Go Daddy Operating Company, LLC

Locu, Inc.

Google Inc.

Celo Foundation

Jason Ansel

Marek Olszewski

Keir Banks Mierle

Rene Reinsberg

Adam Marcus

Stylianos Sidiroglou

Marc Piette

David Sehr

Ulfar Erlingsson

Cliff L Biffle

Silas Boyd-Wickizer

A training set may be created to train a machine learning engine, such as an artificial neural network (ANN), to value a target domain name using data from previously sold domain names. The training set may comprise a plurality of word features vector of real numbers (information related to the words or tokens within the sold domain names), a plurality of word embedding vector of real numbers (word embedding of the words within the sold domain names), a plurality of context embedding vector of real numbers (sale context, i.e., location and date of a sale of a sold domain name), a plurality of DNS embedding vector of real number (DNS information of the sold domain name) and/or a plurality of domain name features vector of real numbers (data regarding the sold domain name). The ANN may then be trained on the training set, using the methods of gradient descent and back propagation, to value a target domain name.

Using a value for a domain name determined by a machine learning engine

Systems and techniques are provided for a tree structure for byzantine fault tolerance. A validator computing device may receive a candidate block from a block producer. The block producer may be a second validator computing device of the distributed network. The validator computing device may determine whether or not the candidate block is valid. The validator computing device may sign an indication of whether or not the candidate block is valid to generate a signature. The validator computing device may send the signature to a third validator computing device that is a sibling node to the validator computing device. The validator computing device may receive a second signature from the third validator computing device. The validator computing device may combine the signature and the second signature to generate a combine signature if the signature and the second signature agree.

Tree structure for byzantine fault tolerance

Illustrative embodiments improve upon prior machine learning techniques by introducing an additional classification layer that mimics human visual pattern recognition. Building upon classification passes that extract contextual information, illustrative embodiments look for hints of high-level semantic categorization that manifest as visual artifacts in the document, such as font family, font weight, text color, text justification, white space, or CSS class name. An improved lightweight markup language enables display of machine-categorized tokens on a screen for human correction, thereby providing ground truths for further machine classification.

Method and apparatus for forming a structured document from unstructured information

Systems and methods of the present invention provide for one or more server computers communicatively coupled to a network and configured to: access a source code for each of a plurality of web pages within a website hosted on the server computer; identify, within the source code of each of the plurality of web pages, a plurality of hyperlinks for navigating to at least one of the plurality of web pages; generate a plurality of link groups each comprising at least one common hyperlink between the plurality of hyperlinks; aggregate the plurality of link groups into a unique link group wherein each of the plurality of hyperlinks appears in only one link group; and determine an order of hyperlinks within the unique link group based on an original order of the plurality of hyperlinks.

Ordering documents within a crawled website

A system and method for document classification are presented. An input document is received (e.g., by at least one server communicatively coupled to a network). A plurality of features are identified in the input document. The plurality of features include sequences of text extracted from the input document. A feature vector of the input document is generated based upon the sequences of text, and the feature vector of the input document is compared to each of a plurality of signature vectors to determine a primary type of the input document. The primary type of the input document is stored into a storage system in communication with the at least one server.

Identifying and processing a number of features identified in a document to determine a type of the document

A system and method for data classification are presented. A plurality of training tokens are identified by at least one server communicatively coupled to a network. Each training token includes a token retrieved from a content source and a classification of the token. For each training token in the plurality of training tokens, a plurality of n-gram sequences are identified, a plurality of features for the plurality of n-gram sequences are generated, and first training data is generated using the token retrieved from the content source, the plurality of features, and the classification of the token. A first classifier is trained with the first training data, and the first classifier is stored into a storage system in communication with the at least one server.

Automated data classification

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for runtime language-independent sandboxing of software. In one aspect, a system implements an extended Software Fault Isolation (SFI) software sandboxing system configured to provide a user-mode program interface for receiving runtime requests for modifying verifiably safe executable machine code. Requests can include dynamic code creation, dynamic code deletion, and atomic modification of machine code instructions. A runtime modification of a verifiably safe executable memory region is made in response to each received runtime request, and code within the modified memory region has a guarantee of safe execution.

Growing community of inventors

Cambridge, MA, United States of America

Jason Ansel