected stroke darkness, and a non-pressure-sensitive condition, where stroke darkness was uniform. University students solved arithmetic problems involving addition, subtraction, multiplication, and division. The results showed no significant difference in accuracy for addition, subtraction, or multiplication, but significantly higher accuracy in division problems under the pressure-sensitive condition. Pen pressure analysis also revealed differences in value distribution between conditions. Furthermore, error analysis suggested that the inability to modulate stroke darkness might lead to misinterpretation of auxiliary digits or other handwritten calculation marks during complex tasks. These findings indicate that pen pressure sensitivity may support more effective problem-solving in arithmetic tasks that involve complex, multi-step reasoning, by improving the visual clarity of handwritten information.

In Japan, handwriting practice is an important part of education and often involves repeatedly writing the same character, which can become monotonous and reduce motivation. We propose a melody-based handwriting system that maps pen strokes to pitches and stops playback when the stroke leaves a predefined spatial gate, providing spatially-contingent auditory feedback that encourages accurate tracing while maintaining engagement. We implemented a browser-native, low-latency prototype that synchronizes visual input and audio in real time. In a controlled experiment comparing continuous feedback, restricted feedback (proposed), and no feedback, the proposed method led to slower, more attentive writing and a higher proportion of strokes within the designated range, indicating improved precision.

Intrinsic motivation is known to improve task performance when individuals make their own choices. However, when multiple tasks are available, people often choose easier ones even when more difficult or troublesome tasks may be more beneficial. This study investigates whether the phrasing of spoken options can influence such decisions in Voice-based interfaces by leveraging the cognitive and articulatory effort required for vocalization. We conducted a controlled experiment with 40 participants, systematically varying the linguistic complexity of Japanese adverbial phrases in a pointing task and comparing Voice-based and Click-based selection. Results indicated a clear tendency in the voice condition to avoid the most complex phrase and revealed a modality-specific positional tendency in which left-positioned options were chosen more often and right-positioned options were avoided. To our knowledge, this is the first empirical study to demonstrate that vocalization cost can systematically bias task selection in Japanese voice interfaces. These findings suggest that carefully designed spoken language can subtly guide task selection, providing implications for fair and effective voice interface design.

Measuring the weight or amount of ingredients is necessary when cooking but it requires tools and can often feel burdensome to the cook. This study developed a method to estimate the weight of an object without special tools by observing the change in fingernail color when holding an object (the color of human fingernails changes when force is applied to the fingertips). We constructed a dataset of nail images with varying weights of held objects. Our analysis revealed a tendency for the values of G and B in the nail's RGB color and H and S in HSV to be related to the object's weight. Linear regression was used for the estimation, and it was found that, although there were individual differences, estimation was possible with an error of 20.99 g in the most accurate case.

Some aquariums give names to the individual animals in their care in order to encourage visitors to observe them. However, it is not easy for visitors to remember names without their having a detailed memory of the individual animals. We previously proposed a method to enable visitors to aquariums to identify individual penguins by drawing the animals’ distinctive abdominal patterns and analyzed the similarity of the drawings. In this study, we developed a prototype system that enables users to retrieve penguins' names by drawing abdominal patterns on a smartphone and investigated its effectiveness. The experiment's results suggested that observations using this system improved the ability of some of the experiment's participants to remember the penguins.

Since comics are diverse, there are some depictions that readers like and some that they do not like. If readers find a part of a comic that they dislike, they can continue reading it even if they can skip that part, but there is a possibility that they will stop reading it entirely if they actually read it. Therefore, we propose a method for allowing readers to enjoy comics without worrying about depictions that they do not like. To realize this method, we developed a system that allows readers to flag depictions they dislike while reading comics and conducted data collection experiments. We also implemented a system in which jirai (content to be avoided) flags and announcements are given to readers while reading a comic, and we examined the feasibility of the system by operating it for about four weeks. We confirmed that the flagging and jirai announcements were performed during the system operation. In addition, the evaluation of jirai judgments for comics using the Vision API showed that while there is potential for AI to make judgments, there are still difficulties in judging detailed depictions.

In reading manga, it is common to look back at the storyline when following a serialized work. Although there are services that assist comic re-reading through quizzes, searching for specific parts related to the quiz takes a lot of time and complicates the review process. Therefore, in this study, we examined whether it is possible to estimate the scenes related to the quiz based on the quiz questions, answers, and manga-specific features. To achieve this, we extracted key elements from the comic and proposed two estimation methods: a word-based CS method and a context-based GPT method. Furthermore, we discussed extractable and difficult-to-estimate scenes in comics. The results showed that the pages containing the answers could be estimated with a probability of 66.7%. Pages containing specific keywords or events were easier to estimate, while those requiring an understanding of the comic’s overall time series and context were more difficult to estimate. In addition, since the accuracy varied greatly depending on the presence or absence of the answer text, it can be considered that the content as close as possible to the topic of the quiz can be estimated if important keywords such as the answer text are included.

