The progress of Natural Language Processing through the eighties and the nineties allowed us to build large collections of text in electronic format annotated with all kind of linguistic information. These corpora are often called "treebanks", alluding to the tree-like structure of the syntactic analysis of natural language. More recently, researchers have started to employ more sophisticated tools to analyze the meaning of natural language, thus creating "meaning banks". In this talk I will present the Groningen Meaning Bank, a free resource of semantically annotated English text, with particular emphasis on the software toolchain that performs the analysis of the meaning. In the second part of the talk, I will present the work in progress on integrating the GMB software pipeline with the Linked Open Data ecosystem via entity linking. The integration aims to create an end-to-end system inspired by the Learning by Reading paradigm. The system extracts DBpedia entities and relations from raw text, and it is being used in the context of the ALOOF project in order to enable domestic robots learn new information about their environment by "reading" the Web.