Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS
Chatbots in educational settings have grown significantly, facilitating interaction between students and learning platforms. However, current systems, such as Rasa, Moodle Integrated Chatbots, and ChatterBot, present significant limitations in precision, adaptability, and response time, affecting th...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10804145/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850101029230608384 |
|---|---|
| author | William Villegas-Ch Jaime Govea Rommel Gutierrez |
| author_facet | William Villegas-Ch Jaime Govea Rommel Gutierrez |
| author_sort | William Villegas-Ch |
| collection | DOAJ |
| description | Chatbots in educational settings have grown significantly, facilitating interaction between students and learning platforms. However, current systems, such as Rasa, Moodle Integrated Chatbots, and ChatterBot, present significant limitations in precision, adaptability, and response time, affecting their effectiveness in resolving academic queries and personalizing learning. To address these shortcomings, this work proposes the development of an advanced educational chatbot that combines large language models (LLMs) with knowledge graphs, allowing for more accurate and contextualized responses and offering valuable suggestions to enrich the learning process. The system is evaluated based on its ability to adjust to different student profiles and offer fast and accurate responses. The results show that the proposed chatbot achieves a precision of 85%, outperforming Rasa and ChatterBot, which achieved accuracies of 83% and 81%, respectively. Furthermore, the chatbot reduces response times to 0.41 seconds, improving efficiency compared to other solutions. The system also demonstrates adaptability, effectively adjusting to students’ learning styles and academic levels. This work indicates that knowledge graph integration and hyperparameter optimization are crucial to improving educational chatbots’ precision, speed, and adaptability, presenting an innovative solution that overcomes the limitations of current systems. |
| format | Article |
| id | doaj-art-436cbc8c2fbe4e8a9418d6a96c3c52f0 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-436cbc8c2fbe4e8a9418d6a96c3c52f02025-08-20T02:40:08ZengIEEEIEEE Access2169-35362024-01-011219199419201210.1109/ACCESS.2024.351895210804145Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMSWilliam Villegas-Ch0https://orcid.org/0000-0002-5421-7710Jaime Govea1Rommel Gutierrez2https://orcid.org/0009-0004-3230-4129Escuela de Ingeniería en Ciberseguridad, FICA, Universidad de Las Américas, Quito, EcuadorEscuela de Ingeniería en Ciberseguridad, FICA, Universidad de Las Américas, Quito, EcuadorEscuela de Ingeniería en Ciberseguridad, FICA, Universidad de Las Américas, Quito, EcuadorChatbots in educational settings have grown significantly, facilitating interaction between students and learning platforms. However, current systems, such as Rasa, Moodle Integrated Chatbots, and ChatterBot, present significant limitations in precision, adaptability, and response time, affecting their effectiveness in resolving academic queries and personalizing learning. To address these shortcomings, this work proposes the development of an advanced educational chatbot that combines large language models (LLMs) with knowledge graphs, allowing for more accurate and contextualized responses and offering valuable suggestions to enrich the learning process. The system is evaluated based on its ability to adjust to different student profiles and offer fast and accurate responses. The results show that the proposed chatbot achieves a precision of 85%, outperforming Rasa and ChatterBot, which achieved accuracies of 83% and 81%, respectively. Furthermore, the chatbot reduces response times to 0.41 seconds, improving efficiency compared to other solutions. The system also demonstrates adaptability, effectively adjusting to students’ learning styles and academic levels. This work indicates that knowledge graph integration and hyperparameter optimization are crucial to improving educational chatbots’ precision, speed, and adaptability, presenting an innovative solution that overcomes the limitations of current systems.https://ieeexplore.ieee.org/document/10804145/Educational chatbotslarge language modelsknowledge graphslearning personalization |
| spellingShingle | William Villegas-Ch Jaime Govea Rommel Gutierrez Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS IEEE Access Educational chatbots large language models knowledge graphs learning personalization |
| title | Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS |
| title_full | Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS |
| title_fullStr | Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS |
| title_full_unstemmed | Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS |
| title_short | Optimizing Language Model-Based Educational Assistants Using Knowledge Graphs: Integration With Moodle LMS |
| title_sort | optimizing language model based educational assistants using knowledge graphs integration with moodle lms |
| topic | Educational chatbots large language models knowledge graphs learning personalization |
| url | https://ieeexplore.ieee.org/document/10804145/ |
| work_keys_str_mv | AT williamvillegasch optimizinglanguagemodelbasededucationalassistantsusingknowledgegraphsintegrationwithmoodlelms AT jaimegovea optimizinglanguagemodelbasededucationalassistantsusingknowledgegraphsintegrationwithmoodlelms AT rommelgutierrez optimizinglanguagemodelbasededucationalassistantsusingknowledgegraphsintegrationwithmoodlelms |