We describe our research on automatically generating rich semantic annotations of text and making it available on the Semantic Web. In particular, we discuss the challenges involved in adapting the OntoSem natural language processing system for this purpose. OntoSem, an implementation of the theory of ontological semantics under continuous development for over fifteen years, uses a specially constructed NLP-oriented ontology and an ontologicalsemantic lexicon to translate English text into a custom ontology-motivated knowledge representation language, the language of text meaning representations (TMRs). OntoSem concentrates on a variety of ambiguity resolution tasks as well as processing unexpected input and reference. To adapt OntoSem's representation to the Semantic Web, we developed a translation system, OntoSem2OWL, between the TMR language into the Semantic Web language OWL. We next used OntoSem and OntoSem2OWL to support SemNews, an experimental web service that monitors RSS news sources, processes the summaries of the news stories and publishes a structured representation of the meaning of the text in the news story. 2 The alternative on the supply side is, then, automatic annotation. Within the current state of the art, automatically produced annotations are roughly at the level attainable by the latest information extraction techniques -a reasonably good level of capturing named entities with a somewhat less successful categorization of such entities (e.g., decidingwhether Jordan is used as the first name of an individual or a reference to the Hashemite kingdom). Extracting more advanced types of semantic information, for example, types of events (to say nothing about determining semantic arguments, "case roles" in AI terminology), is not quite within the current information extraction capabilities, though work in this direction is ongoing. Indeed, semantic annotation is at the moment an active subfield of computational linguistics, where annotated corpora are intended for use by machine learning approaches to building natural language processing capabilities.On the demand side of the Semantic Web, a core capability is improving the precision of the Web search which will be facilitated by detailed semantic annotations that are unambiguous and sufficiently detailed to support the search engine in making fine-grained distinctions in calculating scores of documents. Another core capability is to transcend the level of document retrieval and instead return as answers to user queries specially generated pragmatically and stylistically appropriate responses. To attain this capability, intelligent agents must rely on very detailed semantic annotations of texts. We believe that such annotations will be, for all intents and purposes, complete text meaning representations, not just sets of semantic or pragmatic markers (and certainly not templates filled with uninterpreted snippets of the input text that are generated by the current information extraction methods).To attain such goals, Semantic Web agents must be ...