Applying LDA topic model to a corpus of Italian Supreme Court decisions

We investigate the use of topic models such as latent Dirichlet allocation (LDA) on two real-world problems: classifying documents and retrieving similar documents to a given document, with reference to a corpus of Italian Supreme Court decisions. A topic model is a generative model that specifies a simple probabilistic procedure by which documents in a corpus can be generated. In the LDA approach, a topic is a probability distribution over a fixed vocabulary of terms, each document is modeled as a mixture of K topics, and the mixing coefficients can be used for representing documents as points on K-1 dimensional simplex spanned by the topics. Approximate posterior inference is performed in order to learn the hidden topical structure from the observed data, i.e., the words in the documents.