Purpose

This study is to propose a more effective and efficient analytic methodology based on within-site clickstream associated with path visualization to explore the channel dependence of consumers' latent shopping intent and the related behaviors, with which in turn to gain insight concerning the interactivity between webpages.

Design/methodology/approach

The primary intention of the research is to design and develop a more effective and efficient approach for exploring the consumers' latent shopping intent and the related behaviors from the clickstream data. The proposed methodology is to use text-mining package, consisting of the combination of hierarchical recurrent neural networks and Hopfield-like neural network equipped with Laplacian-based graph visualization to visualize the consumers' browsing patterns. Based on the observed interactivity between webpages, consumers' latent shopping intent and the related behaviors can be understood.

Findings

The key finding is to evidence that consumers' latent shopping intent and related behaviors within website depend on channels the consumers click through. The accessing consumers through channels of paid search and display advertising are identified and categorized as goal-directed and exploratory modes, respectively. The results also indicate that the effect of the content of webpage on the consumer's purchase intent varies with channels. This implies that website optimization and attribution of online advertising should also be channel-dependent.

Practical implications

This is important for the managerial and theoretical implications: First, to uncover the channel dependence of consumer's latent shopping intent and browsing behaviors would be helpful to the attribution of the online advertising for the sales promotion. Second, in the past, webmasters did not understand users' preferences and make decisions of reorganization purely on the user's browsing path (sequential page view) without appraising psychological perspective, that is, user's latent shopping intent.

Originality/value

This study is the first to explore the channel dependences of consumer's latent shopping intent and the related browsing behaviors through within-site clickstream associated with path visualization. The findings are helpful to the attribution of the online advertising for the sales promotion and useful for webmasters to optimize the effectiveness and usability of their websites and in turn promote the purchase decision.

中文翻译：

---

一种通过可视化浏览模式探索消费者潜在购物意图和相关行为的渠道依赖性的新方法

目的

本研究旨在提出一种基于站内点击流与路径可视化相关联的更有效和高效的分析方法，以探索消费者潜在购物意图和相关行为的渠道依赖性，从而深入了解网页之间的交互性.

设计/方法/方法

该研究的主要目的是设计和开发一种更有效和高效的方法，用于从点击流数据中探索消费者的潜在购物意图和相关行为。所提出的方法是使用文本挖掘包，由分层循环神经网络和配备基于拉普拉斯的图形可视化的 Hopfield 类神经网络的组合组成，以可视化消费者的浏览模式。基于观察到的网页之间的交互性，可以了解消费者的潜在购物意图和相关行为。

发现

关键发现是证明消费者在网站内的潜在购物意图和相关行为取决于消费者点击的渠道。通过付费搜索和展示广告渠道访问消费者分别被识别和分类为目标导向和探索模式。结果还表明，网页内容对消费者购买意愿的影响因渠道而异。这意味着网站优化和在线广告的归因也应该依赖于渠道。

实际影响

这对于管理和理论意义很重要：首先，揭示消费者潜在购物意图和浏览行为的渠道依赖性将有助于在线广告促销的归因。其次，以往站长不了解用户的喜好，单纯根据用户的浏览路径（顺序页面浏览）进行重组决策，没有从心理角度，即用户的潜在购物意图进行评估。

原创性/价值

本研究首次通过与路径可视化相关的站点内点击流来探索消费者潜在购物意图和相关浏览行为的渠道依赖性。研究结果有助于在线广告促销的归因，有助于网站管理员优化其网站的有效性和可用性，进而促进购买决策。