Loading...
Publication
AUTOMATED CYBERSECURITY DOCUMENTATION REPOSITORY
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.49 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A digitalização trouxe consigo inúmeras vantagens, mas à medida que a sociedade
se torna mais dependente dos sistemas informáticos e também que os dados se
tornam mais valiosos, os riscos no ciberespaço vão aumentando de igual forma.
Para mitigar estes riscos, muitas organizações e entidades governamentais emitem
documentos legais e orientações técnicas que devem ser seguidos e implementados.
Embora essenciais para a cibersegurança, estes documentos podem possuir um grau
de complexidade bastante elevado e a sua análise pode ser exaustiva. Além disso,
o número existente deste tipo de documentação aumenta continuamente, sendo
que hoje em dia podemos encontrar documentos técnicos e legais relacionados com
cibersegurança dispersos por vários repositórios na Internet.
Nesta tese, desenvolvemos o “Automated Repository”, que utiliza dois modelos
NLP, GPT-4o e GPT-4o mini, assim como tecnologias de recolha de documentos
e ferramentas de visualização para auxiliar os utilizadores a recolher, organizar,
analisar e extrair informações cruciais da documentação relacionada com a cibersegurança.
O desenvolvimento da aplicação é descrito ao longo deste relatório e as
funcionalidades implementadas são avaliadas face aos objetivos definidos.
Com base na informação recolhida e gerada pela aplicação, foi realizada uma
análise da evolução da documentação de cibersegurança ao longo do tempo, destacandose
os documentos mais relevantes e que mais marcaram a área. Por fim foi possivel
concluir que o “Automated Repository” pode auxiliar significativamente os utilizadores
nas tarefas de recolha e análise de documentação sobre cibersegurança.
Além disso, o sistema pode ser de grande utilidade em outras áreas e setores da
sociedade. Exemplos são as áreas da saúde, ciências sociais, engenharias e muitas
outras onde os profissionais têm de seguir e consultar regularmente grandes volumes
de documentação frequentemente.
Digitalization has introduced numerous advantages, but as society becomes more reliant on computer systems and data becomes increasingly valuable, the risks in cyberspace are increasing accordingly. To mitigate these risks, many organizations and government entities issue legal documents and technical guidelines that must be followed and implemented. Although essential for cybersecurity, these documents can be quite complex and their analysis can be exhaustive. Furthermore, the existing number of this type of documents is continuously increasing, and today it is possible to find technical and legal documentation related to cybersecurity distributed across various repositories on the Internet. In this thesis, we have developed the “Automated Repository”, which uses two NLP models, GPT-4o and GPT4o mini, as well as document collection technologies, and visualization tools to help users collect, organize, analyze, and extract crucial information from cybersecurity-related documentation. The development of the application is described throughout this report and the functionalities implemented are evaluated against the defined objectives. Based on the information collected and generated by the application, an analysis of the evolution of cybersecurity documentation was performed, highlighting the most relevant documents and those that have influenced the area the most over time. Finally, it was possible to conclude that the “Automated Repository” can significantly assist users in the tasks of collecting and analyzing cybersecurity documentation. In addition, the system can be of great use in other areas and sectors of society. Examples are the areas of healthcare, social sciences, engineering and many others where professionals have to regularly follow and consult large volumes of documentation.
Digitalization has introduced numerous advantages, but as society becomes more reliant on computer systems and data becomes increasingly valuable, the risks in cyberspace are increasing accordingly. To mitigate these risks, many organizations and government entities issue legal documents and technical guidelines that must be followed and implemented. Although essential for cybersecurity, these documents can be quite complex and their analysis can be exhaustive. Furthermore, the existing number of this type of documents is continuously increasing, and today it is possible to find technical and legal documentation related to cybersecurity distributed across various repositories on the Internet. In this thesis, we have developed the “Automated Repository”, which uses two NLP models, GPT-4o and GPT4o mini, as well as document collection technologies, and visualization tools to help users collect, organize, analyze, and extract crucial information from cybersecurity-related documentation. The development of the application is described throughout this report and the functionalities implemented are evaluated against the defined objectives. Based on the information collected and generated by the application, an analysis of the evolution of cybersecurity documentation was performed, highlighting the most relevant documents and those that have influenced the area the most over time. Finally, it was possible to conclude that the “Automated Repository” can significantly assist users in the tasks of collecting and analyzing cybersecurity documentation. In addition, the system can be of great use in other areas and sectors of society. Examples are the areas of healthcare, social sciences, engineering and many others where professionals have to regularly follow and consult large volumes of documentation.
Description
Keywords
Automação Cibersegurança Documentação GPT Legislação Modelos NLP Repositório