Dr. Steven Skiena on “Representing Knowledge through Word and Graph Embeddings”

Posted on February 16, 2020 by Clara Tran

On February 11, 2020, Dr. Steven Skiena presented the first Spring STEM Lecture at the University Libraries about the useful representation of word and graph embeddings in data science.

Dr. Skiena presented his work in an easy to understand way. Dr. Skiena and his team used natural language processing (NLP) to show how one language maps to another language. They also used word embeddings trained over texts taken from different time periods to detect how words’ meaning changes over time. Additionally, they used phone data to show language clusters in Belgium. These are just a few of the deep learning examples applied in various situations.

The intriguing lecture attracted faculty and students not only from computer science, but also from other disciplines. During the lecture, Dr. Skiena interacted with attendees through engaging questions and answers. Attendees mentioned that the lecture was very easy to understand, interesting, and informative.