Abstract: Computer-readable media, systems and methods may improve automatic summarization of transcripts of financial earnings calls. For example, a system may generate segments, such as by disambiguating sentences, from a transcript to be summarized. The system may use an estimator that assesses whether or not the segment should be included in the summary. Different types of estimators may be used. For example, the estimator may be rule-based, trained based on machine-learning techniques, or trained on based on machine-learning with language modeling using natural language processing to fine-tune language models specific to financial earnings calls.

Abstract: Computer-readable media, systems and methods may improve classification of content based on a machine-learning natural language processing (ML-NLP) classifier. The system may train a general language model based on a general corpus, further train the general language model based on a domain-specific corpus to generate a domain-specific language model, and conduct supervised machine-learning based on the domain-specific language using topic-specific corpus labeled as relating to topics of interest to generate the ML-NLP classifier. Accordingly, the ML-NLP classifier may be trained on a general corpus, further trained on a domain-specific corpus, and fine-tuned on a topic-specific corpus. In this manner, domain-specific content may be classified into topics of interest. The ML-NLP classifier may classify content into the topics of interest.

Abstract: The invention relates to systems and methods for distributing market data. In one implementation, the system may generate a new encryption key at each market data update, and use that key to encrypt each market participant's data in that update before it is sent. Among other factors, characteristics of modern computer networks may cause participants to be sent (and to receive) their encrypted data in that update at different times. After the participants have all been sent their data in that update the system may then simultaneously transmit to those participants the key that will enable them to decipher their data. In an implementation, the key may be transmitted via a multicast transport protocol which can be used to ensure all recipients receive it at the same time. In this manner the invention may ensure that although participants receive their data in a given update at different times, they are unable to decipher that data until substantially the same time.

Abstract: Systems and methods to facilitate credit risk assessment are described herein. The systems and methods described herein relate to implementing and training a credit risk model comprising a document model and a company model. The document model may be configured to read text of a document, understand long range relationships between words, phrases, and the occurrence of one or more financial events, and create a document score that indicates whether the financial events are likely to occur based on that document. A document-model-state vector may be generated that represents important features and relationships identified within each document and across a set of documents for a given entity based on the document scores. The company model may produce a sequence of default probability scores representing overall likelihoods of the occurrence of the financial events for an entity based on the document-model-state vector for documents associated with that entity.

Abstract: An electronic trading venue (ETV) may prioritize maker orders that specify or are associated with cancel delay values by selecting a maker order to be matched with taker orders before competing maker orders, at the same price level for a given instrument, that have smaller or no cancel delay values. Such prioritization may occur even if the competing maker orders were received before the maker order. A cancel delay value indicates an amount of time that must pass after a request to cancel a maker order is received before the ETV actually cancels the maker order. Thus, in exchange for prioritization, maker orders will be subject to a cancelation delay that starts after a cancel request is received and during which a maker order will not be canceled. Takers may benefit from this by being able to match with a maker order before it can be canceled.

Abstract: The invention relates to a system and method for providing a latency floor for an electronic trading venue in which market participants who can respond within the value the floor and choose to compete in a specific race to make or take a price may each have a substantially equal chance of winning that race. The system may detect and distinguish individual “races” that occur on an electronic trading venue. Upon detection of the first order (or message) in such a race, the system may create a batch and a timer for that race. As orders pertaining to that race are received, they are added to its batch. Upon the timer reaching a predetermined value, typically the value of the floor, the race is determined to have ended and the orders are drained from the batch for processing (e.g., against the instrument's central limit order book (CLOB)).

Abstract: Computer-readable media, systems and methods may improve an ability to validate a multipart transaction involving multiple parties and sub-transactions and securely store and access multipart transaction data through a peer-to-peer computer network and a distributed ledger. A multipart transaction, such as a Murabaha transaction, may refer to a transaction having multiple parts involving multiple parties in connection with an agreement between two or more of the parties in which a commodity sale or other vehicle is created based on the agreement. The commodity may be tokenized for transfer in the multipart transaction. Validation of digitally signed data relating to the sub-transactions may provide secure authentication of the sub-transactions and atomic transfers of the tokenized commodity between relevant parties of sub-transactions in connection with the agreement.

