Architecture for data processing and user experience to provide decision support

Abstract

A system for processing data to generate an output includes an automated tagging system that receives data from a plurality of alternate data providers, each of the plurality of data providers having different types of data; a company financial data unit that provides published information comprising annual reports, press releases, information from the social media of spokespersons or executives, and published pricing information; a modelling system that receives the standardized data of the automated tagging system and the company financial data, the modelling system including one or more handle generators and forecast builders that are applied to the artificial intelligence-based system that employs the neural networks to generate, as an output, a forecast; and a revenue prediction unit that receives the output as the forecast, comprising one or more revenue predictions, to generate a final output as a revenue prediction.

I/W claim: 
     
         1 . A system for processing data to generate an output, the system comprising:
 an automated tagging system that receives data from a plurality of alternate data providers, each of the plurality of data providers having different types of data;   wherein the automated tagging system standardizes the different types received from the different data providers by performing filtering, de-duplication, normalization, and classification;   wherein the automated tagging system performs updating and feedback by providing an artificial intelligence system that employs neural networks that use back propagation for periodic updates to a training phase;   a company financial data unit that provides published information comprising annual reports, press releases, information from the social media of spokespersons or executives, and published pricing information;   a modelling system that receives the standardized data of the automated tagging system and the company financial data, the modelling system including one or more handle generators and forecast builders that are applied to the artificial intelligence-based system that employs the neural networks to generate, as an output, a forecast;   a revenue prediction unit that receives the output as the forecast, comprising one or more revenue predictions, to generate a final output as a revenue prediction.   
     
     
         2 . The system of  claim 1 , wherein the different types of data comprise financial transaction information, location information, or consumer behavior information. 
     
     
         3 . The system of  claim 1 , wherein the input to the modelling system includes proprietary data associated with the merger and acquisition information of the company, and a fiscal quarters calendar of the company. 
     
     
         4 . The system of  claim 1 , wherein the final output comprises a recommendation to buy, hold or sell, as a transaction, a score, or an execution instruction associated with the transaction without involving the user. 
     
     
         5 . The system of  claim 1 , wherein a rule-based or other deterministic approach is combined the neural network. 
     
     
         6 . A system for processing data to generate an output, the system comprising:
 a plurality of inputs, each of the inputs being received from a data source, and each of the inputs being of a different type;   a big data system configured to receive each of the plurality inputs, the big data system including,
 a plurality of adapters corresponding to the plurality of inputs, each of the adapters configured to normalize, deduplicate and classify the data received from the plurality of inputs, to generate outputs, and 
 a modeling system that receives the generated outputs, and provides the generated outputs to a plurality of multi-panel generators, multi-forecast builders and forecaster modules, to generate a revenue prediction; and 
   an access point and provides the revenue prediction in the format of a file, an API, a console or a custom format.   
     
     
         7 . The system of  claim 6 , wherein data miners and analysts supervise and update the plurality of adapters. 
     
     
         8 . The system of  claim 6 , wherein the outputs of the plurality of adapters is combined with external data comprising financial reports or other publicly available information associated with historical or present characteristics of an entity. 
     
     
         9 . A computer-implemented method of processing data, the method comprising:
 a first phase associated with data processing, the first phase comprising, receiving input data from a plurality of sources of different types, normalizing the received data by mapping the received input data from organic formats in which the was received, to a standardized data format that provides consistency across the organic formats by accounting for the differences in the input data from the plurality of sources of different types, and deduplication, to generate normalized data; applying tagging rules to the normalized data, the tagging rules comprising rules that are specific to a credit card or debit card, geo-fencing rules for GPS data, and rules associated with browser history or application usage, to generated tagged data;   a second phase associated with development of a panel and calibration, the second phase comprising,
 performing panelization by establishing a sample of users as a panel based on one or more of input data churn rate, user transaction patterns, census data balancing, or another rules that associates a characteristic of user transaction behavior with a transaction grouping the panels by symbol to associate one or more brands of a company with the tagged data, wherein the grouping is performed in real time to incorporate mergers, acquisitions, spinoffs, bankruptcies, rebranding, listing or delisting associated with the company associated with the symbol; and 
 applying one or more corrections to the grouped panels by applying one or more respective patterns associated with a financial institution to calibrate the grouping, wherein the respective patterns comprise weekend postings of information, posting delays typically associated with a financial institution, and pending but not yet posted transactions, removing anomalies associated with anomalous transactions or anomalous users; and 
   a third phase associated with creating and applying a prediction model, the third phase comprising,
 generating a prediction model, wherein training data comprises historical data associated with the company including financial parameters of the company, stock price, and historical measurements, 
 generating a forecast by using the generated prediction model and features associated with a current time period, the forecast comprising a prediction of a future stock price. 
   
     
     
         10 . The computer-implemented method of  claim 9 , wherein the input data comprises timeseries data from multiple vendors, including credit card transactions, debit card transactions or other electronic purchase transactions. 
     
     
         11 . The computer-implemented method of  claim 9 , wherein the tagging comprises, for a series of financial transactions, labeling each of the financial transactions by applying the tagging rules to the series of financial transactions, wherein the tagging rules include an inclusive filter or an exclusive filter, and further comprising applying natural language processing based on neural networks.