Abstract: Techniques are disclosed for abstractive summarization process for summarizing documents, including long documents. A document is encoded using an encoder-decoder architecture with attentive decoding. In particular, an encoder for modeling documents generates both word-level and section-level representations of a document. A discourse-aware decoder then captures the information flow from all discourse sections of a document. In order to extend the robustness of the generated summarization, a neural attention mechanism considers both word-level as well as section-level representations of a document. The neural attention mechanism may utilize a set of weights that are applied to the word-level representations and section-level representations.

Abstract: Techniques are disclosed for abstractive summarization process for summarizing documents, including long documents. A document is encoded using an encoder-decoder architecture with attentive decoding. In particular, an encoder for modeling documents generates both word-level and section-level representations of a document. A discourse-aware decoder then captures the information flow from all discourse sections of a document. In order to extend the robustness of the generated summarization, a neural attention mechanism considers both word-level as well as section-level representations of a document. The neural attention mechanism may utilize a set of weights that are applied to the word-level representations and section-level representations.

Abstract: A system that can be used for targeted radiology resident training can include a memory storing computer-executable instructions and a processor to access the memory and execute the computer-executable instructions to at least receive a preliminary report and a corresponding final report; determine a difference between the final radiology report and the preliminary radiology report; classify the difference as substantive or stylistic based on a property of the difference; and produce an output including the difference when classified as substantive. The output can include one or more critical errors reflected in the substantive difference. The one or more critical errors can be used to facilitate radiology resident training.