Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts

Detalhes bibliográficos
Autor(a) principal: Silva, Jorge Sassaki Resende
Data de Publicação: 2021
Tipo de documento: Dissertação
Idioma: eng
Título da fonte: Repositório Institucional da UFLA
Texto Completo: http://repositorio.ufla.br/jspui/handle/1/49017
Resumo: The evolution in information access caused by the Internet has expanded access to information for everyone. Consequently, the availability of information sources and teaching content on the Web has broadened access to knowledge. Still, people with disabilities face accessibility barriers. In this work, we focus on people with visual impairment and their problems when reading long documents. Even with assistive technologies such as screen readers, consulting the Web when seeking information can be exhausting for this group. They lack options for speed reading or skimming large amounts of text since these tools read texts aloud in sequence. The commands available in screen readers through shortcuts helped users create strategies when searching for information, such as increased reading speed and using webpage markup to navigate (e.g. headers or paragraphs). However, they often come at the cost of the cognitive load caused by the attention needed to understand and remember the information at high speeds. Navigating inside texts also proved to be a problem. Studies have shown that a common complaint of visually impaired people is having to create a text map to re-find information mentally. To counter these problems, we have conducted a systematic mapping to gather previous work that aimed to help screen reader users when using the Web and categorize them based on their proposed approach. The methods encountered were Content Filtering, Text Reduction, Navigation, Concurrent Speech, Auditory Overview and Recommendation Systems. Based on this analysis, we observed a lack of topicalization techniques and their effects on navigation for blind people. Thus, the goal of this work was to propose an algorithm to generate headers aiming to help users in information-seeking tasks automatically. The algorithm was divided into two tasks: segmenting a document into topic segments and labelling a text segment. We adapted the C99 segmenting algorithm to use BERT and improved error rates for long texts. Then, the study followed with the implementation of a labelling algorithm based on keywords, and labels are made of words from the text segment that were ranked according to repetition. We conducted a user study with 8 participants and a prototype composed of preprocessed texts 720-1131 words long to test the algorithm. Users had to answer questions based on the information in these texts for comparison in two scenarios: one with automatically generated headers and the other without. We measured the time taken in each text and the cognitive load participants perceived while completing it. A post-test interview was also conducted to gather feedback. Our analysis could not confirm our hypothesis significantly due to a small volunteer sample, but interviews indicated users benefited from the proposed tool. Either by helping navigate inside the text, or re-finding information, participants agreed they would like to have this tool in their screen readers. With this work, we provide design implications and alternatives to implement a plugin for the screen reader.
id UFLA_eb7912275adb0c28064a9fc760a25707
oai_identifier_str oai:localhost:1/49017
network_acronym_str UFLA
network_name_str Repositório Institucional da UFLA
repository_id_str
spelling Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based textsCabeçalhos automaticamente gerados como mecanismo de leitura rápida para usuários cegos usando softwares leitores de tela em textos não-marcados na webAcessibilidade digitalDeficiência visualProcessamento de linguagem naturalLeitores de telaDigital accessibilityBlind peopleNatural language processingScreen readerCiência da ComputaçãoThe evolution in information access caused by the Internet has expanded access to information for everyone. Consequently, the availability of information sources and teaching content on the Web has broadened access to knowledge. Still, people with disabilities face accessibility barriers. In this work, we focus on people with visual impairment and their problems when reading long documents. Even with assistive technologies such as screen readers, consulting the Web when seeking information can be exhausting for this group. They lack options for speed reading or skimming large amounts of text since these tools read texts aloud in sequence. The commands available in screen readers through shortcuts helped users create strategies when searching for information, such as increased reading speed and using webpage markup to navigate (e.g. headers or paragraphs). However, they often come at the cost of the cognitive load caused by the attention needed to understand and remember the information at high speeds. Navigating inside texts also proved to be a problem. Studies have shown that a common complaint of visually impaired people is having to create a text map to re-find information mentally. To counter these problems, we have conducted a systematic mapping to gather previous work that aimed to help screen reader users when using the Web and categorize them based on their proposed approach. The methods encountered were Content Filtering, Text Reduction, Navigation, Concurrent Speech, Auditory Overview and Recommendation Systems. Based on this analysis, we observed a lack of topicalization techniques and their effects on navigation for blind people. Thus, the goal of this work was to propose an algorithm to generate headers aiming to help users in information-seeking tasks automatically. The algorithm