232
Barbara Carminati and Elena Ferrari
able to translate the query //River//coord/X[contains(.,100)] into an encrypted query
that can be evaluated by the publisher directly on the SE-ENC document, that is,
//E k13 (River)//E k13 (coord)/E k13 (X)//Query-Info [contains(.,‘PF(100’)], where PF()
is the partitioning function which returns the partition id associated with value 100.
The publisher is thus able to evaluate the submitted query directly on the outsourced document (see Fig. 10.5), and returns to the user only those nodes answering it (that is, only the first E k13 (River) element). Only the users provided with the
proper key (i.e. k13 ) are able to decrypt and access the data. Since encryption keys
are distributed according to access control policies satisfied by users, this ensures
confidentiality with respect to the users.
It is important to note that the proposed framework also allows the evaluation of
more complex queries. For instance, if CambridgeCityModel stored the X,Y coordinates into attributes instead of elements, we could evaluate XPath predicates, whose
conditions are based on =, <, <=, >, >= operators. Thus, for instance, it could be
possible to perform queries like “Retrieve all rivers contained in a given rectangle.”
This can easily be evaluated by an XPath predicate on the attribute storing X coordinate (Y coordinate, respectively), which verifies that the coordinate is included
between the X coordinates of the rectangle.
10.6 Conclusions
The chapter dealt with a new and promising paradigm for data management, that is,
data outsourcing. The chapter, besides illustrating the basic concepts of this paradigm
and its possible applications on the GIS domain, focused on security issues arising
when data are outsourced to a third-party. Enforcing security requirements of both
final users and data owners is a primary need to make data outsourcing widely accepted. The chapter discussed main security requirements and analyzed the related
literature in view of these requirements. Then, it presented a comprehensive framework for secure outsourcing of XML data and illustrated its application to geographical data.
Data outsourcing is a new and emerging area, as such interesting research issues
still need to be addressed. For instance, a possible extension is considering privacy
as a further security requirement. This implies investigating several issues. Indeed,
besides the protection of user queries that is achieved in our framework by query
encryption, there is the need to protect user personal data as well as to consider concerns on possible data mining operations performed by publishers. Additionally, an
interesting issue is how techniques for intellectual property protection (see Chap. 11)
can be integrated into the proposed framework. Finally, the investigation of encryption strategies and related query processing for more complex queries, such as the
ones supported by Web feature service interface standard (WFS) [29], is a further
interesting research direction.
Barbara Carminati and Elena Ferrari
able to translate the query //River//coord/X[contains(.,100)] into an encrypted query
that can be evaluated by the publisher directly on the SE-ENC document, that is,
//E k13 (River)//E k13 (coord)/E k13 (X)//Query-Info [contains(.,‘PF(100’)], where PF()
is the partitioning function which returns the partition id associated with value 100.
The publisher is thus able to evaluate the submitted query directly on the outsourced document (see Fig. 10.5), and returns to the user only those nodes answering it (that is, only the first E k13 (River) element). Only the users provided with the
proper key (i.e. k13 ) are able to decrypt and access the data. Since encryption keys
are distributed according to access control policies satisfied by users, this ensures
confidentiality with respect to the users.
It is important to note that the proposed framework also allows the evaluation of
more complex queries. For instance, if CambridgeCityModel stored the X,Y coordinates into attributes instead of elements, we could evaluate XPath predicates, whose
conditions are based on =, <, <=, >, >= operators. Thus, for instance, it could be
possible to perform queries like “Retrieve all rivers contained in a given rectangle.”
This can easily be evaluated by an XPath predicate on the attribute storing X coordinate (Y coordinate, respectively), which verifies that the coordinate is included
between the X coordinates of the rectangle.
10.6 Conclusions
The chapter dealt with a new and promising paradigm for data management, that is,
data outsourcing. The chapter, besides illustrating the basic concepts of this paradigm
and its possible applications on the GIS domain, focused on security issues arising
when data are outsourced to a third-party. Enforcing security requirements of both
final users and data owners is a primary need to make data outsourcing widely accepted. The chapter discussed main security requirements and analyzed the related
literature in view of these requirements. Then, it presented a comprehensive framework for secure outsourcing of XML data and illustrated its application to geographical data.
Data outsourcing is a new and emerging area, as such interesting research issues
still need to be addressed. For instance, a possible extension is considering privacy
as a further security requirement. This implies investigating several issues. Indeed,
besides the protection of user queries that is achieved in our framework by query
encryption, there is the need to protect user personal data as well as to consider concerns on possible data mining operations performed by publishers. Additionally, an
interesting issue is how techniques for intellectual property protection (see Chap. 11)
can be integrated into the proposed framework. Finally, the investigation of encryption strategies and related query processing for more complex queries, such as the
ones supported by Web feature service interface standard (WFS) [29], is a further
interesting research direction.