Electronic publishing, although Information Technologies advancements, are still based in the print text model. The textual format prevents programs to semantic process articles content. A semantic model of scientific electronic publishing is proposed, in which conclusion are prompted by author and recorded in machine-understandable format, enabling semantic retrieval, identification of traces of scientific discoveries and knowledge misunderstandings. The model is based on concepts as deep, or semantic, structure of human language (CHOMSKY, 1975), of microstructure, macrostructure and superstructure (KINTSH & VAN DIJK, 1972), of rhetoric structure of scientific articles (HUTCHINS, 1977), (GROSS, 1990) and on scientific methodology semantic elements, such as problem, question, objective, hypothesis, experiment and conclusion. It results from analysis of 89 biomedical articles. A prototype system was developed which partially implements the model. Questionnaires with authors were used to test the prototype development. The prototype was also tested with several researchers-authors. Four patterns of reasoning and sequencing of semantic elements were identified in articles analyzed. The content model is implemented as a computational ontology. A prototype of a web author’s submission interface to a electronic journal system was developed and tested.