Can anyone help correct possible error in the abstract? The tourists’ attitudes and the potential problems coming along with tourism projects were mainly analyzed. Thousands of comments aiming at five different ancient China towns coming from website were collected. Then with text mining, terms frequency was counted, and the hot topics were analyzed. After that, the scores were received with sentiment analysis. Based on these scores, the possible factors contributing to variation of scores were investigated by analysis of variance and ultimately the fuzzy comprehensive evaluation model can be built. The final results are able to give references for creating a better system for ancient towns traveling.