Meta: A Unified Toolkit For Text Retrieval And Analysis

META is developed to unite machine learning, information retrieval, and natural language processing in one easy-to-use toolkit. Its focus on indexing allows it to perform well on large datasets, supporting online classification and other out-of-core algorithms. META's liberal open source license encourages contributions, and its extensive online documentation, forum, and tutorials make this process straight-forward. We run experiments and show META's performance is competitive with or better than existing software.