Kosei Fume

Table structure analysis based on cell classification and cell modification for XML document transformation

A new method of table structure analysis based on cell classification and cell modification is proposed in this paper as the basis of an OCR which can convert a variety of printed tables into XML documents in accordance with a specified XML schema. The outline of this method is described as follows. Firstly, cell features defined by ruled lines, which correspond to data fields, are extracted from the input image of a table. After that, each cell is classified to identify the irregular table whose ruled lines are not gridded and is modified to form regular cell arrangement. Next, the hierarchical table structure consisting of a regular row structure of cells is extracted from the modified regular table and is described using a DOM tree. In this case, logical objects within a cell are extracted and are converted into a sub-tree in the DOM tree. Finally, this DOM tree is transformed into a target XML document by an XML parser with information extraction process. Experimental results show the method is effective in transforming various printed tables to various XML documents.

TTS-Based DAISY Content Creation System: Implementation and Evaluation of DaisyRingsTM

Digital Accessible Information System (DAISY) content is expected to gain popularity gradually among the visually impaired according to the prevalence of e-book reading devices and the development of text-to-speech (TTS) technology. However, the development of DAISY-formatted e-books, which is undertaken by volunteers, is a time-consuming process, making it difficult to meet end user requirements. In this report, we propose a content transliteration system that can convert plain text to DAISY content including formatted HTML and audio data via automatic TTS technology. Furthermore, using the graphic user interface of the proposed system, users can correct text and accent information by inputting ruby-type data. Through this functionality, we aim to target support for transliteration workers such as volunteers, teachers, and parents to make and edit contents easily and quickly for the visually impaired. Finally, we present the results of a preliminary evaluation using the proposed method in order to compare it with the conventional method.

Implementation of Automatic Captioning System to Enhance the Accessibility of Meetings

In terms of information accessibility tools for hearing-impaired people, in order to understand meetings, expectations for real-time captioning utilizing speech recognition technology are increasing, from manual handwritten abstracts. However, it is still difficult to provide automatic closed captioning with a practical level of accuracy stably, without regard to various speakers and content. Therefore, we develop a web-based real-time closed captioning system that is easy to use in contact conferences, lectures, forums, etc., through trial and feedback from hearing-impaired people in the company. In this report, we outline this system as well as the results of a simple evaluation conducted inside and outside the company.

Practical Evaluation of DaisyRings: Text-to-Speech-Based DAISY Content Creation System

The Digital Accessible Information System (DAISY) content is gradually gaining popularity among the visually impaired, according to “The Act on the Elimination of Discrimination against Persons with Disabilities” enforced in Japan since April 2016. However, compilation of DAISY-formatted e-books, generally undertaken by volunteers, is time-consuming, and forces users to wait a long time for the delivery of such books, which they might want to read as soon as possible. Therefore, we attempted to improve this situation by providing a text-to-speech (TTS)-based DAISY content creation tool to volunteers. To practically evaluate our concept, we ran a web application site from Fall 2013 to the end of 2015. The site provides a transliteration system that converts plain text to DAISY content, including formatted HTML and audio data, by using TTS technology. The results of the practical evaluation were demonstrated in this study, and show the effectiveness of our proposal method.

Pattern-based semantic tagging for ontology population

Ontology population has emerged as an increasingly importantproblem in semantic web services. In this paper, we propose a method usingnamed entity recognition that extracts keywords from Web pages in order topopulate a product ontology. The semantic classification determines meaningsof terms and phrases by heuristic rules after the morphological analysis. Inaddition, our method classifies vocabularies into different semantic tags. Firstly,it records several lists of semantic tags to a history database. Then, we definesome rules from the lists to extract a product name. Finally, the rules build andrefine the product ontology semi-automatically. According to an evaluation,proposed method achieved 87.1% precision and 87.4% recall. Thus, it cansuggest some instances, and it decreases cost of updating the ontology.

Model-based document categorization employing semantic pattern analysis and local structure clustering

We propose a document categorization method based on a document model that can be defined externally for each task and that categorizes Web content or business documents into a target category in accordance with the similarity of the model. The main feature of the proposed method consists of two aspects of semantics extraction from an input document. The semantics of terms are extracted by the semantic pattern analysis and implicit meanings of document substructure are specified by a bottom-up text clustering technique focusing on the similarity of text line attributes. We have constructed a system based on the proposed method for trial purposes. The experimental results show that the system achieves more than 80% classification accuracy in categorizing Web content and business documents into 15 or 70 categories.