437
in this section, we will provide a view of how different types of data affects the
computer science approach to be used in nexus research.
From a computer science perspective, data is a collection of objects that are associated with attributes and their values. These attributes can be nominal, ordinal,
interval or ratio. Nominal attributes represent different categories of data. For
example, the type of cars (e.g., hatchback, SUV, sedan) can be a nominal attribute.
Ordinal attribute refers to quantities with some order. For example, the grades of a
class can be ordinal. Interval attribute is like the ordinal attribute, but the intervals
are equally split. For example, the temperature readings from a sensor can have
interval attribute in Celsius or Fahrenheit scale. Finally, ratio attribute is an interval
data with a natural zero point. For example, temperature readings in Kelvin scale are
ratio datasets because they have an absolute zero value and the steps of each degree
have a same degree of magnitude. Those attributes depending on the application
domain can also be discrete or continuous; the former has only finite number or countably infinite set of values whereas the latter has real numbers as attribute values.
Datasets can also be divided into categories depending on their type (Srivastava
et al. 2005). Record datasets include data matrices, document data and transaction
datasets. Graph datasets, where each data item is considered as nodes and their
relationships are considered edges between nodes, include molecular structures,
social network datasets (e.g., friendship relation), or websites with the links
between each other. Finally, ordered datasets include spatial, temporal, and spatiotemporal data and sequential data.
Datasets that are used in computer science tasks comes from a variety of sources
including texts, videos, and photographs from social networks, web, and so on and
numerical data sets that are collected from sensors and/or generated by mathematical models. These collected datasets can be used in a variety of tasks in computer
science from data mining to machine learning and artificial intelligence. For example, a set of photos can be used to train a deep learning tool (Schmidhuber 2015),
that can further be later used to identify the objects in new input photos. Similarly,
using text mining from social network data streams (e.g., tweets) may help identify
the implicit connections between people. Since nexus datasets are often associated
with a location (e.g., longitude, latitude) as well as temporal information, the datasets that will be described in this section will only cover spatial and spatiotemporal
datasets. However, interested readers may read the referenced sources in this chapter to understand better the types of datasets that are used in other fields.
16.4.2 Spatial and Spatiotemporal Datasets
Spatial data represents the location, size, and shape of an object on planet Earth such
as a building, lake, mountain or township. Spatial data may also include attributes
that provide more information about the entity that is being represented. If those
datasets include temporal information associated with them, they are considered as
spatiotemporal datasets.
16 Computing
in this section, we will provide a view of how different types of data affects the
computer science approach to be used in nexus research.
From a computer science perspective, data is a collection of objects that are associated with attributes and their values. These attributes can be nominal, ordinal,
interval or ratio. Nominal attributes represent different categories of data. For
example, the type of cars (e.g., hatchback, SUV, sedan) can be a nominal attribute.
Ordinal attribute refers to quantities with some order. For example, the grades of a
class can be ordinal. Interval attribute is like the ordinal attribute, but the intervals
are equally split. For example, the temperature readings from a sensor can have
interval attribute in Celsius or Fahrenheit scale. Finally, ratio attribute is an interval
data with a natural zero point. For example, temperature readings in Kelvin scale are
ratio datasets because they have an absolute zero value and the steps of each degree
have a same degree of magnitude. Those attributes depending on the application
domain can also be discrete or continuous; the former has only finite number or countably infinite set of values whereas the latter has real numbers as attribute values.
Datasets can also be divided into categories depending on their type (Srivastava
et al. 2005). Record datasets include data matrices, document data and transaction
datasets. Graph datasets, where each data item is considered as nodes and their
relationships are considered edges between nodes, include molecular structures,
social network datasets (e.g., friendship relation), or websites with the links
between each other. Finally, ordered datasets include spatial, temporal, and spatiotemporal data and sequential data.
Datasets that are used in computer science tasks comes from a variety of sources
including texts, videos, and photographs from social networks, web, and so on and
numerical data sets that are collected from sensors and/or generated by mathematical models. These collected datasets can be used in a variety of tasks in computer
science from data mining to machine learning and artificial intelligence. For example, a set of photos can be used to train a deep learning tool (Schmidhuber 2015),
that can further be later used to identify the objects in new input photos. Similarly,
using text mining from social network data streams (e.g., tweets) may help identify
the implicit connections between people. Since nexus datasets are often associated
with a location (e.g., longitude, latitude) as well as temporal information, the datasets that will be described in this section will only cover spatial and spatiotemporal
datasets. However, interested readers may read the referenced sources in this chapter to understand better the types of datasets that are used in other fields.
16.4.2 Spatial and Spatiotemporal Datasets
Spatial data represents the location, size, and shape of an object on planet Earth such
as a building, lake, mountain or township. Spatial data may also include attributes
that provide more information about the entity that is being represented. If those
datasets include temporal information associated with them, they are considered as
spatiotemporal datasets.
16 Computing