Interpretable Topic Extraction and Word Embedding Learning Using Row-Stochastic DEDICOM

Hillebrand, Lars PatrickLars PatrickHillebrandBiesner, DavidDavidBiesnerBauckhage, ChristianChristianBauckhageSifa, RafetRafetSifa2022-03-142022-03-142020https://publica.fraunhofer.de/handle/publica/41000710.1007/978-3-030-57321-8_22The DEDICOM algorithm provides a uniquely interpretable matrix factorization method for symmetric and asymmetric square matrices. We employ a new row-stochastic variation of DEDICOM on the pointwise mutual information matrices of text corpora to identify latent topic clusters within the vocabulary and simultaneously learn interpretable word embeddings. We introduce a method to efficiently train a constrained DEDICOM algorithm and a qualitative evaluation of its topic modeling and word embedding performance.en005006629Interpretable Topic Extraction and Word Embedding Learning Using Row-Stochastic DEDICOMconference paper