# US 12,217,178 B2 · Computer-based systems configured for detecting and splitting data types in a data file and methods of use thereof > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,217,178 B2 - Document ID: US12217178B2 - Status: Granted - Kind: B2 - Application number: 18/295,153 - Filed: 2023-04-03 - Published: 2023-08-03 - Granted: 2025-02-04 - Earliest claimed: 2019-10-29 - Assignees: Capital One Services, LLC - Applicants: Capital One Services, LLC - Inventors: Galen Rafferty; Reza Farivar; Jeremy Goodsitt; Anh Truong; Austin Walters - Web page: https://patentgrep.com/patent/US12217178B2 ## CPC - G06F 40/279: Recognition of textual entities - G06N 3/08: Learning methods - G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)} - G06F 40/166: Editing, e.g. inserting or deleting - G06F 40/284: Lexical analysis, e.g. tokenisation or collocates - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/09: Supervised learning - G06N 20/10: using kernel methods, e.g. support vector machines [SVM] - G06F 40/18: of spreadsheets (form-filling G06F40/174) - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A method for training a neural network model includes generating a training dataset with a plurality of data types and word samples belonging to each data type. A plurality of character strings stored in a plurality of data fields in a first data file are received where the plurality of character strings includes at least one word belonging to at least one data type in the plurality of data types. The at least one word from each of the plurality of character strings in each of the data fields are split and matched to the at least one data type using the neural network model. An ad hoc second data file with a plurality of data vectors is constructed based on a user selection of data field labels where each data vector includes words matched to a data type with a respective data field label. ## Claims ### Claim 1 A method, comprising: receiving, by a processor, a plurality of character strings stored in a plurality of data fields in a first data file; based on at least one neural network model: (i) assigning, by the processor, to each character or at least one group of characters in each character string, a probability of belonging to at least one specific data type; and (ii) splitting, by the processor, each character string into at least one word based on the probability of belonging to the at least one specific data type; wherein the at least one neural network model has been trained to match word samples to data types in a plurality of data types based on a training dataset that comprises the plurality of data types and the word samples belonging to each data type in the plurality of data types; generating, by the processor, at least one data vector based on at least one predefined format; wherein the at least one data vector comprises the at least one word from each of the plurality of character strings for the at least one specific data type; and constructing, by the processor, a second data file based on the at least one data vector with the at least one predefined format. ### Claim 2 (depends on 1) The method according to claim 1, wherein the first data file and the second data file are selected from the group consisting of a data table, a spreadsheet, an Excel spreadsheet, a key-value store, a JSON object, an AVRO file, and a database file. ### Claim 3 (depends on 1) The method according to claim 1, wherein the plurality of data fields is arranged in an array of rows and columns. ### Claim 4 (depends on 3) The method according to claim 3, wherein each column in the array comprises data fields of one data type from the plurality of data types designated with a respective data type label from a plurality of data type labels. ### Claim 5 (depends on 4) The method according to claim 4, wherein constructing the second data file comprises arranging the at least one data vector as columns in the array of the second data file and formatting the columns in the array according to the at least one predefined format. ### Claim 6 (depends on 1) The method according to claim 1, further comprising generating the training dataset to train the at least one neural network model by assembling word samples belonging to each data type in the plurality of data types from a corpus. ### Claim 7 (depends on 1) The method according to claim 1, wherein the at least one neural network model is selected from the group consisting of a classifier neural network model, a convolutional neural network (CNN) model, a recurrent neural network (RNN) model, and a multilayer perceptron network. ### Claim 8 (depends on 1) The method according to claim 1, wherein outputting each character string with indicia indicating the at least one word comprises highlighting the at least one word in each character string. ### Claim 9 A system, comprising: a memory; and a processor configured to: receive a plurality of character strings stored in a plurality of data fields in a first data file; based on at least one neural network model: (i) assign to each character or at least one group of characters in each character string, a probability of belonging to at least one specific data type; and (ii) split each character string into at least one word based on the probability of belonging to the at least one specific data type; wherein the at least one neural network model has been trained to match word samples to data types in a plurality of data types based on a training dataset that comprises the plurality of data types and the word samples belonging to each data type in the plurality of data types; generate at least one data vector based on at least one predefined format; wherein the at least one data vector comprises the at least one word from each of the plurality of character strings for the at least one specific data type; and construct a second data file based on the at least one data vector with the at least one predefined format. ### Claim 10 (depends on 9) The system according to claim 9, wherein the first data file and the second data file are selected from the group consisting of a data table, a spreadsheet, an Excel spreadsheet, a key-value store, a JSON object, an AVRO file, and a database file. ### Claim 11 (depends on 9) The system according to claim 9, wherein the plurality of data fields is arranged in an array of rows and columns. ### Claim 12 (depends on 11) The system according to claim 11, wherein each column in the array comprises data fields of one data type from the plurality of data types designated with a respective data type label from a plurality of data type labels. ### Claim 13 (depends on 12) The system according to claim 12, wherein the processor is configured to construct the second data file by arranging the at least one data vector as columns in the array of the second data file and formatting the columns in the array according to the at least one predefined format. ### Claim 14 (depends on 9) The system according to claim 9, wherein the processor is configured to generate the training dataset for training the at least one neural network model by assembling word samples belonging to each data type in the plurality of data types from a corpus. ### Claim 15 (depends on 9) The system according to claim 9, wherein the at least one neural network model is selected from the group consisting of a classifier neural network model, a convolutional neural network (CNN) model, a recurrent neural network (RNN) model, and a multilayer perceptron network. ### Claim 16 (depends on 9) The system according to claim 9, wherein the processor is configured to output each character string with indicia indicating the at least one word by highlighting the at least one word in each character string.