Now the tourism information is overflow in the internet, but the useful information is had to find out. So, it takes the recommendation technology in common use as the research object, and adopts the comparison method to study different recommendation technologies, such as the association rules, collaborative filtering and item-based recommendation. The Apriori algorithm is discussed in the association rules first, therefore an improved Apriori method has been considered as an appropriate recommendation ways for tourism information service. Second, there are more and more evaluation record for those new or old tourist destinations or scenic spot by visitor can be found in internet, so it is easy to generate recommendation itemset. Then the collaborative filtering recommendation is selected as a promising recommend technology for tourism. Besides, the best feature of the item-based collaborative filtering recommendation is its expansibility so it is also a useful method for massive tourism information service. Lastly, the applicable tourism information recommendation method has been defined based on the user modeling.