Xiaochuan Wang

4Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Xiaochuan Wang focuses on information retrieval, web search user behavior, and search result diversification. Their work spans the analysis of user examination patterns across mobile and image search interfaces, as well as the development of large-scale web corpora. The most recent research in their trajectory investigates search intent distributions and mobile search dynamics.

Research Trajectory

Xiaochuan Wang’s research in the available corpus spans 2017 to 2018, focusing primarily on information retrieval (IR) systems and the empirical analysis of user search behavior. Early work in this period examined how users interact with specialized search interfaces and contributed to foundational datasets for the IR community.

In 2017, Wang contributed to research on user interaction with non-traditional search engine result pages (SERPs). As a co-author of Investigating Examination Behavior of Image Search Users, Wang participated in analyzing how image search engines alter traditional user behavior. The study contrasted general web search with image search interfaces, noting how two-dimensional result placements, image snapshots, and implicit pagination change the way users examine results. During the same year, Wang was part of the team that introduced SogouT-16: A New Web Corpus to Embrace IR Research. This work addressed the high cost of large-scale web data collection by providing a massive, free-of-charge public Chinese web corpus, which was subsequently adopted for the ad-hoc retrieval task in NTCIR-13.

In 2018, Wang took a leading role in analyzing mobile search dynamics. As the first author of A Large-Scale Study of Mobile Search Examination Behavior, Wang directed a comprehensive investigation into how users examine mobile SERPs, particularly in the absence of click signals. Utilizing large-scale search logs from a commercial mobile search engine, the study leveraged viewport metrics to estimate user attention, providing insights into position bias, the relationship between examination and click behavior, and how user attention shifts as a search session continues.

Also in 2018, Wang contributed to algorithmic IR research concerning search result diversification. In Treating Each Intent Equally: The Equilibrium of IA-Select, the research team examined the objective functions used to satisfy multiple user subtopic intents. The paper critiqued the standard cascade assumption—which posits that a user's information need drops once a subtopic is satisfied—demonstrating empirically that this assumption can cause diversified results to deviate from the actual, often unbalanced, distribution of user intentions.

Current Directions

The most recent papers in the evidence date to 2018. Based on these publications, Wang's latest documented research addresses the intersection of user behavior analytics and algorithmic search optimization. A primary focus is the precise measurement of user attention on mobile devices, moving beyond traditional click-through rates to utilize viewport metrics for a more granular understanding of search examination. Concurrently, the recent work investigates the theoretical underpinnings of search result diversification, specifically addressing how algorithms can better align with real-world, unbalanced distributions of user search intents.

Profile updated

See something wrong? Report a mistake