| 配分額 *注記 |
6,370千円 (直接経費: 4,900千円、間接経費: 1,470千円)
2022年度: 2,210千円 (直接経費: 1,700千円、間接経費: 510千円)
2021年度: 2,470千円 (直接経費: 1,900千円、間接経費: 570千円)
2020年度: 1,690千円 (直接経費: 1,300千円、間接経費: 390千円)
|
|---|
| 研究実績の概要 |
研究成果を国際学会と国内研究会において発表を行った。
国際学会の発表テーマは、Analysis of Questions to Support Dementia Patient Family Focusing on Modality and Consideration on Chatbot Support System,The purpose of this study is to develop an AI chatbot system that gives answers to various questions about dementia patients for their families and relatives with dementia patients. The basis of the chatbot system replies to one advice against one question. However, it is not easy to predict what kind of questions will be asked. So, this study has proposed an algorithm of an AI chatbot system that can reply to appropriate advice against unknown asked in the following way.
1) Collect the data for questions and advices for dementia patients. 2) Give the words that construct question to data the 40 labels data such as question, confirmation, behavior, attachment, feeling, agreement, ability, denial, forgetting, method, distress and confirmation. 3) Classify the question data by transforming to 40 degrees non negativevector via the above process. 4) If question data d(i) was given the clustering label k, then the advice data a(i) is also given the same label data k, where i and k are non-negative integsers.
5) Check contradict between questions and advices, by generating random question-advice pairs from questions and answers belonging to the same class.
|
