For annotation in cancer genomic medicine, oncologists have to refer to various knowledge bases worldwide and retrieve all information (e.g., drugs, clinical trials, and academic papers) related to a gene variant. However, oncologists find it difficult to search these knowledge bases comprehensively because there are multiple paraphrases containing abbreviations and foreign languages in their terminologies including diseases, drugs, and genes. In this paper, we propose a novel search method considering deep paraphrases, which helps oncologists retrieve essential annotation resources swiftly and effortlessly. Our method recursively finds paraphrases based on paraphrase corpora, expands a source document, and finally generates a paraphrase lattice. The proposed method also feedbacks beneficial information regarding the paraphrases applied for a search, which is useful for selecting search results and considering a query for the succeeding search. The results of an experiment demonstrated that our method could retrieve important annotation information that could not be retrieved using a conventional search system and simple paraphrasing. Additionally, annotation experts evaluated our method and found it to be practical.

The purpose of this research is to support users' mealtime communication by changing their utterance rate during meals. By logging and analyzing the utterance rate, or the ratio of utterance length to mealtime length, we determined the relationship between the utterance rate and the distribution of the community to which each meal companion belonged. Using this relationship, we developed a real-time feedback system that presented a pre-estimated utterance rate with rate at the present time during a meal. This utterance rate was estimated according to logged utterance rate data and input data about members at the table. In order to evaluate our system, we asked six users to use it. As a result, we found out that users almost always wanted to increase the utterance rate and users' utterance rates increased as they wanted after using our feedback system.

This study tries to increase the effects of odors by controlling the odor emissions. We developed an olfactory feedback system to maintain user's concentration effectively based on biological information. The system presumes the level of user's concentration based on an electrocardiogram analysis. The analysis is carried out using variance of R-R intervals and frequency spectra. According to the presumed concentration level, the system presents an odor with an awakening effect to maintain user's concentration. We performed experiments in which subjects did simple tasks with this system, changing the method of controlling odor emissions. We compared the results of each method and evaluated the effectiveness of our olfactory feedback system and the methods of controlling odor emissions. The comparison of the results showed that presenting an odor when a user loses his/her concentration is effective to decrease errors of addition tasks.

Olfactory displays which exist now can only present the set of scents which was prepared beforehand because a set of "primary odors" has not been found. In this paper, we focus on development of an olfactory display using cross modality which can represent more patterns of scents than the patterns of scents prepared. Due to cross modal effect between vision and olfaction, human tends to feel not olfactory but visual stimulation as scents. First, we asked subjects to smell various aroma chemicals and evaluate their similarity. Based on the data of similarity among aromas, we built a map of smell distance. Next, we selected a few aroma chemicals from the smell distance map and implemented a visual and olfactory display. Then we conducted an assessment experiment of the display. We presented various pictures and the selected aroma chemicals to subjects and asked them what kind of scent they smelled like. In this experiment, we succeed in giving subjects feeling more patterns of scents than the number of selected aroma chemicals. Also, we succeed in making not olfactory stimulation by aroma chemicals but visual stimulation by pictures as scents. In particular, we find that the visual effect on olfactory sensation is more strong when the distance between the picture and the aroma chemicals is close than when the distance is far.

Controllable particle display has been proposed, that controls the position and blinking patterns to yield a visual representation and the representation to touch these particles. Additionally, as an example of its implementation, a controllable particle display using water drops as the particles was proposed. In this system, objects are represented by projecting images upward onto falling water drops designed to form a plane surface, depending on the positions of the water drops. However, this method has a problem in terms of the brightness of the object. In this paper, we propose a method by which images are projected onto falling water drops at an angle, and users observe the images from in front of a projector.

We discuss application possibilities of virtual reality technologies such as immersive projection technology, to the field of genome science, and the guiding principles for the visualization of genome information are proposed. In addition, we develop a visualization methodology by using immersive projection technology for pair-wise comparison between cluster sets generated from different gene expression datasets. Our methodology could display the distribution of overlaps between two hierarchical cluster sets based on hepatocellular carcinomas and hepatoblastomas.

In this paper we discuss application possibilities of virtual reality technology such as immersive projection technology to the field of genome science. The prototype of the visualization environment is implemented and used in analysis to elucidate the important genes in categorizing liver cancer.