Title: Rule based Complex Event Processing for an Air Quality Monitoring System in Smart City

Abstract: In recent years, smart city-based development has gained momentum due to its versatile nature in architecture and planning for the systematic habitation of human beings. According to World Health Organization (WHO) report, air pollution causes serious respiratory diseases. Hence, it becomes necessary to real-time monitoring of air quality to minimize effect by taking time-bound decisions by the stakeholders. The air pollution comprises various compositions such as NH3, O3, SO2, NO2, etc., and their concentrations vary from location to location.The research work proposes an integrated framework for monitoring air quality using rule-based Complex Event Processing (CEP) and SPARQL queries. CEP works with the data stream based on predefined rules to detect the complex pattern, which helps in decision support for stakeholders. Initially, the dataset was collected from the Central Pollution Control Board (CPCB) of India and this data was then preprocessed and passed through Apache Kafka. Then a knowledge graph developed based on the air quality paradigm. Consequently, convert preprocessed data into Resource Description Framework (RDF) data, and integrate with Knowledge graph which is ingested to CEP engine using Apache Jena for enhancing the decision support . Simultaneously, rules are extracted using a decision tree, and some ground truth parameters of CPCB are added and ingested to the CEP engine to determine the complex patterns. Consequently, the SPARQL query is used on real-time RDF dataset for fetching the condition of air quality as good, poor, severe, hazardous etc based on complex events detection. For validating the proposed approach various chunks of RDF are used for the deployment of events to the CEP engine, and its performance is examined over time while performing simple and complex queries.