Types of Attributes in

Datasets
A data object represents an entity. For example, in the university database, the objects
may be students, professors, etc. The rows of the database correspond to data objects,
the columns correspond to attributes as shown in the below figure. The type of
attribute is determined by the set of possible values the attribute can have.

There are five main types of data attributes which are discussed below one by one.
1. Nominal or Categorical attributes
Nominal attributes are attributes that are symbols or names of things. Each value
represents some kind of category, code, state, etc., and hence can also be referred to
as categorical. The attribute values do not have any meaning order between them. In
the example above, the attributes Name (represents student names), Course
(represents course category), and Gender (male or female) are Nominal attributes.
2. Binary attributes
A Binary Attribute is the same as a Nominal attribute but with only two
categories, for example, 0/1, True/False (Boolean), Pass/Fail, Male/Female,
Yes/No, positive/negative, etc. In our example, the attribute Gender is a
binary attribute as it only consists of two values: Male and Female.
A binary attribute is symmetric if both of its states are equally valuable and
carry the same weights. For example, in our case, the attribute Gender is
symmetric as both of its values (Male and Female) carry equal weight (meaning
none of the values is higher or lower importance than the other).

A binary attribute is asymmetric if both of its states are not equally important
and do not carry the same weight. For example, the outcome of an HIV test
(positive/negative) as a positive value is more concerning than a negative
value.

3. Ordinal attributes
An ordinal attribute is an attribute with possible values that have a meaningful
order or ranking among them. Ordinal attributes are qualitative. They describe
a feature of an object without giving an actual size or quantity. In our example,
the attribute Grades is an ordinal attribute as its values A+, A, B, and C have an
order.

Ordinal values can also be obtained by discretization of numeric quantities by

splitting the value range into a finite number of ordered categories. Let us
understand this with an example of the age of people which say is ranging from
13 to 100. This age attribute contains quantitative values which can be
converted to ordinal attribute as Teen (13–19), Adult (20–49), Elder (50–79),
and Old (80+). The values Teen, Adult, Elder, and Old have got an order
among them.

4. Numeric attributes
Numeric attributes are measurable quantities represented using integers or
real values. Further, this is classified as:
1. Interval-scaled: It has no true zero and can represent values below
zero. This allows us to compare and quantify the difference between the
values. It is measured on a scale of equal-sized units. Example:
 Temperature is interval-scaled as the difference between 20 and 21
degrees is the same as the difference between 200 and 201 degrees.
2. Ratio-scaled: This is a numeric attribute with an inherent zero-point i.e. if a
measurement is ratio-scaled, we can speak of a value as being a multiple of
another value. The value never falls below zero. example, the Height attribute is
a ratio-scaled numeric attribute as it is measured from 0 and above. Another
example can be the age of people.
5. Discrete and Continuous attributes
A discrete attribute is one that has a finite or countable set of values. It can be in
nominal or numeric form. In our example above, attributes Name, Course, Gender,
and Grades are discrete attributes.
A continuous attribute is one that has an infinite set of values. If an attribute is
not discrete it is continuous. They are typically represented as floating point values. In
the given example, attribute Height is a continuous attribute as it can take any value
in a given range; possibly the range being infinite. Another example can be the
temperature of a day.