Hongming Gao, Di Deng, Hongwei Liu, Zhouyang Liang
{"title":"利用用户数字足迹分析在线旅行社移动点击决策的顺序框架","authors":"Hongming Gao, Di Deng, Hongwei Liu, Zhouyang Liang","doi":"10.1007/s40558-024-00294-z","DOIUrl":null,"url":null,"abstract":"<p>In the hotel booking market, high click-through rates are essential for online travel agencies (OTAs) to earn commissions. Given the dominance of mobile devices in web traffic, analyzing the mobile click-through decision-making process plays a vital role in search engine optimization. This study proposes a sequential framework that leverages Bayesian inference to model individual users’ click-through behaviors using user digital footprints, which encompass sequences of search, browse, compare, and click-through actions. This framework extracts three categories of information based on the degrees of dynamism in the hotel search process, ranging from less dynamic to highly dynamic levels: static hotel attributes, information cues in the search results, and temporal characteristics of user behaviors. Extensive experiments on a global OTA mobile clickstream dataset with over 600,000 observations reveal the substantial superiority of the proposed framework over the baseline models like probit regression and Naive Bayes. Notably, temporal characteristics emerge as the most important category. Drawing on our model, we delve into the interpretability of these three information categories. Additionally, we compare their varying impacts across different devices. Beyond these findings, this study offers valuable managerial implications for mobile OTA search engine marketing and optimization.</p>","PeriodicalId":46275,"journal":{"name":"Information Technology & Tourism","volume":"26 1","pages":""},"PeriodicalIF":6.3000,"publicationDate":"2024-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Sequential framework for analyzing mobile click-through decision in online travel agency with user digital footprints\",\"authors\":\"Hongming Gao, Di Deng, Hongwei Liu, Zhouyang Liang\",\"doi\":\"10.1007/s40558-024-00294-z\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>In the hotel booking market, high click-through rates are essential for online travel agencies (OTAs) to earn commissions. Given the dominance of mobile devices in web traffic, analyzing the mobile click-through decision-making process plays a vital role in search engine optimization. This study proposes a sequential framework that leverages Bayesian inference to model individual users’ click-through behaviors using user digital footprints, which encompass sequences of search, browse, compare, and click-through actions. This framework extracts three categories of information based on the degrees of dynamism in the hotel search process, ranging from less dynamic to highly dynamic levels: static hotel attributes, information cues in the search results, and temporal characteristics of user behaviors. Extensive experiments on a global OTA mobile clickstream dataset with over 600,000 observations reveal the substantial superiority of the proposed framework over the baseline models like probit regression and Naive Bayes. Notably, temporal characteristics emerge as the most important category. Drawing on our model, we delve into the interpretability of these three information categories. Additionally, we compare their varying impacts across different devices. Beyond these findings, this study offers valuable managerial implications for mobile OTA search engine marketing and optimization.</p>\",\"PeriodicalId\":46275,\"journal\":{\"name\":\"Information Technology & Tourism\",\"volume\":\"26 1\",\"pages\":\"\"},\"PeriodicalIF\":6.3000,\"publicationDate\":\"2024-06-26\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Information Technology & Tourism\",\"FirstCategoryId\":\"91\",\"ListUrlMain\":\"https://doi.org/10.1007/s40558-024-00294-z\",\"RegionNum\":3,\"RegionCategory\":\"管理学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"HOSPITALITY, LEISURE, SPORT & TOURISM\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Information Technology & Tourism","FirstCategoryId":"91","ListUrlMain":"https://doi.org/10.1007/s40558-024-00294-z","RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"HOSPITALITY, LEISURE, SPORT & TOURISM","Score":null,"Total":0}
引用次数: 0
摘要
在酒店预订市场,高点击率是在线旅行社(OTA)赚取佣金的关键。鉴于移动设备在网络流量中的主导地位,分析移动点击决策过程在搜索引擎优化中起着至关重要的作用。本研究提出了一个顺序框架,利用贝叶斯推理方法,使用用户数字足迹(包括搜索、浏览、比较和点击操作的顺序)对单个用户的点击行为进行建模。该框架根据酒店搜索过程的动态程度(从低动态到高动态)提取三类信息:静态酒店属性、搜索结果中的信息提示和用户行为的时间特征。在全球 OTA 移动点击流数据集上进行的大量实验显示,与 probit 回归和 Naive Bayes 等基线模型相比,所提出的框架具有很大的优越性。值得注意的是,时间特征成为最重要的类别。根据我们的模型,我们深入研究了这三类信息的可解释性。此外,我们还比较了它们在不同设备上的不同影响。除了这些发现,本研究还为移动 OTA 搜索引擎营销和优化提供了宝贵的管理启示。
Sequential framework for analyzing mobile click-through decision in online travel agency with user digital footprints
In the hotel booking market, high click-through rates are essential for online travel agencies (OTAs) to earn commissions. Given the dominance of mobile devices in web traffic, analyzing the mobile click-through decision-making process plays a vital role in search engine optimization. This study proposes a sequential framework that leverages Bayesian inference to model individual users’ click-through behaviors using user digital footprints, which encompass sequences of search, browse, compare, and click-through actions. This framework extracts three categories of information based on the degrees of dynamism in the hotel search process, ranging from less dynamic to highly dynamic levels: static hotel attributes, information cues in the search results, and temporal characteristics of user behaviors. Extensive experiments on a global OTA mobile clickstream dataset with over 600,000 observations reveal the substantial superiority of the proposed framework over the baseline models like probit regression and Naive Bayes. Notably, temporal characteristics emerge as the most important category. Drawing on our model, we delve into the interpretability of these three information categories. Additionally, we compare their varying impacts across different devices. Beyond these findings, this study offers valuable managerial implications for mobile OTA search engine marketing and optimization.
期刊介绍:
Information Technology & Tourism stands as the pioneer interdisciplinary journal dedicated to exploring the essence and impact of digital technology in tourism, travel, and hospitality. It delves into challenges emerging at the crossroads of IT and the domains of tourism, travel, and hospitality, embracing perspectives from both technical and social sciences. The journal covers a broad spectrum of topics, including but not limited to the development, adoption, use, management, and governance of digital technology. It supports both theory-focused research and studies with direct relevance to the industry.