Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating rendering data that when rendered on a display device presents a graphical user interface that displays a first visual representation of a value curve, a time period selection window that is configured to move along the first visual representation of the value curve to select one or more portions of the first visual representation of the value curve, a second visual representation of the value curve that is based on the selected one or more portions of the first visual representation of the value curve, a first events bar that includes two or more first event icons that are each associated with a different type of event, and a second events bar that includes two or more second event icons that are each associated with a same type of event.

Abstract: Methods and systems for dynamically updating machine learning models such as text classifiers. One of the methods includes: receiving first data; producing a first machine learning model using the first data; releasing the first machine learning model for use; receiving second data after receipt of the first data; determining that the second data has a difference metric relative to the first data that exceeds a difference threshold; retraining the first machine learning model using at least part of the second data, the retraining producing a second machine learning model; and releasing the second machine learning model for use.

Abstract: Techniques for prediction of financial instrument returns, identifying statistical history, the discovery of pricing anomalies, and financial instrument visualization are disclosed. In one particular exemplary embodiment, the techniques may be realized as a method for identifying financial instrument returns and pricing anomalies including matching, using at least one computer processor one or more portions of current market data associated with a financial instrument with historical market data, averaging outcomes of matched historical market data, and providing a probabilistic outcome for financial instrument returns, pricing anomalies, or other metrics based on the matched historical market data and the current market data.

Abstract: Methods, and systems for named entity recognition and disambiguation. One of the methods includes: receiving a document; extracting a plurality of mentions within the document; clustering the plurality of mentions to produce at least one cluster; identifying a plurality of candidate entities for the cluster; pairing each candidate entity and the cluster to create a plurality of candidate entity-cluster pairings; generating features for each candidate entity-cluster pairing; and selecting an entity, from the plurality of candidate entities, for the cluster based at least in part on features of a candidate entity pairing.

Abstract: Methods, and systems for a natural language interface with real-time feedback. One of the methods includes receiving an n-gram of characters; identifying a first prefix token within the n-gram of characters; identifying a plurality of possible interpretations of the n-gram of characters; for each possible interpretation, attaching the first prefix token to a token sequence; removing the first prefix token from the n-gram of characters to produce an n-gram remainder; identifying a second prefix token within the n-gram remainder using a tokenizer; identifying a plurality of remaining possible interpretations based on the first prefix token and the second prefix token; for each remaining possible interpretation attaching the second prefix token to a token sequence; applying grammatical rules to each token sequence to determine that each token sequence is valid; and forwarding interpretation data associated with at least one valid token sequence.

Abstract: Methods and systems for dynamic data clustering. One of the methods includes: i) receiving a vector representing a (data) item; in response to receiving the vector: a) obtaining clusters within at least one specified metric of the vector to determine a neighborhood of clusters; b) filtering the neighborhood of clusters based at least in part on an age parameter of the clusters to determine whether there are filtered clusters; c) creating a new cluster based at least in part on the item when there are no filtered clusters; and d) storing the new cluster data for the new cluster, wherein the obtaining, filtering, creating and storing steps are performed in near real-time from receipt of the vector; ii) receiving a request for cluster data; and iii) providing cluster data including the new cluster data in response to the request.

Abstract: Methods, systems, apparatus, including computer programs encoded on computer storage medium, for generating rendering data. The method includes actions of receiving a stream of real-time data, analyzing the stream of real-time data to detect a set of one or more attributes associated with a real-time event, identifying, based on the one or more attributes, one or more data records representing an event-outcome pair, wherein an outcome in the event-outcome pair includes data identifying a content item, obtaining historical data associated with the content item, and generating rendering data including data for rendering a first layer representing real-time data specifying a current value of the content item as an overlay to a second layer representing the obtained historical data associated with the content item, with each of the rendered layers being juxtaposed to one or more specified portions of the graphical user interface.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for a search engine for identifying analogies. One of the methods includes: receiving data that includes a set of user selected features and a user selected feature weight for at least one of the user selected features; generating a reference vector based on the received data; determining a level of similarity between the generated reference vector and each of a plurality of analogy vectors that each represent a potential analogy; providing, for output on a display of a user device, a set of one or more analogies based on the determined level of similarity; receiving a subsequent request to evaluate the set of one or more analogies based on a common attribute; determining, for each of the one or more analogies, a representation of the common attribute for each analogy over a time period; and providing, for display on the user device, a representation of the common attribute for each analogy over the time period.

Abstract: Methods, systems, apparatus, including computer programs encoded on computer storage medium, for predicting an outcome in response to a detected event. In one aspect, the method includes actions of maintaining a database of data records with a knowledge graph, receiving a stream of real-time data that includes data records, and from time-to-time as the real-time data is received: identifying a set of one or more attributes associated with a real-time event represented by one of the data records received, identifying a data record in the knowledge graph representing an event-outcome pair, determining that an event of the event-outcome pair has historically been associated with triggering an outcome of the event-outcome pair, and, executing a set of rules defining logic associated with the event-outcome pair specifying that the outcome will occur at a time that is subsequent to a time in which the one of the data records is received.

Abstract: This specification describes systems and methods for a domain-based search engine. One aspect of the invention provides a computer implemented method including: receiving a query that identifies an entity; identifying a domain that has the entity as a member; obtaining an entity-to-domain weight; identifying a precedent set connected to the domain; obtaining a precedent-set-to-domain weight; identifying an event in the precedent set; determining a score for the event based at least in part on (i) the entity to domain weight and (ii) the precedent set to domain weight; and providing rendering data for the event when the score satisfies a specified threshold. The rendering data can include a selectable link and the method can further include receiving a selection of the selectable link and in response to receiving the selection, providing rendering data that is indicative of an impact of the event on an entity.