Toshinari Itoko

Age-Based task specialization for crowdsourced proofreading

Crowdsourcing can efficiently produce accessible digital books for people with print disabilities. However, particularly in Japan, the proofreading step tends to be expensive because of language-related issues. The elderly population is a promising source of proofreaders. Our surveys found that they have strong linguistic skills and want to contribute to society. So why do they rarely participate in Internet-based work scenarios such as crowdsourcing? We introduce a collaborative crowdsourcing model that aims to fully utilize the linguistic skills of the elderly by encouraging younger people to support the elderly in overcoming their limited technical skills. We decompose each proofreading task into several types of sub-tasks, where some tasks require more linguistic skills while the other tasks need more technical skills, so that the linguistic and technical tasks can be distributed to older and younger participants, respectively. We also discuss other scenarios that may be suitable for such multi-generational crowdsourcing model.

Involving Senior Workers in Crowdsourced Proofreading

Seniors have a wealth of knowledge and free time, so they are a promising workforce for crowdsourced tasks. Currently senior workers are hardly involved in real applications. We have started an experimental project that crowdsources proofreading micro-tasks to volunteer workers to efficiently produce accessible digital books. By design, the majority of the workers in this project are senior citizens. In this paper, we report the findings of our experiment in which we tested four working hypotheses about the behavioral characteristics of senior workers. We also discuss skill management to improve task performance and motivation encouragement for long-term involvement of senior workers.

Transforming Japanese archives into accessible digital books

Digitized physical books offer access to tremendous amounts of knowledge, even for people with print-related disabilities. Various projects and standard activities are underway to make all of our past and present books accessible. However digitizing books requires extensive human efforts such as correcting the results of OCR (optical character recognition) and adding structural information such as headings. Some Asian languages need extra efforts for the OCR errors because of their many and varied character sets. Japanese has used more than 10,000 characters compared with a few hundred in English. This heavy workload is inhibiting the creation of accessible digital books. To facilitate digitization, we are developing a new system for processing physical books. We reduce and disperse the human efforts and accelerate conversions by combining automatic inference and human capabilities. Our system preserves the original page images for the entire digitization process to support gradual refinement and distributes the work as micro-tasks. We conducted trials with the Japanese National Diet Library (NDL) to evaluate the required effort for digitizing books with a variety of layouts and years of publication. The results showed old Japanese books had specific problems when correcting the OCR errors and adding structures. Drawing on our results, we discuss further workload reductions and future directions for international digitization systems.

Computational Geometric Approach to Submodular Function Minimization for Multiclass Queueing Systems

This paper presents an efficient algorithm for minimizing a certain class of submodular functions that arise in analysis of multiclass queueing systems. In particular, the algorithm can be used for testing whether a given multiclass M/M/1 system achieves a required average performance by an appropriate control policy. With the aid of the topological sweeping method for line arrangement, our algorithm runs in O(n) time, where n is the cardinality of the ground set. This is much faster than direct applications of general submodular function minimization algorithms.

Online Learning for Long-Query Reduction in Interactive Search for Experienced Workers

For domain specific document searches like job matching, long queries are often given as a detailed information of targets. Previous studies found that higher quality results can be obtained by searching with an optimal subset of words excerpted from a long query. To excerpt the optimal subset of words, query reduction using machine learning techniques has been studied. Supervised learning requires training data with annotation, which is especially difficult for in-domain data because of its specific terminology. In this study, we propose a model that integrates machine learning techniques and manual processing for long-query reduction. We integrated our model into a job matching system that collects manual “interactions” and used them as training data to learn query reduction. Furthermore, we evaluated our model with actual job offerings and expert profile data obtained from a recruitment agency. We found that our proposed model outperformed the baseline in precision, recall, and F-measure. The result suggests that our model could be used for query reduction of interactive search systems of specific domain data.

Work Motivating Factors of the Communications in a Crowd-Powered Microvolunteering Site

Various systems have been developed to support and motivate volunteer activities for people with disabilities. “Minna de DAISY” is a microvolunteering system, powered by an open-source social networking system and web application, of character corrections. This system demonstrated that digital books could be efficiently produced for the visually impaired through the manual correction of errors using volunteers. According to participant interviews, those who were aware of their social contribution had a higher motivation to do social work. The feeling of community involvement promoted continuous participation. However, these studies do not precisely discuss the contribution of communication between system managers and participants. Some literature reported that interactions in a social network improved participant awareness of the purpose of their work. In this study, we aim to demonstrate the effect of communication on participant motivation and likelihood to continue microvolunteering.