Sketching Word Vectors Through Hashing

We propose a new fast word embedding technique using hash functions. The method is a derandomization of a new type of random projections: By disregarding the classic constraint used in designing random projections (i.e., preserving pairwise distances in a particular normed space), our solution exploits extremely sparse non-negative random projections. Our experiments show that the proposed method can achieve competitive results, comparable to neural embedding learning techniques, however, with only a fraction of the computational complexity of these methods. While the proposed derandomization enhances the computational and space complexity of our method, the possibility of applying weighting methods such as positive pointwise mutual information (PPMI) to our models after their construction (and at a reduced dimensionality) imparts a high discriminatory power to the resulting embeddings. Obviously, this method comes with other known benefits of random projection-based techniques such as ease of update.

Challenges in Persian Electronic Text Analysis

Farsi, also known as Persian, is the official language of Iran and Tajikistan and one of the two main languages spoken in Afghanistan. Farsi enjoys a unified Arabic script as its writing system. In this paper we briefly introduce the writing standards of Farsi and highlight problems one would face when analyzing Farsi electronic texts, especially during development of Farsi corpora regarding to transcription and encoding of Farsi e-texts. The pointes mentioned may sounds easy but they are crucial when developing and processing written corpora of Farsi.