Systems and methods for generating data dictionaries are disclosed. In one example, a system comprises a candidate list generator module to generate a list of keywords from search query information, a validation module to validate candidate token pairs based on a divergence value wherein candidate token pairs having divergence values exceeding a divergence threshold are valid token pairs, and a data dictionary module to receive the validated token pairs as entries in a vocabulary, and provide the validated token pairs in response to a search query containing a keyword corresponding to at least one entry in the vocabulary.