Method for determining a sentiment from a text

Patent metadata

Field	Value
Publication number	US-9965443-B2
Application number	US-201214111101-A
Country	US
Kind code	B2
Filing date	Mar 12, 2012
Priority date	Apr 21, 2011
Publication date	May 8, 2018
Grant date	May 8, 2018

Abstract

A method for determining a sentiment, including determining, from a text including formatting information related to parts of the text, a sentiment expressed by at least one of the parts, wherein the sentiment is determined automatically using a microprocessor and depends on formatting information related to the at least one of the parts.

First claim

The invention claimed is:

1. A method for determining a sentiment, comprising:
receiving or accessing, using a microprocessor, a text;
processing the received or accessed text and determining, from the text including formatting information related to parts of the text, a sentiment expressed by at least one of the parts, wherein
the sentiment is determined automatically using the microprocessor and is determined based on formatting information related to the at least one of the parts,
the determining of the sentiment is based on an analysis of an order of sentences in the text,
the formatting information includes at least one of an underlining, an italic printing, a color, a font style, and/or a font size of characters,
for each of a plurality of the parts, a respective sentiment and a respective level of importance are determined by performing analysis of the text using the microprocessor,
a score is generated depending on values assigned to the respective sentiments of the plurality of the parts of the text, the score being generated by determining a weighted sum of the respective sentiments of the plurality of the parts, a weight of a respective sentiment being determined based on a respective level of importance and/or on a respective strength of the respective sentiment, and
the sentiment is determined depending on a cultural back-ground of an author of the text, by performing an analysis of particular information associated with the text including choice of words;
generating a visual indication associated with the score and the sentiment that is based on the formatting information, the order of sentences, and the cultural back-ground;
outputting the generated visual indication to a display; and
in response to receiving or accessing a plurality of texts, evaluating, for the plurality of texts, a respective sentiment with respect to a semantic content of at least one respective part of each of the texts by using statistical methods.

2. The method according to claim 1 , further comprising:
determining whether the sentiment corresponds to a positive or negative feeling of the author of the text with respect to a semantic content of the at least one of the parts of the text.

3. The method according to claim 1 , further comprising:
determining a strength of the sentiment.

4. The method according to claim 1 , further comprising:
determining a semantic content of the at least one of the parts of the text; and
evaluating the sentiment with respect to the semantic content.

5. The method according to claim 4 , wherein
the semantic content is related to a product or a feature of the product.

6. The method according to claim 1 , wherein
the formatting information includes at least one of a font type, a bold type, a paragraph alignment, a paragraph side margin, an itemization character, a punctuation character, an abbreviation for sentiment expression, a numbering, and/or a sequence of paragraphs used in the text.

7. A device for automated text evaluation, comprising:
circuitry configured to
receive or access from storage devices accessible via a network, texts related to a predetermined topic,
process the received or accessed texts to
determine parts of the texts including respective formatting information related to the parts of the texts,
determine, for each of the parts of the texts, a respective sentiment and a respective semantic content, wherein the determination of the respective sentiment is based on the respective formatting information, based on an analysis of an order of sentences in the respective part of the texts, and based on a cultural back-ground of an author of a respective text, the cultural back-ground being associated with performing an analysis of particular information associated with the respective text including choice of words, and

evaluate, for each of the parts of the texts, the respective semantic content with respect to the respective sentiment by using statistical methods, wherein

the formatting information includes at least one of an underlining, an italic printing, a color, a font style, and/or a font size of characters,
the circuitry is configured to generate a score depending on values assigned to the respective sentiments of the parts of the texts, the score being generated by determining a weighted sum of the respective sentiments of the parts of the texts, a weight of a respective sentiment being determined based on a respective level of importance and/or on a respective strength of the respective sentiment,
the circuitry is configured to generate a visual indication associated with the score and the sentiment that is based on the formatting information, the order of sentences, and the cultural back-ground, and
the circuitry is configured to output the generated visual indication to a display.

8. The device according to claim 7 , wherein
the predetermined topic is related to a product or a feature of the product.

9. The device according to claim 8 , wherein
the circuitry is configured to report a result of an evaluation to adapt a technical feature of the product in accordance with the result of the evaluation.

10. The device according to claim 8 , wherein
depending on a result of an evaluation, the circuitry is configured to report the result of the evaluation to a product development department to repair a malfunction of the product.

11. The device according to claim 7 , wherein
depending on a result of an evaluation, the circuitry is configured to report the result of the evaluation to adapt a product distribution and/or a supply chain.

12. The device according to claim 7 , wherein
a user profile of the author of the respective text is adapted in accordance with the result of the evaluation.

13. The device according to claim 12 , wherein
based on the adapted user profile, a recommendation is provided to the author.

14. The method according to claim 1 , wherein
the formatting information includes three or more of: a capitalization, the underlining, a font type, the font size, a bold type, an italic type, the color, a paragraph alignment, a paragraph side margin, an itemization character, an abbreviation for sentiment expression, a numbering, and/or a sequence of paragraphs used in the text.

15. The method according to claim 1 , wherein
the formatting information includes at least one of the underlining and the italic printing.

16. The method according to claim 1 , wherein
the formatting information includes three or more of: a capitalization, the underlining, a bold type, an italic type, a paragraph alignment, a paragraph side margin, an itemization character, an abbreviation for sentiment expression, a numbering, and/or a sequence of paragraphs used in the text.

Assignees

• Eggink Jana
• Kemp Thomas
• Schenk Niko
• Duran Daniel
• Sony Corp

Inventors

• Eggink Jana
• Kemp Thomas
• Schenk Niko
• Duran Daniel

Classifications

• G06F40/253 Primary

  Grammatical analysis; Style critique · CPC title

• G06F40/10 Primary

  Text processing (natural language analysis G06F40/20; semantic analysis G06F40/30; processing or translation of natural language G06F40/40) · CPC title

• G06F40/30

  Semantic analysis · CPC title

• G06F40/268

  Morphological analysis · CPC title

• G06F17/2785

  Physics · mapped topic