Requirement-oriented integration of Deep Web information refers to acquiring meaningful information from tremendous Web databases according to users’ needs.On the basis of the research survey in the past 10 years

a new framework for the requirement-oriented integration of Deep Web information is proposed

which is composed of four components: search engine for Deep Web

interface integration

source description and query execution.According to this framework

the research work in the last decades classified and discussed in-depth.The key evaluation factors of the requirement-oriented integration for Deep Web information are concluded

namely

personalization

user’s satisfaction and query efficiency.At last

the suggestions for future research directions are pointed out.