Abstract: Generating natural language text from structured data using a fusion model is disclosed. Based on an input dictionary, a first sequence of vectors is generated by a first encoder and a second sequence of vectors is generated by a second encoder. The first and second sequences of vectors are provided to an attention function which generates a modified sequence of vectors. A decoder decodes the modified sequence of vectors to generate a plurality of ordered sequences corresponding to a target natural language sentence. A predetermined number of candidate sentences are determined based on the plurality of ordered sequences and are ranked to select a sentence as the target natural language sentence.

Abstract: Methods, systems, and computer program products for extracting from a descriptive document the value of a slot associated with a target entity described in the descriptive document are disclosed. The descriptive document is split into a set of sentences, and the sentences are filtered to generate a set of candidate sentences. Candidate sentence-entity pairs are determined from the candidate sentences, where each candidate sentence-entity pair includes a candidate sentence and an associated entity of the same type as the slot. The candidate sentence-entity pairs are compared to a set of gold sentences associated with the slot to calculate a plurality of similarity measures. A candidate sentence-entity pair associated with a maximum similarity measure is determined, and the value of the associated entity of the determined candidate sentence-entity pair is assigned to the slot.