On June 23, 2026, SFL invited Professor R. Harald Baayen, a leading professor of quantitative linguistics in the Department of Linguistics at the University of Tübingen, Germany, and a member of Academia Europaea, to give an online academic lecture on “Applications of Word Embeddings in Linguistics.” The lecture was hosted by Dean Cao Ruonan.
Dean Cao first introduced and welcomed Professor Baayen. In the lecture, Professor Baayen explained the core concepts of word embeddings, noting that the technology can convert semantics into high-dimensional numerical vectors, quantitatively analyze linguistic patterns, and address the subjectivity and limited empirical basis of traditional qualitative linguistic research. He then presented three empirical studies to show the value of word embeddings in linguistic research: English noun plurals involve implicit semantic categorization rather than a single grammatical rule; word embeddings can clearly distinguish the semantic differences between Chinese AABB and ABAB reduplication structures; and a vector prediction model for Chinese compounds was built to compare mainstream models and analyze the semantic composition of compounds with different numbers of characters. He also shared his latest research, showing that contextual word embeddings can predict subtle changes in spoken tones, prompting a reexamination of the intrinsic relationship between linguistic form and meaning.
During the Q&A session, faculty and students actively raised questions on cross-lingual word embedding alignment, limitations of compound semantic prediction, and differences between Chinese and English semantic spaces. Professor Baayen answered each question and advised young linguistics researchers to build on basic statistical tools and use open-source word embedding models for quantitative research.
The lecture provided a high-level academic exchange platform, helped faculty and students grasp frontier developments in computational linguistics, clarified the application prospects of word embeddings in language research, and offered ideas and references for quantitative and interdisciplinary exploration in linguistics.
Speaker: Professor R. Harald Baayen
Host: Professor Cao Ruonan
Venue
School Homepage
中文
