Structured representation of temporal document collections by diachronic linguistic periodization

Language is our main communication tool. Deep understanding of its evolution is imperative for many related research areas including history, humanities, social sciences, etc. as well as for effective temporal information retrieval. To this end, we are interested in the task of segmenting long-term document corpora into naturally coherent periods based on the embodied evolving word semantics. There are many benefits of such segmentation including better representation of content in longterm document collections and support for modeling and understanding semantic drift. We propose a two-step framework for learning time-aware word semantics and periodizing document archive. The effectiveness of our model is demonstrated on the New York Times corpus spanning from 1990 to 2016.