Schools are among the primary avenues for public healthcare interventions. With resource limitations posing challenges to the routine conduct of health and wellness checks in Philippine public schools, the deployment of a chatbot-assisted health monitoring system may provide an alternative method. However, deriving insights from raw conversations is not straightforward due to the expressiveness of natural language that causes variances in the input. In this paper, we present a process for transforming unstructured dialogues into a structured schema. The process comprises four stages: (i) processing the dialogues through entity extraction and data aggregation, (ii) storing them as NoSQL documents on the cloud, (iii) transforming them into a star schema for online analytical processing and building an extract-transform-load workflow, and (iv) creating a web-based dashboard for visualizing summarized data and reports. Performance evaluation of this dashboard showed that increasing the number of stored dialogues by a factor of 100,000 increased the loading time for the display of roll-up, drill-down, and filter results by around only one second.

翻译：学校是公共卫生干预的主要途径之一。由于资源限制，菲律宾公立学校难以常规开展健康与 wellness 检查，部署聊天机器人辅助的健康监测系统或可提供替代方案。然而，由于自然语言的表现力导致输入存在差异，从原始对话中提取洞察并非易事。本文提出一种将非结构化对话转换为结构化模式的过程。该过程包含四个阶段：（i）通过实体提取和数据聚合处理对话，（ii）将其作为 NoSQL 文档存储在云端，（iii）将其转换为用于在线分析处理的星型模式并构建提取-转换-加载工作流，（iv）创建基于网页的仪表板以可视化汇总数据和报告。对该仪表板的性能评估显示，存储的对话数量增加100,000倍时，上卷、下钻和过滤结果的显示加载时间仅增加约一秒。