Information Technology Reference
In-Depth Information
and retrieval over the distributed contents. There are different data models for
implementing NoSql.
have a simple data model based on key-
value pairs, which resembles an associative map or a dictionary. The key uniquely
identi
'
Key-value stores
'
es the data value and is used to store and retrieve it from the data store. The
data value can be of any type. In
the data are stored in a
column-oriented way. One example might be where the dataset consists of several
rows, each of which is addressed by a unique row key, also known as a primary
key. Each row is composed of a set of column families, and different rows can have
different column families. Similarly to key-value stores, the row key resembles the
key, and the set of column families resembles the value represented by the row key.
However, each column family further acts as a key for the one or more columns that
it holds, where each column consists of a name-value pair.
'
column-family stores
'
'
provide another derivative of the key-value store data model by using keys to locate
documents inside the data store. Each document can be highly heterogeneous and
so the store can provide the capability to index also on the document contents.
'
'
Document stores
originated from graph theory and use graphs as their data model.
By using a completely different data model to the other 3 types, graph databases can
ef
Graph databases
'
'
'
between different data nodes. Graph databases
are specialized in handling highly interconnected data and therefore are very effi-
ciently store the
relationships
-
cient in traversing relationships between different entities. NewSql is based more on
the relational model, where clients would interact in terms of table and relations. Its
internal data model however might be different and there can be semi-relational
models as well.
A navigational database is a type of database in which its records or objects are
found primarily by following references from other objects. Navigational interfaces
are usually procedural, though some modern systems like XPath ( 2014 ), can be
considered to be simultaneously navigational and declarative. Navigational dat-
abases therefore use a tree indexing system and can fall under the graph-based
category of NoSql. These graph-based databases therefore look more similar to a
concept base or concept tree. While the problems of semi-structured or unstructured
data remain, these new databases do offer general architectures and indexing sys-
tems. One criticism of graph-based ones however, is that they tend to lead to very
messy sets of indexing links that do not have very much structure. This is possible
for concept trees as well, but as the concept tree might have a more speci
c
construction process, it can provide some kind of mathematical foundation to help
with the organisation.
3.4 Semantic Environment
As well as a sensorised environment, a concept base is also closely related to the
Web 3.0, that is, the Semantic Web (Berners-Lee et al. 2001 ) combined with Service
Oriented Architectures (SOA) (OASIS 2014 ). This is because they can also produce
individual pieces of semantic information dynamically and computer-to-computer
Search WWH ::




Custom Search