基于多项logit模型的在线机票代理商选择行为
发布时间：2018-07-21 21:03
【摘要】：用户在线机票代理商选择行为的分析对旅游搜索网站机票搜索产品系统的改进非常重要。该文基于某旅游搜索网站机票搜索和预订用户的在线数据,对该网站的用户在线代理商选择行为进行分析。运用了多项logit模型理论及方法,将有关代理商的机票报价、评价、增值服务、页面位置以及市场份额的变量引入模型,并建立了可以解释该选择行为的对数价格多项logit模型。结果显示:用户的确最关注的是代理商的机票报价;搭售保险和评价人数过少都不利于用户对代理商的选择;增值服务的提供和较高的评分会增加用户对代理商的兴趣;用户更倾向于选择展示页面中靠前位置的代理商,其次是靠后的代理商;市场份额大的代理商对用户的选择行为会产生显著影响。预测检验表明该文建立的模型比历史数据预测方法有更高的预测精度。
[Abstract]:The analysis of the behavior of online ticket agent selection is very important to the improvement of ticket search product system of travel search website. Based on the online data of ticket search and reservation users of a travel search website, this paper analyzes the online agent selection behavior of the users. By using the theory and method of multiple logit models, this paper introduces the variables of ticket quotation, evaluation, value-added service, page location and market share into the model, and establishes the logarithmic price multivariate logit model which can explain the choice behavior. The results show that the users really pay most attention to the ticket quotation of the agent, the low number of insurance and evaluation is not conducive to the choice of the agent, the provision of value-added service and the higher score will increase the interest of the user in the agent; Users are more likely to choose the agent in the front position of the display page, followed by the next agent; the agent with large market share will have a significant impact on the user's choice behavior. The prediction test shows that the model established in this paper has higher prediction accuracy than the historical data prediction method.
【作者单位】： 清华大学工业工程系;内蒙古大学交通学院;
【基金】：国家自然科学基金资助项目(71622006)
【分类号】：F724.6;F562.6
[Abstract]:The analysis of the behavior of online ticket agent selection is very important to the improvement of ticket search product system of travel search website. Based on the online data of ticket search and reservation users of a travel search website, this paper analyzes the online agent selection behavior of the users. By using the theory and method of multiple logit models, this paper introduces the variables of ticket quotation, evaluation, value-added service, page location and market share into the model, and establishes the logarithmic price multivariate logit model which can explain the choice behavior. The results show that the users really pay most attention to the ticket quotation of the agent, the low number of insurance and evaluation is not conducive to the choice of the agent, the provision of value-added service and the higher score will increase the interest of the user in the agent; Users are more likely to choose the agent in the front position of the display page, followed by the next agent; the agent with large market share will have a significant impact on the user's choice behavior. The prediction test shows that the model established in this paper has higher prediction accuracy than the historical data prediction method.
【作者单位】： 清华大学工业工程系;内蒙古大学交通学院;
【基金】：国家自然科学基金资助项目(71622006)
【分类号】：F724.6;F562.6
【相似文献】
相关期刊论文 前10条
1 王爽;赵鹏;;基于Logit模型的客运专线旅客选择行为分析[J];铁道学报;2009年03期
2 刘文歌;赵胜川;;基于Logit模型的道路交通统计生命价值[J];交通运输系统工程与信息;2013年01期
3 刘振;周溪召;;巢式Logit模型在交通方式选择行为中的应用[J];上海海事大学学报;2006年03期
4 南娟;王U，
本文编号：2136850
本文链接：https://www.wllwen.com/jingjilunwen/jtysjj/2136850.html
