Converting unlabeled data into labeled data

1. A computer system comprising:

one or more processors;

one or more computer-readable storage media coupled to the one or more processors; and

a plurality of instructions, stored on the one or more computer-readable storage media and executable by the one or more processors to perform operations, the operations comprising:

extracting information from a plurality of paragraphs associated with a plurality of respective purchase orders, wherein the plurality of paragraphs has inconsistent formats, and generating purchase orders from the plurality of respective purchase orders with individual labeled fields representing the information in a consistent format by:

receiving a document comprising a paragraph associated with a purchase order, wherein the plurality of respective purchase orders comprises the purchase order;

converting the paragraph into digital information containing text; and

parsing the text to identify unlabeled data, to remove the unlabeled data, and to generate a data structure associated with the purchase order, wherein the generating comprises creating the data structure and populating the data structure with data, wherein the unlabeled data becomes labeled data by the populating the data structure, wherein the parsing comprises:

identifying, based on an email regular expression, an email address in the text;

removing the email address from the text;

adding the email address to an email field in the data structure;

identifying, based on a postal code regular expression, a postal code;

removing the postal code from the text;

adding the postal code to a postal code field in the data structure;

searching for digits in the text;

identifying a phone number from the digits in the text;

removing the phone number from the text;

adding the phone number to a phone number field in the data structure;

determining a street name and a street number in the text;

removing the street name and street number from the text;

adding the street name and street number to a first address field in the data structure;

determining a city name, a contact name, and a company name in the text;

removing the city name, contact name, and company name from the text;

storing the city name, contact name, and company name in the data structure, and

generating a plurality of pre-determined number of fields in the data structure.

2. The computer system of claim 1 , wherein converting the paragraph into text comprises:

using Optical Character Recognition (OCR) to convert the paragraph into text.

3. The computer system of claim 1 , wherein parsing comprises:

identifying one or more delimiters in the text;

breaking the text into words based on the one or more delimiters; and

identifying a particular portion of an address associated with the purchase order based on the one or more delimiters.

4. The computer system of claim 1 , wherein parsing comprises:

using a parse tree to identify a pre-determined number of fields.

5. The computer system of claim 1 , wherein parsing comprises:

identifying, based on machine learning, a company name or a contact name.

6. The computer system of claim 1 , wherein determining a city name, a contact name, and a company name in the text comprises:

using keywords associated with a city name, a contact name, or a company name to search the text.

7. The computer system of claim 1 , wherein the operations further comprise:

determining that the paragraph is stored in a first size of computer memory;

determining that the pre-determined number of fields are stored in a second size of computer memory; and

determining that the second size is less than the first size.

8. A computer-implemented method comprises:

extracting information from a plurality of paragraphs associated with a plurality of respective purchase orders, wherein the plurality of paragraphs has inconsistent formats, and generating purchase orders from the plurality of respective purchase orders with individual labeled fields representing the information in a consistent format by:

receiving a document comprising a paragraph associated with a purchase order, wherein the plurality of respective purchase orders comprises the purchase order;

converting the paragraph into digital information containing text;

repeatedly parsing, by a parser, the text to identify a particular field from the text, wherein the particular field is unlabeled data;

removing the particular field from the text; and

storing the particular field in a data structure associated with the purchase order, wherein the storing comprises creating the data structure and populating the data structure with data wherein the unlabeled data becomes labeled data by the populating the data structure;

until a pre-determined number of fields has been identified, removed, and stored.

9. The computer-implemented method of claim 8 , wherein converting the paragraph into text comprises:

using Optical Character Recognition (OCR) to convert the paragraph into text.

10. The computer-implemented method of claim 8 , wherein parsing comprises:

identifying, based on a phone number regular expression or one or more delimiters, a phone number in the text, the phone number comprising an extension of zero, three, four, or five digits.

11. The computer-implemented method of claim 8 , wherein parsing comprises:

identifying, based on a pre-determined list of state names in full form and abbreviated form, a state name in the text.

12. The computer-implemented method of claim 8 , wherein parsing comprises:

identifying, based on machine learning or a street name regular expression, a street name in the text.

13. The computer-implemented method of claim 8 , wherein parsing comprises:

identifying words corresponding to spelling of numbers;

converting the words into digits; and

identifying a street number based on the digits.

14. The computer-implemented method of claim 8 , wherein parsing comprises:

identifying, based on machine learning or a location regular expression, a floor number, a department name or number, a post office box number, a room number, a suite number, or a mail stop number.

15. One or more non-transitory computer-readable storage media to store instructions that are executable by one or more processors to perform operations, the operations comprising:

extracting information from a plurality of paragraphs associated with a plurality of respective purchase orders, wherein the plurality of paragraphs has inconsistent formats, and generating purchase orders from the plurality of respective purchase orders with individual labeled fields representing the information in a consistent format by:

receiving a document comprising a paragraph associated with a purchase order, wherein the plurality of respective purchase orders comprises the purchase order;

converting the paragraph into digital information containing text;

parsing, by a parser, the text to identify particular portions associated with a state name, a zip code, a city name, a street name and number, a phone number, an email address, a contact name, and a company name based on a pre-determined sequential order, wherein the particular portions are unlabeled data;

removing the particular portions from the text based on the pre-determined sequential order; and

storing the particular portions to fields in a data structure, wherein the storing comprises creating the data structure and populating the data structure with data wherein the unlabeled data becomes labeled data by the populating the data structure.

16. The one or more non-transitory computer-readable storage media of claim 15 , wherein converting the paragraph into text comprises:

using Optical Character Recognition (OCR) to convert the paragraph into text.

17. The one or more non-transitory computer-readable storage media of claim 15 , wherein parsing comprises:

identifying one or more delimiters in the text;

breaking the text into words based on the one or more delimiters; and

identifying one or more of the particular portions based on the one or more delimiters.

18. The one or more non-transitory computer-readable storage media of claim 15 , wherein parsing comprises:

identifying, based on an email address regular expression, an email address in the text.

19. The one or more non-transitory computer-readable storage media of claim 15 , wherein the pre-determined sequential order is as follows:

an email address;

a state name;

a zip code;

a phone number;

a contact name;

a street name and number;

a city name;

a floor number, a department name or number, a post office box number, a room number, a suite number, or a mail stop number;

a company name; and

a contact name.

20. The one or more non-transitory computer-readable storage media of claim 15 , wherein the pre-determined sequential order is as follows:

a contact name;

a zip code;

a state name;

a city name;

a floor number, a department name or number, a post office box number, a room number, a suite number, or a mail stop number;

a street name and number;

a company name;

a contact name;

an email address; and

a phone number.