was divided into two tasks: segmenting a document into topic segments and labelling a text segment. We adapted the C99 segmenting algorithm to use BERT and improved error rates for long texts. Then, the study followed with the implementation of a labelling algorithm based on keywords, and labels are made of words from the text segment that were ranked according to repetition. We conducted a user study with 8 participants and a prototype composed of preprocessed texts 720-1131 words long to test the algorithm. Users had to answer questions based on the information in these texts for comparison in two scenarios: one with automatically generated headers and the other without. We measured the time taken in each text and the cognitive load participants perceived while completing it. A post-test interview was also conducted to gather feedback. Our analysis could not confirm our hypothesis significantly due to a small volunteer sample, but interviews indicated users benefited from the proposed tool. Either by helping navigate inside the text, or re-finding information, participants agreed they would like to have this tool in their screen readers. With this work, we provide design implications and alternatives to implement a plugin for the screen reader.A evolução da Internet ampliou a disponibilidade de informação e de conteúdos didáticos na Web ampliando o acesso ao conhecimento. Mesmo assim, as pessoas com deficiência enfrentam problemas de acessibilidade. Neste trabalho, focamos em pessoas com deficiência visual e seus problemas na leitura de documentos longos. Mesmo com leitores de tela, buscar informações na Web pode ser uma tarefa exaustiva. Faltam opções para leitura rápida de grandes quantidades de texto, uma vez que leitores de tela lêem o texto de forma sequencial. Os comandos disponíveis por meio de atalhos ajudam os usuários a criar estratégias na busca de informações, como aumento da velocidade de leitura e uso de marcação de página da web para navegar. Mas muitas vezes eles têm um impacto na carga cognitiva por causa da atenção necessária para entender e lembrar informações em alta velocidade. Navegar dentro de textos também é um problema, estudos mostram que um problema comum é ter que mapear mentalmente as informações do texto. Pensando nisso, realizamos um mapeamento sistemático para reunir trabalhos anteriores que visavam ajudar os usuários de leitores de tela no uso da Web, e os categorizamos com base em suas estratégias. Os métodos encontrados foram Filtragem de Conteúdo, Redução de Texto, Navegação, Fala Simultânea, Visão Geral Auditiva e Sistemas de Recomendação. Observamos uma falta de trabalhos com topicalização e seus efeitos na navegação de cegos. Assim, o objetivo deste trabalho foi propor um algoritmo para gerar cabeçalhos automaticamente para auxiliar usuários na busca de informação. O algoritmo foi dividido em duas tarefas: segmentar um documento por tópicos e rotular estes segmentos. Adaptamos o algoritmo de segmentação C99 para usar o BERT e observamos melhora nas taxas de erro para textos longos. Em seguida, foi implementado um algoritmo de rotulagem baseado em palavras-chave, os rótulos são feitos das palavras mais repetidas no documento. Para testar o algoritmo, conduzimos um estudo de usuário com 8 participants e um protótipo composto de 4 textos pré-processados de 720-1131 palavras. Os usuários tinham que responder a conjuntos de perguntas com base nas informações desses textos, para comparação em dois cenários: um com cabeçalhos gerados automaticamente e outro sem. Medimos o tempo gasto em cada texto e a carga cognitiva que os participantes sentiram ao realizar as tarefas. Uma entrevista pós-teste também foi conduzida para coletar feedback. Nossa análise não pôde confirmar nossa hipótese com alta significância devido à pequena amostra de voluntários, mas as entrevistas indicaram que os usuários se beneficiaram com a ferramenta proposta. Seja ajudando a navegar dentro do texto ou reencontrando informações, os participantes concordaram que gostariam de ter essa ferramenta em seus leitores de tela. Com este trabalho, fornecemos implicações de design e alternativas para implementar um plugin para leitor de tela.Universidade Federal de LavrasPrograma de Pós-Graduação em Ciência da ComputaçãoUFLAbrasilDepartamento de Ciência da ComputaçãoFreire, André PimentaCardoso, Paula Christina FigueiraPardo, Thiago Alexandre SalgueiroValentim, Natasha Malveira CostaSilva, Jorge Sassaki Resende2022-01-25T17:46:10Z2022-01-25T17:46:10Z2022-01-252021-11-23info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfSILVA, J. S. R. Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts. 2021. 90 p. Dissertação (Mestrado em Ciência da Computação) – Universidade Federal de Lavras, Lavras, 2022.http://repositorio.ufla.br/jspui/handle/1/49017enginfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da UFLAinstname:Universidade Federal de Lavras (UFLA)instacron:UFLA2023-04-13T17:41:26Zoai:localhost:1/49017Repositório InstitucionalPUBhttp://repositorio.ufla.br/oai/requestnivaldo@ufla.br || repositorio.biblioteca@ufla.bropendoar:2023-04-13T17:41:26Repositório Institucional da UFLA - Universidade Federal de Lavras (UFLA)false
dc.title.none.fl_str_mv Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
Cabeçalhos automaticamente gerados como mecanismo de leitura rápida para usuários cegos usando softwares leitores de tela em textos não-marcados na web
title Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
spellingShingle Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
Silva, Jorge Sassaki Resende
Acessibilidade digital
Deficiência visual
Processamento de linguagem natural
Leitores de tela
Digital accessibility
Blind people
Natural language processing
Screen reader
Ciência da Computação
