Abstract: Media, methods, and systems are disclosed for automatically calculating predicted values using deep learning models. One or more input forms having a plurality of input form field values are received. The input form field values are automatically parsed into a set of computer-generated candidate standard field values. The set of candidate standard field values are automatically normalized into a data frame having a set of normalized field values. In response to determining which portions of the data frame to apply to a line calculation neural network, a corresponding line calculation neural network is applied. At least one output form line calculation is performed. A natural language explanation regarding the at least one output form line calculation is generated. Additional user provided inputs are received in response to the natural language explanation, and at least one remaining calculation is carried out based on the one or more additional user provided inputs.

Abstract: Media and method for a user interface for training an artificial intelligence system. Many artificial intelligence systems require large volumes of labeled training data before they can accurately classify previously unseen data items. However, for some problem domains, no pre-labeled training data set may be available. Manually labeling training data sets by a subject-matter expert is a laborious process. An interface to enable such a subject-matter expert to accurately, consistently, and quickly label training data sets is disclosed herein. By allowing the subject-matter expert to easily navigate between training data items and select the applicable labels, operation of the computer is improved.

Abstract: Media, methods, and systems are disclosed for applying a computer-implemented model to a table of computed values to identify one or more anomalies. One or more input forms having a plurality of input form field values is received. The input form field values are automatically parsed into a set of computer-generated candidate standard field values. The set of candidate standard field values are automatically normalized into a corresponding set of normalized field values, based on a computer-automated input normalization model. An automated review model controller is applied to automatically identify a review model to apply to the set of normalized field values, based on certain predetermined target field values. The automatically identified review model is then applied to the set of normalized inputs, and in response to detecting an anomaly, a field value is flagged accordingly.

Abstract: An electronic tax assistant for assisting a user in preparing for a tax return preparation process, wherein the electronic tax assistant performs the following steps: determine a trigger based upon a data input, wherein the trigger is indicative of a potential tax implication; determine a data request based at least in part on the data input, wherein at least one of the data input and the data request is associated with an external device; receive a response based upon the data request, wherein the response is indicative as to whether the trigger is indicative of an actual tax implication; and provide, upon determining the actual tax implication, information in indicative of the data input and the response to an external repository associated with a tax preparation system.

Abstract: Matching a customer with a representative is performed for assisting the customer with an issue. Information related to the customer, the issue, historical information, and other customer information may be used to determine a likely issue. A representative may be selected in real-time to assist the customer in resolving the issue based on the customers experience in resolve the same or similar issues to the likely issue and the information related to the customer and the representative may be updated dynamically in real-time.

Abstract: System and media for determining whether to conduct a transaction with a consumer claiming an identity. Identity points for the identity are stored in a first blockchain. Past transactions conducted by the party are stored in a second blockchain. Permission to read and/or write to the various blockchains are stored in a third blockchain. When the consumer desires to conduct a transaction, they can append a permission for the counterparty to access the identity and/or transaction chains to their permission chain. The counterparty can then make an informed decision as to whether to conduct the transaction with the consumer based on the information stored in the various blockchains.

Abstract: Media, methods, and systems are disclosed for automatically calculating predicted values using deep learning models. One or more input forms having a plurality of input form field values are received. The input form field values are automatically parsed into a set of computer-generated candidate standard field values. The set of candidate standard field values are automatically normalized into a data frame having a set of normalized field values. In response to determining which portions of the data frame to apply to a line calculation neural network, a corresponding line calculation neural network is applied. At least one output form line calculation is performed. A natural language explanation regarding the at least one output form line calculation is generated. Additional user provided inputs are received in response to the natural language explanation, and at least one remaining calculation is carried out based on the one or more additional user provided inputs.

Abstract: System and media for determining whether to conduct a transaction with a consumer claiming an identity. Identity points for the identity are stored in a first blockchain. Past transactions conducted by the party are stored in a second blockchain. Permission to read and/or write to the various blockchains are stored in a third blockchain. When the consumer desires to conduct a transaction, they can append a permission for the counterparty to access the identity and/or transaction chains to their permission chain. The counterparty can then make an informed decision as to whether to conduct the transaction with the consumer based on the information stored in the various blockchains.

Abstract: Methods and a system for generating and employing dimensionality-reduction for generating statistical models. Traditionally, constructing models used for statistical analysis has been a laborious process requiring significant amounts of analyst input and processing time. Embodiments of the inventions address the above problem by providing an efficient method for pre-screening variable to automatically prune the dimensionality of the input space to a tractable size while retaining the most predictive power possible.

