Hujiang Technology Salon: Data Applications – Summaries of Five Expert Talks

On July 23, 2016, Hujiang organized a technology salon centered on data applications, bringing together industry experts to discuss how data can drive operations and product innovation.

The first talk, delivered by Hujiang’s Chief Data Officer Tang Xiaolei, examined data‑driven operations in online education, outlining user conversion models, metric decomposition, goal monitoring, and case studies such as testing models, memory curves, and learning alerts, while highlighting challenges of implementing data‑based interventions in learning.

Hujiang’s data lead Hu Liangjing presented “O2O Data‑Driven Operations,” describing anti‑fraud measures, address‑base construction, order dispatch, churn prediction, and lifecycle management, and showing how an enriched address database and evolving dispatch algorithms improved coverage, accuracy, and delivery efficiency.

Suning Cloud’s Big Data Center Technical Director Chen Minmin shared “Personalized Recommendation and Pitfalls,” focusing on building a recommendation system for e‑commerce, including database read/write optimization, monitoring of over‑supplemented recommendation items, and multi‑metric evaluation of algorithm performance.

Finally, Alibaba Alipay senior data mining engineer Zhang Bin (aka “Bread‑Jun”) discussed “Deep Learning in Image Search,” covering the end‑to‑end pipeline of image recognition, object detection, feature extraction, ranking, quality control, as well as category prediction and index construction to support mobile‑first, “photo‑to‑buy” scenarios.

The event included numerous slides and on‑site photos, which were later made available for download.