It takes several months or years to release a new volume of a comic book after the previous volume. Therefore, when reading a long-awaited newest volume of a comic, it is sometimes difficult to recollect and understand the flow of the story, causing the readers to reread the previous volume or reread from the first volume to check the story. Re-reading can be fun when there is enough time. However, when there are many volumes of the comic or the time is limited, the reader will want to recollect the previous content as soon as possible to read the newest volume. One way to recollect the previous content quickly is to check its summary. However, a synopsis is often not enough to recollect everything and may become a spoiler if the reader has forgotten to read the previous volume. In this paper, we proposed and implemented a system that enables users to recollect the content of the previous volume by quizzes (question-answer pairs). We considered that just reading a question text would not be a spoiler. In addition, we released our system, “ComiQA,” as a Web service and found the characteristics of quizzes made by analyzing the registered 1465 quizzes in our service. We also experimented to investigate and compare the degree of recollection of creating quizzes and writing reviews. We found that creating quizzes helps people recollect the episodes more effectively than writing reviews, and viewing the quiz leads to further recollection.

In online games, some people are disadvantaged due to hearing or vision handicaps that have nothing to do with their abilities. People with color vision diversity, who have difficulty seeing certain colors, are at a disadvantage in games that require color judgments because of the time it takes to read color information. To assist people with color vision diversity, some games support them by using color schemes that match their color vision type, but not all color vision types are supported, and there are limitations to the support provided. In our previous studies, we have conducted experiments in which we implemented a D-type simulating filter to clarify easy colors to recognize both for those with normal color vision and those with D-type color vision, and colors that are close in time to recognize for both users. However, we have not clarified whether these experiments are effective in actual games. In this study, we experimented using the game “Among Us” and examined whether it was possible to control the color handicap. Our results showed that it is possible to control the player's advantages and disadvantages in the game, both for those with normal color vision and those with color vision diversity, depending on the color scheme.

People differ in their assessment of whether they consider spoilers to be problematic or not. Recent studies have investigated the effects of spoilers. However, the definition of spoilers in those studies was ambiguous, so it is difficult to say that the impact of spoilers has been investigated. In this work, we propose a definition of comic spoilers and a dataset construction method to avoid ambiguity in the definition of spoilers. Furthermore, we construct a dataset and compare our dataset with the spoilers used in previous studies to clarify the characteristics of spoilers. Then, we found that the newly constructed dataset not only covered the previous dataset but also detected important spoiler scenes that were not included in the previous dataset, such as goal scenes and scenes that reveal the killer's motive and so on, as spoilers.

Evaluating the validity of an existing user performance model in a variety of tasks is important for enhancing its applicability. The model studied in this work is the steering law for predicting the speed and time needed to perform tasks in which a cursor or a car passes through a constrained path. Previous HCI studies have refined this model to take additional path factors into account, but its applicability has only been evaluated in GUI-based environments such as those using mice or pen tablets. Accordingly, we conducted a user experiment with a driving simulator to measure the speed and time on curved roads and thus facilitate evaluation of models for pen-based path-steering tasks. The results showed that the best-fit models for speed and time had adjusted r^2 values of 0.9342 and 0.9723, respectively, for three road widths and eight curvature radii. While the models required some adjustments, the overall components of the tested models were consistent with those in previous pen-based experimental results. Our results demonstrated that user experiments to validate potential models based on pen-based tasks are effective as a pilot approach for driving tasks with more complex road conditions.

“Oshigatari” is a Japanese term that refers to when a person talks to recommend his/her favorite content, actors, artists, and so on to others not only for his/her enjoyment but also because he/she wants others to be equally passionate about this subject. However, people often have difficulties conveying their enthusiasm and giving convincing recommendations. We conducted surveys using Yahoo! Crowdsourcing and found that 80.3% of respondents have unsuccessful experiences of recommending content to others. Therefore, the objective of our study is to maximize the success of Oshigatari with a little effort. In this study, we proposed a method asking the listeners to imagine generating a search query while listening to the recommendation talk (Oshigatari) to increase the likelihood of the listeners becoming interested. In addition, we experimented and found that our method increased the rate of successful recommendations.

It is difficult for novice drivers to improve their driving skills by themselves. One of the difficulties for novice drivers is the technique for driving on curves, such as the timing and amount of steering. Therefore, in this study, we propose a method to sensitively estimate the amount and timing of steering by using a sine wave of a musical scale corresponding to the steering angle. We called this method the “DoReMi Steering Wheel.” We also implemented a prototype system on a driving simulator and conducted experiments to check the usefulness of our system. The results showed no significant difference in the number of steering corrections between the DoReMi Steering Wheel and the standard steering wheel. Still, there was a significant difference in the subjective evaluation of the ease of driving on curves, suggesting that the DoReMi Steering Wheel may support driving.

