Database Reference
In-Depth Information
Fig. 9.22 SNA visualization
We experimented (for more details, see [
38
]) the process using the data of a
database containing information about PhD students (administrative and technical
information) and also information about relevant personnel. This database contains
1,788 students (+ 80 students per year). To evaluate the scalability and performance
of our method for converting a relational database into a hypernode database, a set
of SQL queries has been designed to observe any differences between the source
relational database and the hypernode database. After comparing the results
between the two databases, the hypernode schema is generated without loss or
redundancy of data. This proves the correctness of this conversion.
We have used the same method to evaluate the transformation into a social
network. For each entity
e
:
, we verify that (1) the same attributes
appear in the relational database and in the social network, and (2) we find the same
relations. For example, in order to verify an entity's attributes:
Select * from h.name
where n
0
.name
¼<
h
,e
n
,N
e
>
n
0
.val
;(n
0
is the first node on the
e
's node set and is usually
corresponding to the primary key or a part of the primary key on the relational
database). For example, for the entity “
Student_Mohsen_Ali”,
the correspondent
query used:
Select * from Student where St_id
¼
03
;
The results obtained from these queries show the accuracy of the transformation
approach. Our approach can transform a relational database into a graph from a
social network perspective without the loss of information.
¼
9.6 Conclusion
In this chapter, we have presented the main graph database models and the
associated graph query languages. Graph database models can model communities
(social networks) and their activities, even though they are complex and dynamic
(using a model based on complex nodes). Moreover, graph query languages are the
most appropriate query languages for querying communities (e.g., better than RDF
Search WWH ::
Custom Search