Abstract: Quality safety early warning with quality information published by consumers on internet is a new perspective of quality governance in network times. The paper studies the construction and implementation of an analysis framework and model based on consumer quality safety online early-warning. The analysis framework of quality safety online early warning is designed from three aspects of quality information classification, risk level and warning division, in order to extract general classification forms of consumer online quality information, namely bodily injury information, insecurity information and character change information. Combining with the characteristics of online information, the three categories of information is divided into five different levels according to the seriousness of risk. The division of early warning is determined based on risk level. Specify the construction process of early warning model with evidence theory from three steps of identifying framework construction, obtaining reliability evaluation and recursive fusion algorithm, mixing the quality information in different types and risk levels. The study provides an idea and method for quality online government in big data times.
Key words: Quality safety; network warning; analysis framework; model