Extracting valuable feedback information from user behavior logs is one of the major concerns in Web search studies. Among the tremendous efforts that aim to improve search performance with user behavior modeling, constructing click models is of vital importance because it provides a direct estimation of result relevance. Most existing click models assume that whether user clicks a result only depends on the examination probability and the content of the result. However, through carefully designed user eye-tracking study we found that users do not make click-through decisions isolatedly. Instead, they also take the context of a result (e.g. adjacent results) into consideration. This finding leads to the designing of a novel click model named Comparison-based Click Model (CBCM). Different from traditional examination hypothesis, CBCM introduces the concept of examination viewport and assumes users click results after comparing adjacent results within a same viewport. Experimental results on a public available user behavior dataset show effectiveness of CBCM.