Vía @jorgecortell una interesanteUna interesante entrada en entopix con muchos datos para implementar soluciones de Procesamiento de Lenguaje Natural, Natural Language Processing (NPL).
So, why is it possible for such complex software to be free? Researchers at universities are spending decades improving performance of many individual components and publish their results in conference papers. Releasing the software as open-source allows them to improve the state-of-the-art by collaborating on the same task over time and across universities, benchmark solutions against each other and using smaller components to build more complex NLP systems. It also gets researchers more citations. Everyone benefits.