Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for normalizing a collection of structured data records generated by an extraction neural network to enable large-scale data processing and analysis for query answering. According to one aspect, there is provided a method that includes extracting text strings from the structured data records, embedding the text strings in a latent space, performing an iterative numerical clustering operation in the latent space to cluster the embeddings of the text strings, and then identifying standardized text strings based on a result of the clustering. The structured data records are normalized using the standardized text strings. The normalized collection of structured data records are then used to generate a response to a user query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing a large corpus of unstructured data using an extraction neural network to generate a corresponding collection of structured data records. According to one aspect, there is provided a method that includes obtaining a set of input text sequences, and generating a collection of structured data records from the set of input text sequences using an extraction neural network. Each structured data record defines a structured representation of a corresponding input text sequence with reference to a predefined schema of semantic categories. The collection of structured data records can be filtered to identify and remove structured data records that are predicted to be unreliable. The collection of structured data records can then be processed to generate an output that is directed to a selected topic and that aggregates information from across multiple structured data records.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for automatically generating an article on a selected topic. In one aspect, a method comprises: obtaining a set of input text sequences; generating a set of structured data records from the set of input text sequences using an extraction neural network, wherein each structured data record defines a structured representation of a corresponding input text sequence; filtering the collection of structured data records to identify and remove structured data records that are predicted to be unreliable; and processing the collection of structure data records to generate an article that is relevant to a selected topic and that aggregates information from across multiple structured data records.

Abstract: A computer-implemented method is disclosed for generating structured data representations from unstructured input text sequences through a neural network-based extraction pipeline. The method involves acquiring input text sequences and processing them with an extraction neural network to produce output text sequences. Each output text sequence includes embedded delimiter tokens that annotate specific text segments, mapping them to semantic categories defined within a structured schema. The annotated text segments encapsulate semantically relevant information extracted from the input sequence. The output text sequences are subsequently processed to generate structured data records.