Augmenting Semantic Representation of Depressive Language: from Forums to Microblogs

Here we share our Adjusted Twitter word Embedding (ATE) and Depression Specific word Embedding (DSE) as described in our paper "Augmenting Semantic Representation of Depressive Language: from Forums to Microblogs", in the following formats respectively:

• ATE-Embedding.p is a pickle file which contains a Python dictionary of words and their corresponding vectors.
• DSE-Embedding.bin is a Gensim Word2Vec file and should be loaded using Gensim library (the files with .npy extensions are required to do this.)
• Example code for loading DSE-Embedding using Gensim: Word2Vec.load(embedding_path), tested on Gensim-3.8

Please do cite our paper if you use our embedding or the idea of creating them.