Alibaba Showcases AI Innovations in Entertainment and Security at IJCAI 2019

At the 50th anniversary IJCAI 2019 conference in Macau, Alibaba highlighted its latest AI technologies for entertainment and security, presenting multiple papers, demos, and exhibitions.

Alibaba's entertainment arm, the MoKu Lab, demonstrated the "Beidou Star" AI platform, which leverages AI and big data to analyze and assist decision‑making across the entire content lifecycle—from investment and production to operation and marketing. A demo showed how the platform provides recommendations and data analysis for each stage of video creation and promotion.

The lab also showcased an intelligent conversational video search system that integrates speech recognition (ASR), natural language understanding (NLU), multi‑turn dialogue management, and text‑to‑speech (TTS) to generate user‑friendly responses and retrieve video content based on complex queries such as “show me movies starring Huang Xiaoming’s wife” or “find Yi Yangqianxi’s floor‑dance video.”

In the security domain, Alibaba’s Turing Lab presented its AI‑driven safety solutions, including:

"Green Net" – a system that scans billions of text and image assets daily to detect harmful information.

"Intellectual Property Brain" – an IP protection engine with over 96% accuracy in detecting counterfeit and infringing goods, built on a massive dataset equivalent to 186 national libraries.

Facial‑recognition technology serving more than 100 scenarios across Alibaba’s ecosystem.

"Tianyan" – a visual‑technology‑based new‑retail security system.

The IP Brain was recently honored as an “AI Innovation Star” at the China AI Summit, and its deployment has helped police in 31 provinces capture 4,439 counterfeit‑related suspects and dismantle 4,289 illegal sites.

Alibaba’s participation also included a series of academic papers—five from the MoKu Lab and several on recommendation, multi‑view learning, and bandit algorithms—highlighting the company’s active research contributions to top AI conferences such as IJCAI, KDD, CVPR, ICML, and CIKM.

Speakers emphasized the tight coupling between industry and academia, noting that large‑scale data and real‑world user feedback are reshaping AI research priorities, especially in areas like model robustness, adversarial samples, and ethical considerations.

Alibaba announced ongoing collaborations with institutions such as Tsinghua University to launch a security‑AI challenger program focused on adversarial samples across text, image, video, and audio domains.

Both the MoKu Lab and Turing Lab are actively recruiting senior algorithm engineers in machine learning, natural language processing, and computer vision.