Toward Public Opinion Monitoring System of Large-Scale Data with Lambda Architecture

The public opinion monitoring system can monitor the latest information at any time and manage the network information activities by focusing on the website information about hot issues and key areas. The traditional Lambda architecture does not write abstract data on the storage layer, and the data latency and cost are relatively high on the storage. In this paper, we use the improved Lambda-based architecture to support stream and batch computing to make computational logic reused. The improved Lambda architecture enables the storage of unified historical data in full volume and online real-time incremental data. The storage can support high-throughput real-time write, as well as multi-field combination search and full-text retrieval.