Expanding network relationships

1. A method, said method comprising:

executing, by a processor of a data processing system, a similarity based clustering algorithm to generate clusters of user profiles, each cluster comprising a group of users of a plurality of users in an electronic community, each cluster representing a relationship between the users in each group that each cluster comprises, wherein relationships between the users in the clusters are non-static and change over time, and wherein interconnections or relationships in the clusters result in time-changing potential interactions among the users;

said processor displaying, on a display device, the relationships between the users in the clusters as hyperlinks that dynamically change over time on the display device in response to the relationships between the users in the clusters dynamically changing over time; and

said processor storing each cluster in a user profile and relationship database,

wherein the similarity based clustering algorithm comprises a member importance function and a member similarity function,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member importance function, an importance value of keywords as a depth of the keywords in a semantic hierarchical tree,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member similarity function, a similarity distance between keywords as a path distance between the keywords in the semantic hierarchical tree, and

wherein said executing the similarity based clustering algorithm comprises: using the member importance function and the member similarity function to ascertain the clusters.

2. The method of claim 1 , wherein the method further comprises:

said processor extracting noun phrases from activities of each user logged in an activity log server, each user having an existing user profile stored in the user profile and relationship database; and

prior to said executing the similarity based clustering algorithm, said processor updating the existing user profiles in the user profile and relationship database from the extracted noun phrases, wherein a keyword within the semantic hierarchical tree is associated with each determined noun phrase, and wherein said updating is based on a usage frequency of the extracted noun phrases and an importance value of the keywords.

3. The method of claim 2 , wherein said updating comprises:

using the member importance function to ascertain a first importance value as a first depth in the semantic hierarchical tree of a first keyword associated with a first noun phrase of the extracted noun phrases;

ascertaining, via use of the member importance function, a second importance value as a second depth in the semantic hierarchical tree of a second keyword whose depth in the semantic hierarchical tree exceeds the first depth and whose meaning is more specific and descriptive than is the meaning of the first keyword; and

in response to ascertaining that the second depth exceeds the first depth, replacing the first noun phrase in a first user profile of the user profiles by a second noun phrase to which the second keyword is associated.

4. The method of claim 2 ,

wherein a digital hierarchical dictionary comprises synsets, each synset being a set of cognitive synonyms consisting of noun phrases, said synsets being interlinked into the semantic hierarchical tree within the digital hierarchical dictionary,

wherein a first noun phrase of the extracted noun phrases is a first name not found in the digital hierarchical dictionary, and wherein a dataset consisting of a text file or a database table comprises mappings of names to respective noun phrases, wherein the mappings comprise a first mapping of the first name to a respective second noun phrase, and

wherein the method further comprises: said processor ascertaining the second noun phrase from the first mapping.

5. The method of claim 1 , wherein the method further comprises said processor calculating a relationship importance value of the relationship represented by a first cluster of the clusters by:

using the member importance function to ascertain importance values of keywords of the first cluster; and

calculating the importance value of the relationship represented by the first cluster as a sum of the ascertained importance values of the keywords of the first cluster.

6. The method of claim 1 , wherein the similarity based clustering algorithm uses a similarity threshold to constrain a total number of clusters returned from said executing the similarity based clustering algorithm, and wherein said executing the similarity based clustering algorithm comprises:

increasing the similarity threshold to decrease the total number of clusters returned from said executing the similarity based clustering algorithm; or

decreasing the similarity threshold to increase the total number of clusters returned from said executing the similarity based clustering algorithm.

7. A data processing system comprising a processor, a memory coupled to the processor, and a computer readable storage device coupled to the processor, said storage device containing program code configured to be executed by the processor via the memory to implement a method, said method comprising:

said processor executing a similarity based clustering algorithm to generate clusters of user profiles, each cluster comprising a group of users of a plurality of users in an electronic community, each cluster representing a relationship between the users in each group that each cluster comprises, wherein relationships between the users in the clusters are non-static and change over time, and wherein interconnections or relationships in the clusters result in time-changing potential interactions among the users;

said processor displaying, on a display device, the relationships between the users in the clusters as hyperlinks that dynamically change over time on the display device in response to the relationships between the users in the clusters dynamically changing over time; and

said processor storing each cluster in a user profile and relationship database,

wherein the similarity based clustering algorithm comprises a member importance function and a member similarity function,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member importance function, an importance value of keywords as a depth of the keywords in a semantic hierarchical tree,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member similarity function, a similarity distance between keywords as a path distance between the keywords in the semantic hierarchical tree, and

wherein said executing the similarity based clustering algorithm comprises: using the member importance function and the member similarity function to ascertain the clusters.

8. The system of claim 7 , wherein the method further comprises:

said processor extracting noun phrases from activities of each user logged in an activity log server, each user having an existing user profile stored in the user profile and relationship database; and

prior to said executing the similarity based clustering algorithm, said processor updating the existing user profiles in the user profile and relationship database from the extracted noun phrases, wherein a keyword within the semantic hierarchical tree is associated with each determined noun phrase, and wherein said updating is based on a usage frequency of the extracted noun phrases and an importance value of the keywords.

