A scalable model-based approach to gain insights in dynamic purchase behavior for large product assortments and customer bases.

Social media is used as a platform to not only make connections and share user-generated content but also to advertise products. We study the role of opinion leadership on the social media networking website of Facebook for product advertisements through news feeds and wall postings. We found that perceived benevolence of the opinion leader posting advertisements was significantly associated wi...

Understanding purchase behavior of people in city environment can have important implications in the design and management of cities and the study of urban economy. Traditional studies have proposed to use gravity-based spatial interaction (Huff) model [1, 2] or discrete choice model [3] to model individual purchase behavior. These models treat individual purchases separately and do not explici...

We propose a new topic model for tracking timevarying consumer purchase behavior, in which consumer interests and item trends change over time. The proposed model can adaptively track changes in interests and trends based on current purchase logs and previously estimated interests and trends. The online nature of the proposed method means we do not need to store past data for current inferences...

DOES SEARCH MATTER? : USING ONLINE CLICK STREAM DATA TO EXAMINE THE RELATIONSHIP BETWEEN ONLINE SEARCH AND PURCHASE BEHAVIOR Neveen F. Awad, Joni L. Jones, Jian Zhang Introduction The ability to track consumer behavior online is considered a prime advantage that online retailing has over brick and mortar retailers. However, firms are still grappling with how to understand their consumers throug...

We propose a new topic model for tracking timevarying consumer purchase behavior, in which consumer interests and item trends change over time. The proposed model can adaptively track changes in interests and trends based on current purchase logs and previously estimated interests and trends. The online nature of the proposed method means we do not need to store past data for current inferences...

Modern e-commerce recommendation systems recommend users products through purchase prediction off of historical purchase data. This signal however has limitations as new and long tail products have little to no such signal to exploit. One signal however that influences user purchase behavior, especially in verticals such as fashion, is visual. In this paper we explore how visual similarity and ...