Yuhki Shiraishi

Simultaneous realization of page-centric communication and search

We present a novel system that combines the advantages of social communication and Web search by simultaneously discovering important pages and users. First, the system provides a communication interface attached to pages, which allows users to talk with each other in real time while browsing the same page, i.e., page-centric communication. Then, the system can efficiently provide two ranking lists of pages and users by analyzing a hybrid structure of hyperlinks (page-page relationship) and social links (page-user relationship and user-user relationship). Thus, users can efficiently search for important pages as well as important users related to their queries through the ranking function, and immediately obtain useful information or knowledge from not only pages themselves but also from other users.

Real-time captioning of sign language by groups of deaf and hard-of-hearing people

In order to realize smooth communication between people with or without difficulty in hearing, we aim to implement an information support system based on crowdsourcing, a problem-solving model, in which numerous people cooperate to accomplish a job. More specifically, we implement a crowdsourced captioning system that enables groups of deaf and hard-of-hearing people to collaboratively interpret a speech or lecture given by sign language to text in real-time. Deaf and hard-of-hearing people are generally considered as the ones who receive support from others, but our proposal helps them become the ones who offer support to others. In this paper, we explore the issues that need be resolved for the practical use of the proposed system through several experiments and succeed in verifying its feasibility.

Group rotation management in real-time crowdsourcing

A common workflow to perform a continuous human task stream is to divide workers into groups, have one group perform the newly-arrived task, and rotate the groups. Usually, more than one worker belongs to each group for improving the quality of task results. We call this type of workflow the group rotation. This paper addresses the problem of how to manage Group Rotation Type Crowdsourcing, the group rotation in a crowdsourcing setting. In the group-rotation type crowdsourcing, we must change the group structure dynamically because workers come in and leave frequently. However, changing the group structure will give workers psychological stress, such as surprise, confusion or irritation. This paper explores a design space for group restructuring algorithms in the group rotation type crowdsourcing and compares implemented strategies in terms of the evaluation results on psychological stress with real-world crowd workers.

Crowdsourced real-time captioning of sign language by deaf and hard-of-hearing people

Purpose The purpose of this paper is to explore the issues on how to achieve crowdsourced real-time captioning of sign language by deaf and hard-of-hearing (DHH) people, such that how a system structure should be designed, how a continuous task of sign language captioning should be divided into microtasks and how many DHH people are required to maintain a high-quality real-time captioning. Design/methodology/approach The authors first propose a system structure, including the new design of worker roles, task division and task assignment. Then, based on an implemented prototype, the authors analyze the necessary setting for achieving a crowdsourced real-time captioning of sign language, test the feasibility of the proposed system and explore its robustness and improvability through four experiments. Findings The results of Experiment 1 have revealed the optimal method for task division, the necessary minimum number of groups and the necessary minimum number of workers in a group. The results of Experiment 2 have verified the feasibility of the crowdsourced real-time captioning of sign language by DHH people. The results of Experiment 3 and Experiment 4 have shown the robustness and improvability of the captioning system. Originality/value Although some crowdsourcing-based systems have been developed for the captioning of voice to text, the authors intend to resolve the issues on the captioning of sign language to text, for which the existing approaches do not work well due to the unique properties of sign language. Moreover, DHH people are generally considered as the ones who receive support from others, but our proposal helps them become the ones who offer support to others.

Searching Comprehensive Web Pages of Multiple Sentiments for a Topic

We have developed a novel system for searching comprehensive Web pages by focusing on multiplicity of sentiments of writers for a topic. Recently, lots of studies and services based on sentiment analysis have been conducted, since it is still difficult to search and summarize information satisfying users’ needs by text analysis only. In this paper, we propose a system for searching and visualizing comprehensive Web pages in terms of sentiments by extracting multiple sentiments of Web pages on a query topic and re-retrieving Web pages using sub-topic keywords. Specifically, this system extracts sentiment features of each Web page using a sentiment dictionary consisting of three sentiment dimensions; “Happy ⇔ Sad,” “Glad ⇔ Angry,” and “Peaceful ⇔ Strained.” Next, in order to conduct a re-retrieval, it extracts sub-topic keywords from Web pages of maximum (or minimum) sentiment features on three sentiment dimensions, respectively. Then, it re-retrieves Web pages using the query topic keyword and the extracted sub-topic keywords. Then, it plots them on sentiment graphs based on their sentiment features. By using the graphs, we can grasp not only sentiment tendency but also comprehensive sentiments for a query topic. In the experiment, we evaluate our proposed method using the developed system.

Web Page Centered Communication System Based on a Physical Property

We have developed a novel communication system that enables ALL users to share information and emotion effectively through ALL Web pages based on a physical property: a user can perceive others browsing the same pages through avatars on the Web browser, like a face to face communication in real world. The system optimize the displayed and communicable users by the relationship among them based on the information searching and sharing activities. Furthermore, the system also enables users to search not only Web pages but appropriate users: we construct a new ranking model by combining the real space information (the user activities) with the Web space information (the hyperlinks). We also verify the effectiveness of the system by the user study experiments in the condition where we suppose the system works especially well, i.e., information retrieval, collaboration tasks, and sharing of feelings.

Personalized Web Search Using Emotional Features

Re-ranking and re-retrieval of search results are useful techniques for satisfying users’ search intentions, since current search engines cannot always return user-desired pages at the top ranks. In this paper, we propose a system for personalized Web search considering users’ emotional aspects. Given a query topic, the system presents the major emotion tendency on this topic that search results returned from search engines are reflecting. The system also enables users to specify the polarities and strengths of their emotions (e.g., happy or sad, glad or angry, peaceful or strained) on this topic and offers a re-ranking list of initial search results based on the similarity of emotions. Particularly, the system can automatically obtain Web pages with minor emotion tendency on the query topic by extracting sub-queries with opposite emotions and conducting a re-retrieval. Experimental evaluations show the re-ranking and the re-retrieval achieve encouraging search results in comparison with initial search results.