Abstract: Tax payoffs are determined for potential income source exclusions, for potential tax deductions, for potential tax credits, and for other tax implications. Identifying the tax payoff may be based at least in part on the likeliness that the tax implication will apply to the taxpayer's situation, the expected monetary impact on the taxpayer's tax burden if it does apply, and the expected effort required to claim the tax implication. Based upon the tax payoffs, embodiments of the invention may then recommend potential tax implications, prioritize potential tax implications, eliminate potential tax implications from consideration, or take other actions to assist the taxpayer in efficiently and effectively prepare their tax return.

Abstract: System, method and media for the provision of relevant information, tax regulations and other advice for the preparation of the return to the user of the tax preparation system. This can be done by developing a profile of the taxpayer as the user moves through the process of preparing the tax return. When a user views a tip or help article, the impact on the outcome of the return is noted. If the impact is a positive one, then the tip or help article may be automatically presented to future users with similar taxpayer profiles. In this way, users can maximize the benefit of the return for the taxpayer while increasing the accuracy of the prepared return.

Abstract: Embodiments of the invention provide systems and methods of determining the authenticity of a document by creating a document fingerprint from known authentic documents and comparing the document to the document fingerprint. A plurality of authenticated documents may be compared to determine like characteristics. The documents containing like characteristics may be categorized based on the similarities while determining that the documents are of the same type, from a same company, and for a same user. Information from a data layer, user interface layer, and a device associated with the document may be obtained and a document fingerprint for the document created using the information. Upon receipt of a document submitted for authenticity, the document may be similarly categorized and compared to the document fingerprint for authenticity. A likelihood of authenticity and a fraud score may be determined and a rule, based on the fraud score and the likelihood, may be applied.

Abstract: Items are identified and valuated based at least in part on an image that shows a depiction of the one or more items. The image is analyzed to identify the items, such as by a categorization. The items are valuated based upon the description, category, and/or additional information from the user about the items. The valuation is also based upon information from a valuation database. The identification and valuation may then be used for any of numerous purposes.

Abstract: A system, a method, and computer-readable media for opportunistically authenticating a taxpayer. Specifically, embodiments of the invention leverage the fact that the user has possession of particular documents or access to certain information as evidence that the user is the person referred to in those documents or information. If the user provides sufficient evidence to authenticate themselves while providing the information required for the financial transaction, no separate authentication step may be required. At a high level, documents or other data imported in a first context (e.g., during the process of preparing a tax return for a user) are used as evidence of the user's authenticity in a second context.

Abstract: Described is identifying entities based at least in part on the location of a user device. A server or the user device will determine the location of the user device and/or whether there are any beacons, such as proximity beacons, in the vicinity. The server or the user device will then determine where the user is and if there are any identifiable entities within a certain distance to that location. In performing these steps, the server and/or the user device may access a database for the geographical area. The user device will then receive information related to a function as indicated by the user, such as the facilitation of a donation, a non-obligated payment, or an obligated payment to the entity.

Abstract: Media and method for presenting a user interface for verifying automatically recognized information from forms. Embodiments of the invention may import an image of a form and break the form down into a series of bounding boxes, each including the value of a form item and a description and/or label for the form item. When making use of a particular value from the form image, embodiments of the invention may present the bounding box containing that value to the user, together with enough context for the user to verify that the correct item has been identified and that the automatically recognized value is correct.

Abstract: A method, system, and computer program for presenting a tax interview in preparation for preparing a tax return and, more particularly, to an advanced tax interview engine that integrates with third-party data sources including social networking sites, professional networking sites, credit reports, data aggregators, and other sources so as to minimize the burden on a user. One or more profiles for the taxpayer are retrieved from their respective third-party data sources so as to pre-populate responses to the tax interview, present additional portions of the tax interview, or determine discrepancies with user-entered information.

Abstract: Methods and media for predicting responses to discovery questions while interviewing a user. Embodiments of the invention begin by collecting information about the user from various sources. Factors affecting likely responses to discovery questions are identified and a score for each discovery question is determined. These scores may be adjusted as additional information is gathered and as the user answers other interview questions. Then, for each discovery question, the score is compared to a first threshold. If the score is above that first threshold, then the discovery question can be skipped. Otherwise the score is compared to a second, lower threshold. If the score is above that threshold, the likely answer to the discovery question can be presented to the user to confirm. Otherwise, the discovery question is presented to the user to answer.

Abstract: Media, method and system are described for creating and operating software capable of performing steps to provide a user with an estimated amount of time until a tax return application may be completed. Embodiments of the invention consider the availability of documents in physical or electronic form. Embodiments of the invention further display the time estimation to enable a user to reserve enough time for completion of the application.