Tsinghua University and TAL Launch Phase II Collaboration on Intelligent Education Research

On the morning of April 15, the Tsinghua University Computer Science Department and the TAL Education Intelligent Education Information Technology Joint Research Center held the launch ceremony for Phase II cooperation, building on Phase I achievements to deepen frontier exploration of intelligent education.

Attending the signing and launch were Liu Yiqun, secretary of the Party Committee of the Computer Science Department, former department head Zhou Lizhu, deputy director Feng Jianhua, Center director Li Guoliang, TAL CTO Tian Mi, TAL Vice President and AI Institute head Wu Zhongqin, Vice President Li Fei, and Center deputy director Liu Xiao, among others.

Phase II will focus on real teaching scenarios from TAL’s flagship products such as Xueersi Premium, Xueersi Online School, and Little Monkey Enlightenment, addressing authentic learning needs across various teaching modes, grades, and subjects. The collaboration aims to explore online teaching environment construction, interaction mechanism design, and intelligent student assessment, while seeking breakthroughs in knowledge‑graph construction and reasoning, intelligent question generation, and automatic problem solving.

Liu Yiqun praised the achievements of Phase I and expressed confidence in Phase II, stating that TAL has always been at the forefront of intelligent education and that the partnership will leverage real‑world scenarios to advance research and apply scientific results.

CTO Tian Mi emphasized the intention to combine Tsinghua’s cutting‑edge research in big data and artificial intelligence with TAL’s educational application strengths, creating new models, technologies, and methods for the integration of technology and education, and opening the next‑generation AI platform for the entire industry.

Since 2018, the two parties have jointly researched educational big data mining, intelligent education theory, automatic problem solving, oral expression evaluation, and personalized speech recognition, publishing dozens of academic papers and integrating many research outcomes into TAL’s products.

One notable project, “Teacher Personalized Emotional Speech Synthesis,” developed a system that can quickly build speaker models from limited data, enabling personalized voice calls and feedback for teachers, which has been deployed in Xueersi’s large‑class live teaching to enhance student engagement.

The “Teacher Emotion Analysis” project quantifies teachers’ emotional states from lecture audio, with results published at ACM Multimedia 2019. The system provides intelligent detection and feedback for lesson preparation and real‑time analysis of online classroom behavior, helping teachers improve teaching quality.

In the area of cultivating student learning habits, the joint research center’s technology has been applied to TAL’s bilingual oral expression assessment solution, delivering real‑time, objective, multi‑dimensional evaluation of students’ spoken responses.

Concluding, the partners reaffirm their mission “Love and technology make education better,” and pledge to explore the uncharted “no‑man’s land” of intelligent education with industry, academia, and research forces, driving steady progress in the sector.