Abstract: Systems and methods to facilitate credit risk assessment are described herein. The systems and methods described herein relate to implementing and training a credit risk model comprising a document model and a company model. The document model may be configured to read text of a document, understand long range relationships between words, phrases, and the occurrence of one or more financial events, and create a document score that indicates whether the financial events are likely to occur based on that document. A document-model-state vector may be generated that represents important features and relationships identified within each document and across a set of documents for a given entity based on the document scores. The company model may produce a sequence of default probability scores representing overall likelihoods of the occurrence of the financial events for an entity based on the document-model-state vector for documents associated with that entity.

Abstract: An electronic trading venue (ETV) may prioritize maker orders that specify or are associated with cancel delay values by selecting a maker order to be matched with taker orders before competing maker orders, at the same price level for a given instrument, that have smaller or no cancel delay values. Such prioritization may occur even if the competing maker orders were received before the maker order. A cancel delay value indicates an amount of time that must pass after a request to cancel a maker order is received before the ETV actually cancels the maker order. Thus, in exchange for prioritization, maker orders will be subject to a cancellation delay that starts after a cancel request is received and during which a maker order will not be canceled. Takers may benefit from this by being able to match with a maker order before it can be canceled.

Abstract: The disclosure relates to systems and methods for symbology-based obfuscation. A system may access an input data value obtained from an input data message. The system may obtain a symbology encoding based on the input data message, the symbology encoding including a precision parameter. The system may generate an output data value based on the input data value and the symbology encoding, the output data value obfuscating the input data value based on the precision parameter defined in the symbology encoding. The system may prepare an output data message based on the output data value that obfuscates the input data value. The system may transmit the output data value obfuscating the input data value to one or more recipients to obfuscate the input data value while providing an indication of the input data value within the precision parameter.

Abstract: Multivariate graphs specific to a company's performance, alone or in relation to a benchmark, and methods for producing same include defining on a display screen an origin having a first value, extending from the origin at least three axes on the display screen, the axes being generally equidistant from each other and representing a respective variable, plotting on the display screen a value of each variable concerning the company as a point on a respective axis, and using the plotted points to interpolate first segments between the axes on the display screen. Benchmark information, extrinsic to the company, can be obtained, processed into a rating in the same manner as the company data, and superimposed for ready visual comparison of the company's performance to the obtained benchmark. The benchmark information, if shown, is distinguished from the company's performance graph.

Abstract: Computer-readable media, systems and methods may propagate lower layer data, in headers that would otherwise be deleted, to upper layers in a network protocol stack that uses a multi-layer communication model. For example, high precision network-level timestamps that indicate a time that binary data is received at a network switch may be propagated upward for online application-level processing. Network switches may ordinarily insert network-level timestamps in frame headers that are ordinarily removed prior to propagation upward. The network-level timestamps may be propagated via in-band propagation or out-of-band propagation.

Abstract: Systems and methods of improved network analytics are disclosed. A system may determine feature propagation in a network of nodes of a graph database. The system may compute, at scale, datasets having complex relationships using graph analysis to determine network effects of entities in a network of entities stored in a graph database. The system may identify entities of interest, which may be associated with a quantitative feature value. The system may compute paths from an entity to the entities of interest, centrality metrics for entities in each of the paths, and path lengths to determine network effects of the entity of interests on the entity. The system may use the computed network effects, taking into account types of relationships between entities in the paths, to determine scaled quantitative feature values for the entity that is subject to the network effects of the entities of interest.

Abstract: The present disclosure relates to methods and systems for ingesting content from data feeds and to generate an enhanced output of relevant content for presentation to a user to facilitate analysis of the relevant content. Content is received from data feeds, and filtered to identify relevant content with respect to a particular context. The relevant content is then processed, e.g., using natural language processes, to extract entities involved, and to also identify particular activities detailed in the relevant content. Activity-mining is applied to the identified relevant data to classify and assigned activity tags to the extracted entity. Based on the extracted and identified information, an enhanced output is generated for presentation to facilitate research operations. The enhanced output may include overlaid graphical annotations, indicators, and graphical controls over the relevant articles to provide a means for updating a database based on the relevant content.

Abstract: The disclosure relates to non-blocking messaging in an electronic trading venue (ETV) in which a matching engine sends groups of price-compatible contra-orders to a credit engine. Doing so may improve the ETV's efficacy and performance by eliminating the need for multiple roundtrips of messages in the case of a failed credit check, and/or by allowing the ETV to more faithfully implement its matching rules. Upon sending an order to the credit engine the matching engine marks it as ‘uncertain’ to indicate that additional requests received on it (e.g., cancels) must also be forwarded to the credit engine. When all pending requests on an order have been acknowledged by the credit engine the matching engine unmarks it as such. In addition to credit checking the credit engine may also perform computations related to advanced order types (e.g., icebergs) indicating the output of such in its acknowledgements to the matching engine.

Abstract: Systems and methods of a deterministic credit system are disclosed to provide high-performance credit-checking while reducing the need for horizontal scaling or implementing larger system architecture. The systems and methods presented herein may achieve this in several ways, for example, by: storing credit lines in memory; optimizing the storage of those lines in memory and performing precomputations on them so as to minimize operations performed on them; and minimizing disk and network I/O by (i) sequencing and persisting inputs to the credit checking computations and (ii) ensuring those computations are deterministic, consequently eliminating the need to persist or transmit the outputs of those computations.

Abstract: An embodiment of a method of providing identity assurance for a decentralized application (DApp) includes executing, by at least one distributed node of a blockchain system, an entitlement contract stored on the blockchain to perform a read call from a DApp contract stored on the blockchain, the read call including an address signing a transaction to the DApp contract. Performing the read call may include reading a list of registered addresses stored on the blockchain, determining whether the list includes the signing address; and providing an output indicating whether the list includes the signing address. The method may further include executing, by the at least one distributed node, a registry contract stored on the blockchain to perform a read call from the DApp contract, the read call including an identifier of the decentralized application.

Abstract: An embodiment of a method of providing financial data to a financial instrument smart contract in a distributed ledger system includes: receiving, by an oracle smart contract in the distributed ledger system, a transaction from the financial instrument smart contract, the transaction including registration data having an identification of financial data to be delivered and an identification of a schedule on which to deliver the financial data; generating a transaction including the financial data; and transmitting the generated transaction to at least one distributed node of the distributed ledger system according to the requested schedule.

Abstract: A method and system for generating and using a master entity associative data network. The master entity associative data network system includes data sources storing data entities. Each data entity is of a specific entity type having specific entity properties, associations to other data entities, and an assigned common identifier label. Also, the master entity associative data network system includes a data processor configured to extract data entities from the data sources and extract associations to other data entities. The data processor stitches the data entities to each other based on at least one or more of the extracted associations, specific entity properties, and the common identifier label to generate one or more master data entities.

Abstract: Systems and methods may use natural language processing (NLP) and machine-learning techniques to detect an impact that an event will have on a domain-specific topic. For example, the system may use multi-stage cleaning using a rules-based and an artificial intelligence (AI)-based filter to filter large quantities of event items that may not be relevant to a domain of interest. The AI-based filter may be trained using labeled event items that were previously known to be impactful. The system may cluster the cleaned event items to group similar event items and eliminate redundancy. The system may then predict and quantify the impact that events described by clustered event items will have on the domain-specific topic. Such prediction may be based on a classifier trained using various model features that correlate with impactful events, including prior similar events.