C.L.A.I.T.O.N - Chatbot Legal Artificial Intelligence Technology On Network
| dc.contributor.advisor | Pereira, Clayton Reginaldo [UNESP] | |
| dc.contributor.author | Roratto, Pedro Henrique [UNESP] | |
| dc.contributor.committeeMember | Napolitano, Carlo Jose [UNESP] | |
| dc.contributor.committeeMember | Brega, José Remo Ferreira [UNESP] | |
| dc.contributor.institution | Universidade Estadual Paulista (Unesp) | |
| dc.date.accessioned | 2025-12-15T17:09:31Z | |
| dc.date.issued | 2025-12-02 | |
| dc.description.abstract | O presente Trabalho de Conclusão de Curso apresenta o desenvolvimento do C.L.A.I.T.O.N (Chatbot Legal Artificial Intelligence Technology on Network), um assistente jurídico voltado ao Direito Penal brasileiro, fundamentado em técnicas de Inteligência Artificial e Processamento de Linguagem Natural. O sistema tem como objetivo oferecer respostas fundamentadas e contextualizadas a consultas jurídicas, integrando consulta semântica, recuperação de documentos e geração de linguagem natural com base em fontes verificáveis. A solução adota a arquitetura de Recuperação Aumentada por Geração, utilizando o LangChain como framework principal. Os documentos jurídicos, compostos por acórdãos do STF, do STJ, do TJSP e pela legislação penal, são processados e indexados no banco vetorial ChromaDB, por meio de embeddings da família E5. As respostas são geradas localmente com o Ollama (Llama 3), o que contribui para a privacidade e o desempenho do sistema. O sistema disponibiliza dois canais de acesso. A interface Web, desenvolvida em Streamlit, oferece recursos de explicabilidade e visualização de fontes. O bot para WhatsApp, integrado via Twilio API, atende ao uso cotidiano com praticidade. As respostas apresentam as fontes originais, com metadados e escores de relevância, promovendo transparência e auditabilidade. Os resultados obtidos demonstram a viabilidade da integração entre recuperação semântica, bancos vetoriais e modelos de linguagem de grande porte no domínio jurídico penal. O C.L.A.I.T.O.N contribui para a modernização da pesquisa e da prática jurídica, evidenciando o potencial da Inteligência Artificial no apoio à análise e à interpretação de normas e precedentes. | pt |
| dc.description.abstract | This Final Course Project presents the development of C.L.A.I.T.O.N (Chatbot Legal Artificial Intelligence Technology on Network), a legal assistant focused on Brazilian Criminal Law, based on Artificial Intelligence and Natural Language Processing techniques. The system aims to offer well-founded and contextualized answers to legal queries, integrating semantic query, document retrieval, and natural language generation based on verifiable sources. The solution adopts an Augmented Generation Retrieval architecture, using LangChain as the main framework. Legal documents, composed of rulings from the STF (Supreme Federal Court), STJ (Superior Court of Justice), TJSP (Court of Justice of São Paulo), and criminal legislation, are processed and indexed in the ChromaDB vector database, using E5 family embeddings. Responses are generated locally with Ollama (Llama 3), which contributes to the system's privacy and performance. The system provides two access channels. The web interface, developed in Streamlit, offers explainability and source visualization features. The WhatsApp bot, integrated via the Twilio API, caters to everyday use with practicality. Responses present the original sources, with metadata and relevance scores, promoting transparency and auditability. The results obtained demonstrate the viability of integrating semantic retrieval, vector databases, and large-scale language models in the criminal law domain. C.L.A.I.T.O.N contributes to the modernization of legal research and practice, highlighting the potential of Artificial Intelligence in supporting the analysis and interpretation of norms and precedents. | en |
| dc.description.sponsorshipId | Não recebi financiamento. | |
| dc.identifier.citation | RORATTO, Pedro Henrique. C.L.A.I.T.O.N - Chatbot Legal Artificial Intelligence Technology On Network. 2025. Trabalho de Conclusão de Curso (Bacharelado em Sistemas de Informação) – Faculdade de Ciências, Universidade Estadual Paulista (UNESP), Bauru, 2025. | |
| dc.identifier.uri | https://hdl.handle.net/11449/317206 | |
| dc.language.iso | por | |
| dc.publisher | Universidade Estadual Paulista (Unesp) | |
| dc.rights.accessRights | Acesso aberto | pt |
| dc.subject | Inteligência artificial | pt |
| dc.subject | Processamento de linguagem natural (Computação) | pt |
| dc.subject | Agentes inteligentes (Software) | pt |
| dc.subject | Direito penal | pt |
| dc.subject | Jurisprudência | pt |
| dc.title | C.L.A.I.T.O.N - Chatbot Legal Artificial Intelligence Technology On Network | pt |
| dc.title.alternative | C.L.A.I.T.O.N - Chatbot Legal Artificial Intelligence Technology On Network | en |
| dc.type | Trabalho de conclusão de curso | pt |
| dspace.entity.type | Publication | |
| relation.isAuthorOfPublication | 039e4083-6de8-4325-919c-983217c11722 | |
| relation.isAuthorOfPublication.latestForDiscovery | 039e4083-6de8-4325-919c-983217c11722 | |
| relation.isOrgUnitOfPublication | aef1f5df-a00f-45f4-b366-6926b097829b | |
| relation.isOrgUnitOfPublication.latestForDiscovery | aef1f5df-a00f-45f4-b366-6926b097829b | |
| relation.isUndergradCourseOfPublication | df1908ee-c0c8-4434-8a19-2e30b9824084 | |
| relation.isUndergradCourseOfPublication.latestForDiscovery | df1908ee-c0c8-4434-8a19-2e30b9824084 | |
| unesp.campus | Universidade Estadual Paulista (UNESP), Faculdade de Ciências, Bauru | pt |
| unesp.examinationboard.type | Banca pública | pt |
| unesp.undergraduate | Bauru - FC - Sistemas de Informação | pt |
Arquivos
Pacote original
1 - 1 de 1
Carregando...
- Nome:
- roratto_ph_tcc_bauru.pdf
- Tamanho:
- 2.43 MB
- Formato:
- Adobe Portable Document Format
Licença do pacote
1 - 2 de 2
Carregando...
- Nome:
- license.txt
- Tamanho:
- 2.14 KB
- Formato:
- Item-specific license agreed upon to submission
- Descrição:
Carregando...
- Nome:
- roratto_ph_autorizacao_bauru.pdf
- Tamanho:
- 400.17 KB
- Formato:
- Adobe Portable Document Format
- Descrição:

