论文部分内容阅读
提出一种基于家庭本体知识库的服务决策机制的设计方法.首先,围绕网络中存储的与家庭服务相关的半结构化文本信息初步构建训练数据集.其次,基于与家庭服务相关的本体知识库构建信息提取机制,针对初步数据集包含的物品信息提取知识库中对应的服务物品属性信息;基于长短时记忆网络(LSTM)设计一种学习模型并命名为服务决策机制,该机制利用知识库作为常识部分进行辅助推理,生成与服务请求对应的决策信息.最后,针对决策信息进行对比实验.实验结果表明:将本体知识库作为常识部分辅助推理,能够提高决策信息的准确性.
This paper proposes a design method of service decision mechanism based on the family ontology knowledge base.Firstly, the training data set is built around the semi-structured text information stored in the network and related to family services.Secondly, based on the ontology knowledge base Construct the information extraction mechanism, extract the attribute information of the corresponding service item in the knowledge base for the item information contained in the initial dataset; design a learning model based on the LSTM and name it as the service decision mechanism, which uses the knowledge base as the Common sense part to generate decision-making information corresponding to service request.Finally, a comparative experiment on decision-making information is carried out.The experimental results show that the ontology knowledge base can be used as common sense part to assist in reasoning, which can improve the accuracy of decision information.