DETERMINING THEMES

1 . (canceled)

2 . A system, comprising:

a processor configured to:

receive reputation data extracted from at least one data source, wherein the reputation data includes a plurality of user-authored reviews;

determine that a first item is present in a first review in the plurality of user-authored reviews;

assign to the first item an identifier that is generated based at least in part on a string from which the first item was derived;

in response to determining that an identifier assigned to a second item is the same as the identifier assigned to the first item, count the first item and the second item as a single item;

determine that a third item, different from but associated with the first item, is present in a second review in the plurality of user-authored reviews;

determine a sentiment for a theme based on the presence of the first and third items; and

provide as output a report that indicates the sentiment for the theme; and

a memory coupled to the processor and configured to provide the processor with instructions.

3 . The system of claim 2 , wherein the identifier for the first item is generated based at least in part on a sentence from which the first item was derived.

4 . The system of claim 2 , wherein the identifier for the first item is generated based at least in part on a clause.

5 . The system of claim 2 , wherein the identifier for the first item is generated at least in part by performing a hash of a string.

6 . The system of claim 5 , wherein determining that the identifier assigned to the second item is the same as the identifier assigned to the first item comprises comparing one or more hash codes.

7 . The system of claim 2 , wherein determining that the first item is present includes performing natural language processing.

8 . The system of claim 2 , wherein the first and second reviews are reviews of an entity.

9 . The system of claim 8 , wherein the processor is further configured to determine whether the first item is present in an ontology associated with the entity.

10 . The system of claim 2 , wherein the processor is further configured to select an ontology associated with an entity being reviewed in the first review.

11 . The system of claim 2 , wherein the processor is further configured to select, from a plurality of ontologies, an ontology that is associated with an industry associated with an entity being reviewed in the first review.

12 . A method, comprising:

receiving reputation data extracted from at least one data source, wherein the reputation data includes a plurality of user-authored reviews;

determining that a first item is present in a first review in the plurality of user-authored reviews;

assigning to the first item an identifier that is generated based at least in part on a string from which the first item was derived;

in response to determining that an identifier assigned to a second item is the same as the identifier assigned to the first item, counting the first item and the second item as a single item;

determining that a third item, different from but associated with the first item, is present in a second review in the plurality of user-authored reviews;

determining a sentiment for a theme based on the presence of the first and third items; and

providing as output a report that indicates the sentiment for the theme.

13 . The method of claim 12 , wherein the identifier for the first item is generated based at least in part on a sentence from which the first item was derived.

14 . The method of claim 12 , wherein the identifier for the first item is generated based at least in part on a clause.

15 . The method of claim 12 , wherein the identifier for the first item is generated at least in part by performing a hash of a string.

16 . The method of claim 15 , wherein determining that the identifier assigned to the second item is the same as the identifier assigned to the first item comprises comparing one or more hash codes.

17 . The method of claim 12 , wherein determining that the first item is present includes performing natural language processing.

18 . The method of claim 12 , wherein the first and second reviews are reviews of an entity.

19 . The method of claim 18 , further comprising determining whether the first item is present in an ontology associated with the entity.

20 . The method of claim 12 , further comprising selecting an ontology associated with an entity being reviewed in the first review.

21 . The method of claim 12 , further comprising selecting, from a plurality of ontologies, an ontology that is associated with an industry associated with an entity being reviewed in the first review.

22 . A computer program product embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

receiving reputation data extracted from at least one data source, wherein the reputation data includes a plurality of user-authored reviews;

determining that a first item is present in a first review in the plurality of user-authored reviews;

assigning to the first item an identifier that is generated based at least in part on a string from which the first item was derived;

in response to determining that an identifier assigned to a second item is the same as the identifier assigned to the first item, counting the first item and the second item as a single item;

determining that a third item, different from but associated with the first item, is present in a second review in the plurality of user-authored reviews;

determining a sentiment for a theme based on the presence of the first and third items; and

providing as output a report that indicates the sentiment for the theme.