Data processing pipelines are a core object of interest for data scientist and practitioners operating in a variety of data-related application domains. To effectively capitalise on the experience gained in the creation and adoption of such pipelines, the need arises for mechanisms able to capture knowledge about datasets of interest, data processing methods designed to achieve a given goal, and the performance achieved when applying such methods to the considered datasets. However, due to its distributed and often unstructured nature, this knowledge is not easily accessible. In this paper, we use (scientific) publications as source of knowledge about Data Processing Pipelines. We describe a method designed to classify sentences according to the nature of the contained information (i.e. scientific objective, dataset, method, software, result), and to extract relevant named entities. The extracted information is then semantically annotated and published as linked data in open knowledge repositories according to the DMS ontology for data processing metadata. To demonstrate the effectiveness and performance of our approach, we present the results of a quantitative and qualitative analysis performed on four different conference series.