East Asia Disability Studies Forum (EADSF) 2023 Oct 27th, 28th(Fri, Sat) 2023, at E-room Center, Korea A Report on the Role of Live Captioning and Transcripts as a Form of Information Accessibility Based on Our Practice YASUDA, Masayuki (安田真之); KUBOSAKI, Yasuki (窪崎泰紀) from UNI, a Specified Non-Profit Organization (特定非営利活動法人ゆに) 0. Overview UNI, based in Kyoto, Japan, offers live captioning services for lectures, training, and other events. This service enables users, especially those with hearing impairments, to receive audio information in text form. The main goal of live captioning is to enable participants who have difficulty acquiring audio information to participate in these events. Event organizers also often request transcript text data generated by live captioning for use as a textual record of the event. The text obtained through live captioning and the documents created after an event for recording or archiving the project contents differ in their purpose and characteristics. There is concern that by focusing on live captioning as an effort to simply convert audio information into text, its original purpose and role will be lost. In this report, we discuss the issues regarding live captioning practices implemented by UNI and the role of live captioning as an element of information accessibility. 1. Live Captioning at UNI (Overview, Purpose) (1) What is UNI? ・A non-profit organization that helps disabled students with their studies and daily activities. ・Its mission is to help disabled students "do whatever they want to do" and "go wherever they want to go". One of its activities is providing live captioning services for people with hearing impairments. (2) What is live captioning? ・An attempt to make sure that people with hearing impairments can participate in events. ・Multiple captioners type audio information such as lectures, Q&A sessions, class start/end chimes, etc. into computers and display it as text information to hearing-impaired users. ・Captioners do not just convert "spoken words" they hear into "written words", but rather transform them into "written information" that allows hearing-impaired users to see and understand the content. The main targets of live captioning by UNI are university and high school classes and events, disability welfare related events, corporate in-house training, etc. 2. Transcripts Generated in Live Captioning (1) What is a transcript text? Text data displayed in live captioning → Usually used for verification of live captioning accuracy or for users to supplement audio information later. (2) Handling of text data in Japanese live captioning and UNI ・ Various views and arguments exist regarding the handling of transcripts. ・ The following are examples of views from some organizations. Zen-nancho, Zen-yoken *1 Live captioning is an interpreting act that converts spoken language into written language. Therefore, like language interpretation, it is something that is completed within the event, and after the interpretation is over, please consider it as something that has disappeared like speech. When using live captioning, we set the system to not leave text data. The roll and paper are the result of the interpreting act and are not something that can be used for secondary purposes. If the user wishes to keep a written record of the content, please prepare a recording method other than captioning. Mojitsuken; from a lecture *2 Documents and text data are generated in captioning. We think this is a good aspect of live captioning. It creates new value for the people who need it. In the case of people with hearing impairments, they look at sign language interpretation and live captioning. It is difficult for them to take notes on their own. The text data and paper records are also valuable because they assist them in this aspect. They can also be used when users take minutes or prepare reports. UNI *3 Basically, our approach is to adhere to the guidelines of the requesting educational institution. We welcome the use of transcripts if it allows users to more easily participate in the event. If there is a request for transcripts in advance, we will provide them. If there is no request for transcripts we will dispose of them at the end of the event for confidentiality reasons. We do not recommend using them as a record, but we will provide them after explaining the characteristics of live captioning. ・ While having an aspect like "interpretation", it also has the aspect of remaining as "text", unlike speech that disappears at the moment of interpretation. There are differences in how these aspects are perceived and handled. *1 From documents published by the All Japan Association of Hard of Hearing and Late-Deafened People (全日本難聴者・中途失聴者団体連合会) and, the National Association for the Study of the Problem of Captioning (全国要約筆記問題研究会) *2 From a lecture transcript from the National Association of the Study of Text Interpreting (全国文字通訳研究会) (3) Requests for provision of transcripts from the event organizers Classification of the purposes of 189 requests for transcript provision to UNI from April 2021 to August 2023. Post-hoc information complementation for live captioning users: 70 cases; Verification of live captioning: 5 cases; Use as a text record of the content: 114 cases. (Pie chart of the data above) (4) Problem: the incompatibility of live captioning being both "interpretation" and "written record" ・It is difficult to reconcile "participation of users" and "making a record" using the same text: "Prioritizing making a record" may interfere with "participation of users". ・Differences between live captioning text and text records. | | Transcripts from live captioning | Text records | | - | - | - | | Purpose | Guarantee of participation (ex-post information supplementation), verification of participation guarantee | Recording of contents | | When used | Simultaneous with audio | After the end of the event | | Source of information | Audio information from lectures | Audio information from lectures | | Format | Text data | Text data | | Priority | Guarantee of participation, immediacy | Accuracy | | Inaccurate information | Included | Not included | | Summaries, supplementary information, paraphrasing | Included to a considerable extent | Included but only minimally | ・The first priority in live captioning is to ensure users are able to participate in the event, so "accuracy as a record" is not a priority. ・In live captioning, priority is given to accuracy as a caption, quickness, and readability. These are essential to participation. ・Live captioning may include the substitution of specific words for demonstratives, organization of repetitions and restatements, and reorganization of word order. 3. Current Situation and Concerns (1) One-sided understanding of live captioning: "live captioning = converting spoken words into written text" → The purpose and essential aspects of information accessibility are not properly understood, and the provision of transcripts in live captioning is requested without understanding. (2) Absence of an understanding of the needs of users ・The majority of live captioning requests for "UNI" are made by event organizers who are not disabled. ・More than half of the requests for text data are for use as a record of the content of the events. → Transcripts are provided without sufficient regard for the user's needs. (3) Concerns about losing sight of the purpose ・One-sided understanding → Excessive emphasis on transcripts by event organizers → Incompatibility of "live captioning" and "text record". → Risk of not enabling the "participation" of the users. 4. Future Practical Issues (1) Raising awareness of live captioning as a guarantee of rights ・Form a common understanding that live captioning = an effort to provide accessibility assurance → Utilization of transcripts as textual records = secondary aim ・The presence of live captioning alone does not ensure accessibility; it is also necessary for speakers and planners to be aware of the need for the effective use of live captioning. (2) Participation of users → Practice live captioning that does not leave the specific needs of the users behind (3) Organizing the role of live captioning in terms of "diversity" and "universal design" Live captioning is effective for ALL participants, not only people with hearing impairments. → While this fact may support the use of live captioning, there are concerns that the "specific needs" of people with hearing impairments may be left behind because the organizers (who are not people with hearing impairments) often make requests for live captioning services. 5.References 1. 一般社団法人 全日本難聴者・中途失聴者団体連合会 特定非営利活動法人 全国要約筆記問題研究会, 2016, 「要約筆記利用時のロールや用紙、ログの扱いについて」, (Accessed 23 Sept. 2023, https://zenyouken.jp/wp-content/uploads/2016/11/161107.pdf) 2. 全国文字通訳研究会, 2019, 「第 9 回パソコン文字通訳シンポジウム 講演録」, (Accessed 23 Sept. 2023, http://mojitsuken.sakura.ne.jp/wp/wp-content/uploads/2019/12/19-01-27-sympo-report.pdf) 3. 特定非営利活動法人ゆに, 2023, 「文字情報保障サービスの利用にあたって」, (Accessed 23 Sept. 2023, https://www.unikyoto.com/project/captioning/livecaptioning_usagepolicy/)

AltStyle によって変換されたページ (->オリジナル) /