title_short Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
title_full Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
title_fullStr Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
title_full_unstemmed Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
title_sort Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts
author Silva, Jorge Sassaki Resende
author_facet Silva, Jorge Sassaki Resende
author_role author
dc.contributor.none.fl_str_mv Freire, André Pimenta
Cardoso, Paula Christina Figueira
Pardo, Thiago Alexandre Salgueiro
Valentim, Natasha Malveira Costa
dc.contributor.author.fl_str_mv Silva, Jorge Sassaki Resende
dc.subject.por.fl_str_mv Acessibilidade digital
Deficiência visual
Processamento de linguagem natural
Leitores de tela
Digital accessibility
Blind people
Natural language processing
Screen reader
Ciência da Computação
topic Acessibilidade digital
Deficiência visual
Processamento de linguagem natural
Leitores de tela
Digital accessibility
Blind people
Natural language processing
Screen reader
Ciência da Computação
description The evolution in information access caused by the Internet has expanded access to information for everyone. Consequently, the availability of information sources and teaching content on the Web has broadened access to knowledge. Still, people with disabilities face accessibility barriers. In this work, we focus on people with visual impairment and their problems when reading long documents. Even with assistive technologies such as screen readers, consulting the Web when seeking information can be exhausting for this group. They lack options for speed reading or skimming large amounts of text since these tools read texts aloud in sequence. The commands available in screen readers through shortcuts helped users create strategies when searching for information, such as increased reading speed and using webpage markup to navigate (e.g. headers or paragraphs). However, they often come at the cost of the cognitive load caused by the attention needed to understand and remember the information at high speeds. Navigating inside texts also proved to be a problem. Studies have shown that a common complaint of visually impaired people is having to create a text map to re-find information mentally. To counter these problems, we have conducted a systematic mapping to gather previous work that aimed to help screen reader users when using the Web and categorize them based on their proposed approach. The methods encountered were Content Filtering, Text Reduction, Navigation, Concurrent Speech, Auditory Overview and Recommendation Systems. Based on this analysis, we observed a lack of topicalization techniques and their effects on navigation for blind people. Thus, the goal of this work was to propose an algorithm to generate headers aiming to help users in information-seeking tasks automatically. The algorithm was divided into two tasks: segmenting a document into topic segments and labelling a text segment. We adapted the C99 segmenting algorithm to use BERT and improved error rates for long texts. Then, the study followed with the implementation of a labelling algorithm based on keywords, and labels are made of words from the text segment that were ranked according to repetition. We conducted a user study with 8 participants and a prototype composed of preprocessed texts 720-1131 words long to test the algorithm. Users had to answer questions based on the information in these texts for comparison in two scenarios: one with automatically generated headers and the other without. We measured the time taken in each text and the cognitive load participants perceived while completing it. A post-test interview was also conducted to gather feedback. Our analysis could not confirm our hypothesis significantly due to a small volunteer sample, but interviews indicated users benefited from the proposed tool. Either by helping navigate inside the text, or re-finding information, participants agreed they would like to have this tool in their screen readers. With this work, we provide design implications and alternatives to implement a plugin for the screen reader.
publishDate 2021
dc.date.none.fl_str_mv 2021-11-23
2022-01-25T17:46:10Z
2022-01-25T17:46:10Z
2022-01-25
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
status_str publishedVersion
dc.identifier.uri.fl_str_mv SILVA, J. S. R. Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts. 2021. 90 p. Dissertação (Mestrado em Ciência da Computação) – Universidade Federal de Lavras, Lavras, 2022.
http://repositorio.ufla.br/jspui/handle/1/49017
identifier_str_mv SILVA, J. S. R. Automatically generated headers as text-skimming mechanisms for blind users using screen reading software in unmarked web-based texts. 2021. 90 p. Dissertação (Mestrado em Ciência da Computação) – Universidade Federal de Lavras, Lavras, 2022.
url http://repositorio.ufla.br/jspui/handle/1/49017
dc.language.iso.fl_str_mv eng
language eng
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universidade Federal de Lavras
Programa de Pós-Graduação em Ciência da Computação
UFLA
brasil
Departamento de Ciência da Computação
publisher.none.fl_str_mv Universidade Federal de Lavras
Programa de Pós-Graduação em Ciência da Computação
UFLA
brasil
Departamento de Ciência da Computação
dc.source.none.fl_str_mv reponame:Repositório Institucional da UFLA
instname:Universidade Federal de Lavras (UFLA)
instacron:UFLA
instname_str Universidade Federal de Lavras (UFLA)
instacron_str UFLA
institution UFLA
reponame_str Repositório Institucional da UFLA
collection Repositório Institucional da UFLA
repository.name.fl_str_mv Repositório Institucional da UFLA - Universidade Federal de Lavras (UFLA)
repository.mail.fl_str_mv nivaldo@ufla.br || repositorio.biblioteca@ufla.br
_version_ 1807835188191821824