System, report, and method for generating natural language news-based stories

• US 8,676,691 B2
• Filed: 07/16/2013
• Issued: 03/18/2014
• Est. Priority Date: 06/06/2007
• Status: Expired due to Fees

• Current Assignee: O2 Media LLC
• Original Assignee: O2 Media LLC
• Inventors: Schiller, Victor Hugo
• Primary Examiner(s): NGUYEN, TIEN C
• Application Number:
  US13/943,383
• Publication Number:
  US 20130305148A1
• Time in Patent Office: 245 Days
• Field of Search: 705/35, 707/769, 707/737, 707/700, 713/156, 704/9, 709/202, 715/234
• US Class Current: 705/36
• CPC Class Codes: G06F 40/166 Editing, e.g. inserting or deleting
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis

Abstract

The present invention generally relates to a system, report, and method for automatically generating a series of natural language news-based stories to be presented via a digital interface or printed publication to a portfolio user. The disclosure relates to a filter or selection of a handful of relevant and desired financial instruments, or events created in a large group of events such as sports results, travel information, auction related data, online shopping tools, social media, retail store promotion generation, search engine daily report, etc. for a specific use. These financial instruments, based on different selections from a portfolio manager via a management tool, are then used to either produce a strategies page where a list of useful covered call trade and hedged trade are displayed in the form of a table, or natural language news-based stories relating to a selected list of financial instruments found in a portfolio. The events are based on different selections from a portfolio manager via a management tool and are then used to either produce a secondary page where a list of the selected event data is displayed or natural language news-based stories relating to a selected list of events found in a portfolio from a large event database.

Claims

1. A method of generating natural language news-based stories as part of a display with a plurality of natural news-based stories, the method comprising the steps of:

using a management tool to select items of interest within an event data database to create a portfolio made from the items of interest;

selecting data of the portfolio from the items of interest using a computer interface, wherein some data is in numerical format;

offering a choice on the computer interface between creating a strategies page where the data is displayed in the numerical format as part of the strategies page or creating a natural language news stories page where the data is displayed as part of a plurality of natural language news stories, wherein the natural language news stories as part of the natural language news stories page each including a date of publication, a story rewritten in natural language using a template from a news brief with a news-generating algorithm with sentence and paragraph construction to simulate natural language from the numerical format data selected from a plurality of templates each with input variables and at least a data field from the data, and a source of origin of the news brief, wherein the algorithm with sentence and paragraph reconstruction includes a series of boilerplate sentences varied using a parsing table; and

creating and displaying either the strategies page or the natural language news stories page to a portfolio user.

2. The method of claim 1, wherein the algorithm with sentence and paragraph reconstruction to simulate natural language include any one of the list:

merging information, removing irrelevant data, conforming dates for convenience, conforming dates for consistency.

3. The method of claim 1, wherein the items of interest are selected from a group consisting of investment instruments, sports journalism, hotel location, travel data, automated daily recap of items on a wish list in an auction, selection of retail product families, daily activity breakdown of friends posts on social media, daily reports on search engines, marketing program promotions.

4. The method of claim 1, wherein the data is selected from a group consisting of a number of records, a number of records, an investment symbol, a headline, an alternate headline, a story, a topic, a chrono, a sport, a number of stories, a team, and a player.

5. A system of generating natural language news-based stories of items of interest held in a portfolio, the system comprising:

a computer connected to an Internet, wherein the computer comprising a memory and a central processing unit in functional communication with the memory, and

a software executed by the central processing unit in the memory, wherein the software being used to selectively execute either a strategies page or a natural language news stories page with a plurality of natural language news stories of a plurality of items of interest held in a portfolio as defined by a management tool of the selection of items of interest; and

a data retrieval system connected to the Internet in data connection with the computer, with a user interface,

a central processing unit connected to the user interface, and

a display, wherein the software generates either a strategies page or a natural language news stories page based on data from each of a plurality of items of interest in the portfolio, wherein each natural language news story as part of the natural language news stories page includes a date of publication, a news item rewritten in natural language using a template from a news brief using an news-generating algorithm with sentence and paragraph construction to simulate natural language from a partly numerical format data selected from a plurality of templates each with input variables and at least a data field from the data, and wherein the data retrieval system retrieves either the strategies page or the natural language news stories page from the personal computer via the Internet, wherein the algorithm with sentence and paragraph construction includes a series of boilerplate sentences to be repeated, and wherein terms part of the boilerplate sentences are varied using a parsing table.

6. The system of claim 5, wherein the algorithm with sentence and paragraph reconstruction to simulate natural language include any one of the list:

merging information, removing irrelevant data, conforming dates of convenience and consistency.

7. The system of claim 5, wherein the items of interest are selected from a group consisting of investment instruments, sports journalism, hotel location, travel data, automated daily recap of items on a wish list in an auction, selection of retail product families, daily activity breakdown of friends posts on social media, daily reports on search engines, marketing program promotions.

8. The system of claim 5, wherein the data is selected from a group consisting of a number of records, a number of records, an investment symbol, a headline, an alternate headline, a story, a topic, a chrono, a sport, a number of stories, a team, and a player.

9. A system of generating natural language news-based stories of events held in a portfolio, the system comprising:

a computer connected to an Internet, wherein the computer comprising a memory and a central processing unit in functional communication with the memory, and

a software executed by the central processing unit in the memory, wherein the software being used to selectively execute a news report a plurality of natural language news stories of a plurality of items of interest held in a portfolio as defined by a portfolio manager, wherein the natural language news stories page comprises a plurality of natural language news stories, each including a date of publication, a story rewritten in natural language using a template from a news brief with a news-generating algorithm with sentence and paragraph construction to simulate natural language from a partly numerical format data selected from a plurality of templates each with input variables and at least a data field from a data selected from the items of interest in the portfolio, and a source of origin of the news brief, and wherein some of the data is numerical and used by the news-generating algorithm to simulate natural language, wherein the algorithm with sentence and paragraph construction includes a series of boilerplate sentences to be repeated, and wherein terms part of the boilerplate sentences are varied using a parsing table.

10. The system of claim 9, wherein the algorithm with sentence and paragraph reconstruction to simulate natural language include any one of the list:

merging information, removing irrelevant data, conforming dates of convenience and consistency.

11. The system of claim 9, wherein the items of interest are selected from a group consisting of investment instruments, sports journalism, hotel location, travel data, automated daily recap of items on a wish list in an auction, selection of retail product families, daily activity breakdown of friends posts on social media, daily reports on search engines, marketing program promotions.

12. The system of claim 9, wherein the data is selected from a group consisting of a number of records, a number of records, an investment symbol, a headline, an alternate headline, a story, a topic, a chrono, a sport, a number of stories, a team, and a player.