This application takes a number of entities, for example keywords or companies, as the input
and transforms it into a
Pajek-network file. It can however be used for basically any type of network; words, literature, communication, inter-organizatinal...
It simply produces a Pajek network file from the input data.
The input should be formatted as follows: (see examples below)
Each row defines one set of related entities.
Each entity in the row is related to every other entity in the row.
The entities on a row should be separated by a comma!
Consider making the text either UPPER or lower case.
Example of input:
business relationships, connectedness, business networks
business relationships, adaptation, commitment, trust
business networks, connectedness, embeddedness, structure
Preparing the Input
To distinguish the nodes
To get rid of characters interfering with the name of the nodes
To prepare for formats that cannot handle such characters
or To avoid duplicates for caused by such differences
Filter the following nodes/words (case insensitive):
Replace certain nodes being the same (or words with the same meaning)
Case insensitivity of replace
(replacement: original1; original2; etc)
Embedded weighting (each node ends with a space and a number, e.g. node +5, missing equals 0)
Off Use sign (+/-) Add absolut numbers Multiply
Look for timestamps or indicator of order.
Place a timestamp of any format on a row of its own, spaces can be replaced with "_" and ":" with "#". The line must start with [TIME] and can end with [comment] (optional), e.g. [TIME]2007-03-31_23#59[First quarter].
Each timestamp will be applied for all consecutive lines, until a new timestamp is found.
If timestamp is on, use entire timestamp or just a part of it? (e.g. for reducing 2007-05-30_10#41 to 2007-05)
Use entire Use part start: length:
Type of connection between nodes
Connect all nodes in each line Connect each node with the