System for generating data from social media messages for the real-time evaluation of publicly traded assets

• US 9,940,672 B2
• Filed: 03/22/2012
• Issued: 04/10/2018
• Est. Priority Date: 03/22/2011
• Status: Active Grant

• Current Assignee: Kubera Logic LLC
• Original Assignee: iSentium, LLC
• Inventors: Di Sciullo, Anna Maria, Karakozoff, Alex, Sastri, Gautham
• Primary Examiner(s): Nguyen, Dustin
• Application Number:
  US13/427,819
• Publication Number:
  US 20120324023A1
• Time in Patent Office: 2,210 Days
• Field of Search: 707 2, 707 3, 707706, 707723, 707748, 707740, 707754, 705 1449, 705 35, 704 9, 704 2, 709206-209, 715744, 706 20
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

A system for generating data from social media messages for the real-time evaluation of publicly traded assets includes an ingest component for ingesting the social media messages and a filter module eliminating expressions not considered useful language from social media messages and configuring input social media message into useful formats to form filtered social media messages. The system also includes a language processor processing the filtered social media messages based upon lexical databases to form filter and processed social media messages. The system further includes a sentiment calculator applying rules to the filtered and processed social media messages so as to compute a representation of sentiment values associated with the social media messages. A graphical user interface displaying the sentiment values is also provided.

Claims

1. A computer implemented system for generating data from social media messages of at most 140 characters for the real-time evaluation of publicly traded equities or commodities, comprising:

an ingest component in the form of a computer for ingesting social media messages of at most 140 characters;

a filter module of the computer eliminating expressions not considered useful language from social media messages and configuring input social media messages into useful formats to form filtered social media messages, wherein the filter module eliminates URLs and hash tags from the social media messages;

a language processor of the computer processing the filtered social media messages based upon lexical databases to form filtered and processed social media messages, wherein the language processor includes a stock-based lexical repository composed of lexical items used in the ingested social media messages, as well as lexical items used in stock exchange and financial news wire, wherein the lexical items of the stock-based lexical repository are associated with specific polarity and strength values;

a sentiment calculator of the computer applying rules based upon the association of the lexical items of the stock-based lexical repository with specific polarity and strength values to the filtered and processed social media messages so as to compute a representation of sentiment values associated with the social media messages as they relate to the publicly traded equities or commodities, wherein the rules are applied in head-complement, modifier-modified, and subject- predicate relations to pairs of the lexical items in their local syntactic context based upon the specific polarity and strength value of the lexical items; and

a computer monitor having a graphical user interface displaying the sentiment values as they relate to the publicly traded equities or commodities, wherein the publicly traded equities or commodities are represented as graphic objects which change in size and color to depict sentiment as a function of intensity and polarity.

2. The computer implemented system according to claim 1, wherein the ingest component acquires social media messages on the basis of keywords or a combination of keywords the ingest component has been programmed to look for.

3. The computer implemented system according to claim 1, further including a language identifier positioned after the ingest component for determining the language used in the social media messages prior to the filter module.

4. The computer implemented system according to claim 1, wherein the language processor includes a sentiment-based repository of frequent lexical items.

5. The computer implemented system according to claim 4, wherein the lexical items are associated with polarity values and strength values.

6. The computer implemented system according to claim 5, wherein the polarity values are positive, negative and neutral.

7. The computer implemented system according to claim 5, wherein the strength values associated with the lexical items range from at least 1 to 3, where 1 is the lowest value and 3 is the highest value.

8. The computer implemented system according to claim 4, wherein the lexical items of the stock-based lexical repository are associated with polarity values and strength values and wherein the lexical items of the sentiment-based lexical repository are associated with polarity values and strength values.

9. The computer implemented system according to claim 8, wherein the polarity values of the lexical items of the stock-based lexical repository are positive, negative and neutral and wherein the polarity values of the lexical items of the sentiment-based lexical repository are positive, negative and neutral.

10. The computer implemented system according to claim 8, wherein the strength values associated with the lexical items of the stock-based lexical repository ranges from at least 1 to 3, where 1 is the lowest value and 3 is the highest value and wherein the strength values associated to the lexical items of the sentiment-based lexical repository ranges from at least 1 to 3, where 1 is the lowest value and 3 is the highest value.

11. The computer implemented system according to claim 1, wherein the language processor includes a POS tagger.

12. The computer implemented system according to claim 1, wherein the polarity values are positive, negative and neutral.

13. The computer implemented system according to claim 1, wherein the language processor includes a parser.

14. The computer implemented system according to claim 1, wherein a sentiment is an integer computed based upon pairs of lexical items in local syntactic context.

15. The computer implemented system according to claim 1, further including an inference engine attempting to provide an answer to a problem, or clarifying uncertainties.

16. The computer implemented system according to claim 15, wherein the inference engine includes a data structure, a knowledge base that uses some knowledge representation structure to capture the knowledge of a specific domain, and a set of inference rules applying to the entities in the relational table and drawing consequences.

17. The computer implemented system according to claim 1, wherein the language processor includes a lemmatizer.

18. The computer implemented system according to claim 1, wherein the language processor includes a part of speech tagger assigning lexical categories to each of the filtered social media messages, a partial parser recovering the structure of syntactic structures of the filtered social media messages, and a lemmatizer reducing the spelling of words of the filtered social media messages to their lexical root.