VisDB: Database Exploration Using Multidimensional Visualization

1
VisDB Database Exploration Using
Multidimensional Visualization

• Maithili Narasimha
• 4/24/2001

2
VisDB

• Scientific and Geographic databases tend to
  have large amounts of data.
• Some of the challenges in dealing with these
  databases are
• Mining these databases for useful information is
  a difficult task due to the sheer volume of data

3
VisDB

• Users do not know what they are looking for
  exactly.
• With traditional query specification languages,
  it is not possible to specify vague queries and
  thus not possible to get approximate results.
• There is no feedback. Result set may contain too
  few or too many points.

4
VisDB

• Requirements for a good Visualization System
  to explore large databases
• Flexible Query Specification
• Good Query Feedback
• Interactive system

5
VisDB

• Also, the users should be able to view as many
  data points as possible to see the patterns and
  clusters.
• Display size and resolution are limiting factors
• Also necessary to display the interdependencies
  between data attributes, Hotspots(anomalies).

6
VisDB

• The goal of the VisDB system is to address the
  tasks of visualization of the results , and that
  of incrementally refining the query to provide an
  effective way to find interesting data
  properties.

7
VisDB

• The approach.
• Use each pixel of the screen to visualize the
  results.
• Provide data items not only fulfilling the result
  exactly , but also those that match
  approximately.

8
VisDB

• Approximate results are determined by a relevance
  factor.
• The relevance factor of a data item is obtained
  by calculating distances for each selection
  predicate and combining them.
• The more the combined distance, the less the
  relevance of the data point.

9
Calculating the Relevance Factor

• Calculate the distance.
• Simple for Quantitative data.
• Nominal and Ordinal?
• Combining distances.
• User and Query dependent.
• Weighting Factor for each attribute.
• Normalizing.
• Arithmetic Mean for AND and Geometric mean for OR
  for combining different condition parts.
• Relevance Factor is the inverse of the Combined
  distance.

10
VisDB

• Basic Visualization Technique
• Sort the data points according to their
  relevance, with respect to the query.
• Assign colors depending on the relevance.
• Plot the sorted, colored points starting from the
  center of the screen moving outwards in a
  rectangular spiral fashion.

11
Overall Result Plotting
12
VisDB

• To relate the visualization of the overall result
  to the visualization of different selection
  predicates, separate windows for each selected
  predicate of the query are created and shown
  along with the result window.
• The position of the data items in all the other
  windows is determined by their position in the
  overall result window.

13
Arrangement of Windows for 5D Data
14
VisDB

• Mapping two dimensions to the axes
• It is possible for the user to assign two
  attributes to the axes and the system will
  arrange the relevance factors according to the
  directions of the distance of the data point from
  the selection predicate.
• With this method it is possible to provide better
  feedback to the user.
• However, we may not be able to use the display
  efficiently in some cases (I.e. some quadrants
  may not be used fully, while others are
  saturated)

15
2D Representation
16
VisDB

• Grouping the dimensions
• The pixels corresponding to the different
  dimensions of one data item are placed in one
  area instead of distributing them in different
  windows.
• Will require more pixels per dimension per data
  item.
• May provide more useful visualizations for data
  sets with larger dimensionality.

17
Grouping multi dimensional data
18
VisDB

• Interactive data exploration
• Users initially specify their queries, using some
  query language.
• Inside the VisDB interactive query and
  visualization interface, it is possible to view
  the visualizations and perform query
  modifications.
• System provides sliders for modifying selection
  predicates, weight factors and other options.

19
VisDB
20
VisDB

• Conclusion
• Useful for identifying and isolating clusters,
  correlations and hotspots in large databases.
• Good Query specification system.
• No Zoom or pan for the visualizations.