It is difficult for novice drivers to improve their driving skills by themselves. One of the difficulties for novice drivers is the technique for driving on curves, such as the timing and amount of steering. Therefore, in this study, we propose a method to sensitively estimate the amount and timing of steering by using a sine wave of a musical scale corresponding to the steering angle. We called this method the “DoReMi Steering Wheel.” We also implemented a prototype system on a driving simulator and conducted experiments to check the usefulness of our system. The results showed no significant difference in the number of steering corrections between the DoReMi Steering Wheel and the standard steering wheel. Still, there was a significant difference in the subjective evaluation of the ease of driving on curves, suggesting that the DoReMi Steering Wheel may support driving.

Because some people appreciate their own handwritten characters blended with the handwritten characters of others, we propose a method for generating good handwriting by the real-time blending of users' handwritten characters with their own past handwritten characters or others' handwritten characters. We also realize the prototype system and show its usefulness.

Many people manage their tasks using tools such as notebooks or personal task management applications in their smartphones. In fact, according to Microsoft's research, 78% of respondents in the United States currently have at least one task management app [1]. However, conventional task lists are sometimes troublesome because tasks usually need to be expressed in words. In addition, it takes time to understand tasks when they are described in words. However, it is known that a person can instantaneously process an image and has the ability to process many images at once [2][3]. Therefore, we propose a system called "PhoToDo" that enables people to use visual images to manage tasks. By using PhoToDo, users can instantly visualize all their tasks and efficiently manage them. In this paper, we propose and implement our system and show its effectiveness by conducting experimental tests.

In many shopping sites such as Amazon.com it is possible to view and write reviews of items (products and content). Reviews of items including stories, such as novels, movies, and comics, include reviewers' opinions. Often, these reviews also include descriptions of the story. In some cases, these descriptions may spoil later reader's or viewer's enjoyment and excitement. Hereinafter, we call these descriptions spoilers. Spoilers may be related to the position in the story line. In this study we use story documents. Story documents are documents that record all of the details of the given story. Using the story documents, we investigate the location to which the content of the spoilers correspond in the story documents. Based on the result of the investigation, we consider how to detect spoilers in reviewers' comments.

This paper presents VRMixer, a system that mixes real world and a video clip letting a user enter the video clip and realize a virtual co-starring role with people appearing in the clip. Our system constructs a simple virtual space by allocating video frames and the people appearing in the clip within the user's 3D space. By measuring the user's 3D depth in real time, the time space of the video clip and the user's 3D space become mixed. VRMixer automatically extracts human images from a video clip by using a video segmentation technique based on 3D graph cut segmentation that employs face detection to detach the human area from the background. A virtual 3D space (i.e., 2.5D space) is constructed by positioning the background in the back and the people in the front. In the video clip, the user can stand in front of or behind the people by using a depth camera. Real objects that are closer than the distance of the clip's background will become part of the constructed virtual 3D space. This synthesis creates a new image in which the user appears to be a part of the video clip, or in which people in the clip appear to enter the real world. We aim to realize "video reality," i.e., a mixture of reality and video clips using VRMixer.

Analysis and monitoring of system logs such as transaction logs and access logs is important for various objectives including trend discovery, update effort determination, and malicious behavior monitoring. However, it is not always an easy task because these logs may be massive, consisting of millions of records containing tens of variables, and therefore it may be difficult or time-consuming to discover significant knowledge. This paper presents a visual analytics tool which enables us to effectively observe system logs. The tool recommends variables that can reveal interesting discoveries and provides feature-based filtering that selects meaningful items from the visualization results. This paper also presents the result of experiments for non-professional users.

This short paper proposes a method to classify music video clips uploaded to a video sharing service into music mood categories such as 'cheerful,' 'wistful,' and 'aggressive.' The method leverages viewer comments posted to the music video clips for the music mood classification. It extracts specific features from the comments: (1) adjectives in comments, (2) lengthened words in comments, and (3) comments in chorus sections. Our experimental results classifying 695 video clips into six mood categories showed that our method outperformed the baseline in terms of macro and micro averaged F-measures. In addition, our method outperformed the existing approaches that utilize lyrics and audio signals of songs.

Seeing the final score of a sports match on the Web often spoils the pleasure of a user who is waiting to watch a recording of this match on TV. This paper proposes four information clouding methods to block spoiling information, and describes implementation of a system using these methods as a browser extension. We then experimentally investigate the usefulness of the methods, taking into account their differences, differences in the variety of content, and differences in the user's interest in sports.

