Document summarization using noun and sentence ranking
View Patent ↗Systems and methods are provided for summarization of electronic text documents. Nouns and sentences are identified in a text document, and the most-prevalent nouns are further identified based on frequency. The sentences in the document are scored based on assigning points based on cumulative presence or absence of each of the most-prevalent nouns. A tag cloud consisting of the most-prevalent nouns is displayed together with the highest-scoring sentences, thereby providing context for the nouns in the tag cloud.
1. A computerized method for providing a summary of a text document, comprising:
identifying at least some nouns and sentences in the text document;
for at least some of the identified nouns, counting the number of times the identified nouns appear in the text document;
identifying a predetermined number of most-prevalent nouns based on the number of times the identified nouns appear in the text document;
scoring each of the identified sentences as a function of the number of times the predetermined number of most-prevalent nouns occurs therein;
displaying at least some of the predetermined number of most-prevalent nouns, wherein the size of each displayed noun is a function of the number of times the displayed noun appears in the text document;
displaying a predetermined number of the scored sentences that have the highest scores, the displayed scored sentences being displayed in proximity to the displayed nouns; and
discarding identified sentences that contain fewer words than a minimum word threshold.
2. The computerized method of claim 1 , the scoring scheme further comprising:
for a given sentence, locating each of the identified nouns that are in the sentence;
for each located noun, identifying a score value to be added;
adding each of the score values for each time each located noun is in the sentence to the score for the sentence; and
dividing the score by the average number of words of all sentences in the text document, thereby normalizing the score over differing sentence lengths.
3. The computerized method of claim 1 , wherein three scored sentences and nine identified nouns are displayed in the displayed nouns and the displayed sentences.
4. The computerized method of claim 1 , wherein the minimum word threshold is six words.
5. The computerized method of claim 1 , further comprising sending the displayed nouns and the displayed sentences to a mobile Internet device.
6. The computerized method of claim 1 , further comprising displaying the displayed nouns in a tag cloud.
7. The computerized method of claim 1 , further comprising placing the most-prevalent noun in the center of a tag cloud.
8. The computerized method of claim 1 , further comprising using case-insensitivity and stemming to identify the nouns and sentences.
9. The computerized method of claim 1 , wherein the predetermined number of scored sentences to be displayed and the predetermined number of most-prevalent nouns to be displayed are configurable.
10. A computing system comprising:
a processor; and
a memory coupled to the processor and including computer-readable instructions that, when executed by the processor, cause the processor to:
identify at least some nouns and sentences in the text document;
for at least some of the identified nouns, count the number of times the identified nouns appear in the text document;
identify a predetermined number of most-prevalent nouns based on the number of times the identified nouns appear in the text document;
score each of the identified sentences as a function of the number of times the predetermined number of most-prevalent nouns occurs therein;
display at least some of the predetermined number of most-prevalent nouns, wherein the size of each displayed noun is a function of the number of times the displayed noun appears in the text document;
display a predetermined number of the scored sentences that have the highest scores, the displayed scored sentences being displayed in proximity to the displayed nouns; and
discard identified sentences that contain fewer words than a minimum word threshold.
11. The computing system of claim 10 , wherein three scored sentences and nine identified nouns are displayed in the displayed nouns and the displayed sentences.
12. The computing system of claim 10 , further comprising computer-readable instructions that, when executed by the processor, cause the processor to:
for a given sentence, locate each of the identified nouns that are in the sentence;
for each located noun, identify a score value to be added;
add each of the score values for each time each located noun is in the sentence to the score for the sentence; and
divide the score by the average number of words of all sentences in the text document, thereby normalizing the score over differing sentence lengths.
13. The computing system of claim 10 , wherein the predetermined number of scored sentences to be displayed and the predetermined number of most-prevalent nouns to be displayed are configurable.
14. The computing system of claim 10 , wherein the minimum word threshold is six words.
15. The computing system of claim 10 , further comprising a mobile Internet device for displaying the displayed nouns and the displayed sentences.
16. The computing system of claim 10 , further comprising computer-readable instructions that, when executed by the processor, cause the processor to display the displayed nouns in a tag cloud.
17. The computing system of claim 10 , further comprising computer-readable instructions that, when executed by the processor, cause the processor to display the most-prevalent noun in the center of a tag cloud.
18. The computing system of claim 10 , further comprising computer-readable instructions that, when executed by the processor, cause the processor to use case-insensitivity and stemming to identify the nouns and sentences.