9. The system of claim 8 , wherein said updating comprises:

using the member importance function to ascertain a first importance value as a first depth in the semantic hierarchical tree of a first keyword associated with a first noun phrase of the extracted noun phrases;

ascertaining, via use of the member importance function, a second importance value as a second depth in the semantic hierarchical tree of a second keyword whose depth in the semantic hierarchical tree exceeds the first depth and whose meaning is more specific and descriptive than is the meaning of the first keyword; and

in response to ascertaining that the second depth exceeds the first depth, replacing the first noun phrase in a first user profile of the user profiles by a second noun phrase to which the second keyword is associated.

10. The system of claim 8 ,

wherein a digital hierarchical dictionary comprises synsets, each synset being a set of cognitive synonyms consisting of noun phrases, said synsets being interlinked into the semantic hierarchical tree within the digital hierarchical dictionary,

wherein a first noun phrase of the extracted noun phrases is a first name not found in the digital hierarchical dictionary, and wherein a dataset consisting of a text file or a database table comprises mappings of names to respective noun phrases, wherein the mappings comprise a first mapping of the first name to a respective second noun phrase, and

wherein the method further comprises: said processor ascertaining the second noun phrase from the first mapping.

11. The system of claim 7 , wherein the method further comprises said processor calculating a relationship importance value of the relationship represented by a first cluster of the clusters by:

using the member importance function to ascertain importance values of keywords of the first cluster; and

calculating the importance value of the relationship represented by the first cluster as a sum of the ascertained importance values of the keywords of the first cluster.

12. The system of claim 7 , wherein the similarity based clustering algorithm uses a similarity threshold to constrain a total number of clusters returned from said executing the similarity based clustering algorithm, and wherein said executing the similarity based clustering algorithm comprises:

increasing the similarity threshold to decrease the total number of clusters returned from said executing the similarity based clustering algorithm; or

decreasing the similarity threshold to increase the total number of clusters returned from said executing the similarity based clustering algorithm.

13. A data processor readable medium, said medium being hardware and not being a transmission medium, said medium comprising program code stored therein, said medium not being a transitory signal, said program code configured to be executed by a processor of a data processing system to perform a method, said method comprising:

said processor executing a similarity based clustering algorithm to generate clusters of user profiles, each cluster comprising a group of users of a plurality of users in an electronic community, each cluster representing a relationship between the users in each group that each cluster comprises, wherein relationships between the users in the clusters are non-static and change over time, and wherein interconnections or relationships in the clusters result in time-changing potential interactions among the users;

said processor displaying, on a display device, the relationships between the users in the clusters as hyperlinks that dynamically change over time on the display device in response to the relationships between the users in the clusters dynamically changing over time; and

said processor storing each cluster in a user profile and relationship database,

wherein the similarity based clustering algorithm comprises a member importance function and a member similarity function,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member importance function, an importance value of keywords as a depth of the keywords in a semantic hierarchical tree,

wherein said executing the similarity based clustering algorithm comprises: ascertaining, by the member similarity function, a similarity distance between keywords as a path distance between the keywords in the semantic hierarchical tree, and

wherein said executing the similarity based clustering algorithm comprises: using the member importance function and the member similarity function to ascertain the clusters.

14. The medium of claim 13 , wherein the method further comprises:

said processor extracting noun phrases from activities of each user logged in an activity log server, each user having an existing user profile stored in the user profile and relationship database; and

prior to said executing the similarity based clustering algorithm, said processor updating the existing user profiles in the user profile and relationship database from the extracted noun phrases, wherein a keyword within the semantic hierarchical tree is associated with each determined noun phrase, and wherein said updating is based on a usage frequency of the extracted noun phrases and an importance value of the keywords.

15. The medium of claim 14 , wherein said updating comprises:

using the member importance function to ascertain a first importance value as a first depth in the semantic hierarchical tree of a first keyword associated with a first noun phrase of the extracted noun phrases;

ascertaining, via use of the member importance function, a second importance value as a second depth in the semantic hierarchical tree of a second keyword whose depth in the semantic hierarchical tree exceeds the first depth and whose meaning is more specific and descriptive than is the meaning of the first keyword; and

in response to ascertaining that the second depth exceeds the first depth, replacing the first noun phrase in a first user profile of the user profiles by a second noun phrase to which the second keyword is associated.

16. The medium of claim 14 ,

wherein a digital hierarchical dictionary comprises synsets, each synset being a set of cognitive synonyms consisting of noun phrases, said synsets being interlinked into the semantic hierarchical tree within the digital hierarchical dictionary,

wherein a first noun phrase of the extracted noun phrases is a first name not found in the digital hierarchical dictionary, and wherein a dataset consisting of a text file or a database table comprises mappings of names to respective noun phrases, wherein the mappings comprise a first mapping of the first name to a respective second noun phrase, and

wherein the method further comprises: said processor ascertaining the second noun phrase from the first mapping.

17. The medium of claim 13 , wherein the method further comprises said processor calculating a relationship importance value of the relationship represented by a first cluster of the clusters by:

using the member importance function to ascertain importance values of keywords of the first cluster; and

calculating the importance value of the relationship represented by the first cluster as a sum of the ascertained importance values of the keywords of the first cluster.

18. The medium of claim 13 , wherein the similarity based clustering algorithm uses a similarity threshold to constrain a total number of clusters returned from said executing the similarity based clustering algorithm, and wherein said executing the similarity based clustering algorithm comprises:

increasing the similarity threshold to decrease the total number of clusters returned from said executing the similarity based clustering algorithm; or

decreasing the similarity threshold to increase the total number of clusters returned from said executing the similarity based clustering algorithm.