Abstract: Ecotourist study is one of the most important parts in ectourism research, which must be based on enough ecotourist samples. Using the foreign experience for reference, the paper proposed an integrated classification technique comprising K-Nearest Neighbor (KNN) and BackPropagation (BP) Networks to identify the ectourists. A questionnaire was designed to collect data through "internet" and "face to face" interviews during the National Day holiday in 2003 in Beijing Baihuashan Nature Reserve. 423 persons answered the questionnaires and 139 out of them were identified as ecotourists. The result of classification is proved to be effective by prediction validity test, content validity test, theory validity test and convergence validity test.In addition, the ecotourists were classified as hard ecotourists, frequent ecotourists and occasional ecotourists farther according to the times they traveled to nature-based destinations every year and their New Ecological Paradigm (NEP) grade which is used to evaluate the visitors' attitude to environment. Frequent ecotourists and occasional ecotourists both fall category of soft ecotourists, so the "hard-soft" classification system is founded, which will help propel ecotourist studies in China to develop towards the current international mainstream of the field. Finally, the paper clarified the concept of hard ecotourists, frequent ecotourists and experienced ecotourists ever used in others's research. Hard ecotourists and frequent ecotourists both belong to experienced ecotourists, but the hard ecotourists evidently have better environmental attitudes than the frequent ecotourists.

Key words: ecotourist, identification, segmentation, Baihuashan

摘要： 生态旅游者是生态旅游研究的重要方面，而足够数量的生态旅游者样本则是该研究得以进行的前提。以北京市百花山自然保护区为例，在总结国外相关研究经验的基础上，采用K阶最近邻（KNN）和反向传播(BP)前馈型多层神经网络所构成的综合分类器对百花山游客中的生态旅游者进行识别，又根据“每年去相对原始的自然区域的旅游次数”和环境态度尺度NEP得分两项指标将生态旅游者进一步细分为严格的生态旅游者、经常的生态旅游者和偶尔的生态旅游者，其中经常的生态旅游者和偶尔的生态旅游者归为一般的生态旅游者，从而构建起严格的－一般的生态旅游者分类体系。分类结果经预测有效性、内容有效性、理论有效性和收敛有效性检验，效果较好。

关键词: 生态旅游者, 识别, 细分, 百花山