System and methods for predicting future trends of term taxonomies usage
A method and system for predicting future trends of terms taxonomies of users generated content are provided. The method includes crawling one or more sources of users generated content to collect phrases mentioned by users of the one or more data sources; periodically analyzing one or more term taxonomies to determine at least a trend of at least a non-sentiment phrase with respect of a plurality of sentiment phrases, wherein a term taxonomy is an association between a non-sentiment phrase and a sentiment phrase, the non-sentiment and sentiment phrases are included in the collected phrases; and generating a prediction of future behavior of the at least trend with respect of the one or more term taxonomies.
1. A system for predicting an expected volume of a future trend of term taxonomies of user generated content, comprising:
a network interface for accessing at least one data source to receive the user generated connect;
a processor; and
a memory, wherein the memory contains instructions that, when executed by the processor, configure the system to:
generate phrases from the user generated content, the phrases including sentiment phrases and non-sentiment phrases, wherein each of the sentiment phrases includes at least one word describing a sentiment;
periodically analyze at least one term taxonomy to determine at least one trend of at least one of the non-sentiment phrases with respect of a plurality of the respective sentiment phrases, wherein the term taxonomy is an association between the non-sentiment phrase and the sentiment phrases;
generate a prediction of a future behavior of the at least one trend with respect of the at least one term taxonomy, wherein the behavior of the trend is the number of mentions of phrases related to the trend and the proximity of the phrases from the trend;
monitor an actual performance of the at least one trend;
compare the actual behavior of the at least one trend and the prediction of future behavior of the at least one trend; and
generate a prediction grade respective of the comparison, wherein the prediction grade is indicative of the expected volume of the future trend.
2. The system of claim 1 , further comprises:
an output unit configured to provide real-time recommendations respective of the prediction grade.
3. The system of claim 1 , wherein the system is further configured to:
separate between variable behavior of the at least one trend and periodic behavior of the at least one trend respective of the prediction grade.
4. The system of claim 3 , wherein the system is further configured to:
define regions of interest related to the at least one trend respective of the periodic behavior of the at least one trend.
5. The system of claim 3 , wherein the system is further configured to:
determine a distribution of a volume of the at least one trend within the period behavior of the at least one trend.
6. The system of claim 1 , wherein the system is further configured to:
determine a context of a plurality of generated phrases.
7. The system of claim 6 , wherein the system is further configured to cluster a plurality of the generated phrases to one or more groups based on the context of the plurality of generated phrases.
8. The system of claim 6 , wherein the system is further configured to:
determine a probability that at least a first phrase and at least a second phrase out of the plurality of generated phrases appear together.
9. The system of claim 1 , wherein the sentiment is any one of: a positive sentiment, a neutral sentiment, and a negative sentiment.
10. The system of claim 1 , wherein the prediction grade is a delta volume between the expected volume and an actual volume of the determined trend.
11. A method for predicting an expected volume of a future trend of term taxonomies of user generated content, comprising:
accessing at least one data source to receive the user generated connect;
generating phrases from the user generated content, the phrases including sentiment phrases and non-sentiment phrases, wherein each of the sentiment phrases includes at least one word describing a sentiment;
periodically analyzing at least one term taxonomy to determine at least one trend of at least one of the non-sentiment phrases respective of a plurality of the respective sentiment phrases, wherein the term taxonomy is an association between the non-sentiment phrase and the sentiment phrases;
generating a prediction of a future behavior of the at least one trend with respect of the at least one term taxonomy, wherein the behavior of the trend is the number of mentions of phrases related to the trend and the proximity of the phrases from the trend;
monitoring an actual performance of the at least one trend;
comparing the actual behavior of the at least one trend and the prediction of future behavior of the at least one trend; and
generating a prediction grade respective of the comparison, wherein the prediction grade is indicative of the expected volume of the future trend.
12. The method of claim 11 , further comprising:
providing real-time recommendations respective of the prediction grade.
13. The method of claim 11 , further comprising:
separating between variable behavior of the at least one trend and periodic behavior of the at least one trend respective of the prediction grade.
14. The method of claim 13 , further comprising:
defining regions of interest related to the at least one trend respective of the periodic behavior of the at least one trend.
15. The method of claim 13 , further comprising:
determining a distribution of a volume of the at least one trend within the period behavior of the at least one trend.
16. The method of claim 11 , further comprising:
determining a context of a plurality of generated phrases.
17. The method of claim 16 , further comprising:
clustering a plurality of the generated phrases to one or more groups based on the context of the of the plurality of generated phrases.
18. The method of claim 16 , further comprising:
determining a probability that at least a first phrase and at least a second phrase out of the plurality of generated phrases appear together.
19. The method claim 18 , wherein the sentiment is any one of: a positive sentiment, a neutral sentiment, and a negative sentiment.
20. The method of claim 11 , wherein the prediction grade is a delta volume between the expected volume and an actual volume of the determined trend.
21. A non-transitory computer readable medium having stored thereon instructions for causing one or more processing units to execute the method according to claim 11 .