This paper demonstrates a photo browser which rearranges photos referring to the persons who were close to the photographer when the photos were taken by consulting Bluetooth device detection information. Most of Bluetooth devices accompany their owners. Each photo is tagged with Bluetooth device-IDs which were detected around the moment when it was taken. Employing the tag information, the system classifies the user's photo archive into a layered cluster tree in terms of tag similarity, and shows its user the photos of her-selected cluster on either a map or timelines.

In this paper, we propose a two-stage system using user's eye movements to accommodate the increasing demands to obtain information from the Web in an efficient way. In the first stage the system estimates a user's search intent as a set of weighted terms extracted based on the user's eye movements while browsing Web pages. Then in the second stage, the system shows relevant information to the user by using the estimated intent for re-ranking search results, suggesting intent-based queries, and emphasizing relevant parts of Web pages. The system aims to help users to efficiently obtain what they need by repeating these steps throughout the information seeking process. We proposed four types of search intent estimation methods (MLT, nMLT, DLT and nDLT) considering the relationship among intents, term frequencies and eye movements. As a result of an experiment designed for evaluating the accuracy of each method with a prototype system, we confirmed that the nMLT method works best. In addition, by analyzing the extracted intent terms for eight subjects in the experiment, we found that the system could estimate the unique search intent of each user even if they performed the same search tasks.

This paper proposes a system called "RerankEverything", which enables users to rerank search results in any search service, such as a Web search engine, an e-commerce site, a hotel reservation site, and so on. This system helps users explore diverse search results. In conventional search services, interactions between users and systems are quite limited and complicated. By using RerankEverything, users can interactively explore search results in accordance with their interests by reranking search results from various viewpoints. Experimental results show that our system potentially help users search more proactively. When using our system, users were more likely to click search results that were initially low ranked. Users also browsed through more diverse search results by reranking search results after giving various types of feedback with our system.

In this paper, we propose a method for helping users explore information via Web searches by using a question and answer (Q&A) corpus archived in a community Q&A site. When users do not have clear information needs and have little knowledge about the task domain, it is difficult for them to create queries that adequately reflect their information needs. We focused on terms like "famous temples," "historical townscapes," and "delicious sweets," which we call "adjective facets", and developed a method of extracting these facets from question and answer archives at a community Q&A site. We evaluated the effectiveness of our adjective facets by comparing them with several baselines.

This paper proposes a system called RerankEverything, which enables users to rerank search results in any search service, such as a Web search engine, an e-commerce site, a hotel reservation site and so on. In conventional search services, interactions between users and services are quite limited and complicated. In addition, search functions and interactions to refine search results differ depending on the services. By using RerankEverything, users can interactively explore search results in accordance with their interests by reranking search results from various viewpoints.

This paper proposes a temporal filtering system called the Anti-Spoiler system. The system changes filters dynamically based on user-specified preferences and the user's timetable. The system then blocks contents that would spoil the user's enjoyment of a previously unwatched content. The system analyzes a user-requested Web content, and then uses filters to prevent portions of the content being displayed that might spoil user's enjoyment. For example, the system hides the final score of football from the Web content before watching it on TV.

We implemented blow displays, which provide force feelings with no contact. Although blow displays can use only wind velocities and directions to represent information, they are less intrusive and less visually polluting to other media than other displays. We propose collaborative ambient systems to utilize blow displays' characteristics of spatiality and compatibility with other media. In collaborative ambient systems, blow displays direct the user to displays that provide rich information. Blow displays can also express information auxiliary to the main content the user is attending to. In this paper, we describe some ongoing applications and discuss their benefits and issues.

We describe a browser for the past web. It can retrieve data from multiple past web resources and features a passive browsing style based on change detection and presentation. The browser shows past pages one by one along a time line. The parts that were changed between consecutive page versions are animated to reflect their deletion or insertion, thereby drawing the user's attention to them. The browser enables automatic skipping of changeless periods and filtered browsing based on user specified query.

A method of automatically indexing broadcast content using live chat on the Web is proposed. The live chat is a Web bulletin board where the viewers post messages in sync with a TV program. Statistical analysis and pattern recognition of these messages can effectively extract metadata related to viewer's viewpoints such as important scenes in the program or responses by a particular viewer. Preliminary experiments indicate that the proposed method can efficiently extract metadata such as the intensity of viewers' responses and degree of emotional delight or depression. They also indicate that a prototype TV viewing system using the extracted metadata enables a new way of viewing TV content from different perspectives reflecting viewers' viewpoints.

We propose a method of generating views of TV programs based on viewer's perspectives expressed in live chats on the Web. Important scenes in a program and responses by particular viewers can be extracted efficiently by statistically computing and/or recognizing live chat data obtained in sync with the broadcast content. We show that by using the computed results, views can be generated that indicate the momentum of reactions by viewers and scenes of interest to particular viewers whose preferences are similar to those of the viewer, etc. This is a new way of viewing TV content from various perspectives.