Fork me on GitHub

ICE: Item Concept Embedding @ SIGIR2017

Chuan-Ju Wang, Ting-Hsiang Wang, Hsiu-Wei Yang, Bo-Sin Chang, Ming-Feng Tsai


ICE is an item concept embedding framework to model item concepts via textual information. With the proposed carefully designed ICE networks, the resulting embedding facilitates both homogeneous and heterogeneous retrieval, including item-to-item and word-to-item retrieval. Below is a visualization on the learned movie and word embeddings of the IMDB dataset via the plots of the movies.

Reset Chart