System and method for predicting user behaviors based on phrase connections
A method and system for generating alerts based on term taxonomies. The method includes analyzing textual content to detect a plurality of phrases in the textual content; identifying each of the detected phrases as any of: a sentiment phrase, and a non-sentiment phrase, wherein each sentiment phrase includes at least one word describing a sentiment; associating each identified non-sentiment phrase with at least one of the identified sentiment phrases to create at least one term taxonomy; analyzing the at least one term taxonomy to determine whether each identified non-sentiment phrase is associated with any of: a highly positive sentiment, and a highly negative sentiment; and generating an alert, when it is determined that one of the non-sentiment phrases is associated with any of: a highly positive sentiment, and a highly negative sentiment.
1. A method for generating alerts based on term taxonomies, comprising:
analyzing textual content to detect a plurality of phrases in the textual content;
identifying each of the detected phrases as any of: a sentiment phrase, and a non-sentiment phrase, wherein each sentiment phrase includes at least one word describing a sentiment;
associating each identified non-sentiment phrase with at least one of the identified sentiment phrases to create at least one term taxonomy;
analyzing the at least one term taxonomy to determine whether each identified non-sentiment phrase is associated with any of: a highly positive sentiment, and a highly negative sentiment; and
generating an alert, when it is determined that one of the non-sentiment phrases is associated with any of: a highly positive sentiment, and a highly negative sentiment.
2. The method of claim 1 , further comprising:
iteratively analyzing the at least one term taxonomy, wherein iteratively analyzing the at least one term taxonomy further comprises:
analyzing additional textual content to detect a plurality of other phrases in the additional textual content;
identifying at least one other sentiment phrase of the plurality of other phrases as any of: another sentiment phrase, and another non-sentiment phrase; and
updating, based on the identified at least one other sentiment phrase, the at least one term taxonomy, wherein the updated at least one term taxonomy is analyzed.
3. The method of claim 2 , further comprising:
generating a sentiment score for each term taxonomy;
comparing each generated sentiment score to at least one sentiment threshold; and
determining, based on the comparison, whether each term taxonomy is associated with any of: a highly positive sentiment, and a highly negative sentiment.
4. The method of claim 3 , wherein the at least one sentiment threshold includes at least one of: a highly positive sentiment threshold, and a highly negative sentiment threshold.
5. The method of claim 3 , wherein each sentiment phrase is associated with a sentiment value, wherein each sentiment score is an aggregated score generated based on the sentiment value associated with each sentiment phrase of the updated term taxonomy.
6. The method of claim 5 , further comprising:
ranking, via a tuning mechanism, the sentiment phrases relative to each other; and
determining, based on the ranking, the sentiment value of each sentiment phrase.
7. The method of claim 1 , wherein analyzing the textual content further comprises:
cleaning the textual content, wherein the cleaning includes at least one of: converting at least one of the detected phrases, and removing at least one of the detected phrases.
8. The method of claim 1 , further comprising:
identifying at least one connection between at least two of the detected phrases, wherein the at least one term taxonomy is created further based on the identified at least one connection.
9. The method of claim 8 , wherein each connection is any of: a direct connection, and a hidden connection.
10. A non-transitory computer readable medium having stored thereon instructions for causing one or more processing units to execute a method, the method comprising:
analyzing textual content to detect a plurality of phrases in the textual content;
identifying each of the detected phrases as any of: a sentiment phrase, and a non-sentiment phrase, wherein each sentiment phrase includes at least one word describing a sentiment;
associating each identified non-sentiment phrase with at least one of the identified sentiment phrases to create at least one term taxonomy;
analyzing the at least one term taxonomy to determine whether each identified non-sentiment phrase is associated with any of: a highly positive sentiment, and a highly negative sentiment; and
generating an alert, when it is determined that one of the non-sentiment phrases is associated with any of: a highly positive sentiment, and a highly negative sentiment.
11. A system for predicting user behaviors based on term taxonomies, comprising:
a processing circuitry; and
a memory, wherein the memory contains instructions that, when executed by the processing circuitry, configure the system to:
analyze textual content to detect a plurality of phrases in the textual content;
identify each of the detected phrases as any of: a sentiment phrase, and a non-sentiment phrase, wherein each sentiment phrase includes at least one word describing a sentiment;
associate each identified non-sentiment phrase with at least one of the identified sentiment phrases to create at least one term taxonomy;
analyze the at least one term taxonomy to determine whether each identified non-sentiment phrase is associated with any of: a highly positive sentiment, and a highly negative sentiment; and
generate an alert, when it is determined that one of the non-sentiment phrases is associated with any of: a highly positive sentiment, and a highly negative sentiment.
12. The system of claim 11 , wherein the system is further configured to:
iteratively analyze the at least one term taxonomy, wherein the system is configured to, at each iteration:
analyze additional textual content to detect a plurality of other phrases in the additional textual content;
identify at least one other sentiment phrase of the plurality of other phrases as any of: another sentiment phrase, and another non-sentiment phrase; and
update, based on the identified at least one other sentiment phrase, the at least one term taxonomy, wherein the updated at least one term taxonomy is analyzed.
13. The system of claim 12 , wherein the system is further configured to:
generate a sentiment score for each term taxonomy;
compare each generated sentiment score to at least one sentiment threshold; and
determine, based on the comparison, whether each term taxonomy is associated with any of: a highly positive sentiment, and a highly negative sentiment.
14. The system of claim 13 , wherein the at least one sentiment threshold includes at least one of: a highly positive sentiment threshold, and a highly negative sentiment threshold.
15. The system of claim 13 , wherein each sentiment phrase is associated with a sentiment value, wherein each sentiment score is an aggregated score generated based on the sentiment value associated with each sentiment phrase of the updated term taxonomy.
16. The system of claim 15 , wherein the system is further configured to:
rank, via a tuning mechanism, the sentiment phrases relative to each other; and
determine, based on the ranking, the sentiment value of each sentiment phrase.
17. The system of claim 11 , wherein the system is further configured to:
clean the textual content, wherein the cleaning includes at least one of: converting at least one of the detected phrases, and removing at least one of the detected phrases.
18. The system of claim 11 , wherein the system is further configured to:
identify at least one connection between at least two of the detected phrases, wherein the at least one term taxonomy is created further based on the identified at least one connection.
19. The system of claim 18 , wherein each connection is any of: a direct connection, and a hidden connection.