BiQL Home page

Welcome to the BiQL web page.
This page contains information on the latest release of the Bison Query Language (BiQL).

Current version: v2.0.2


Release Notes

version 2.0.2 (14 December 2010)

version 2.0.1 (2 August 2010)

version 2.0 (25 July 2010)

version 1.1 (27 Oct 2009)

version 1.0 (20 Oct 2009)

Future Releases

We are working on the following aspects:

Documentation

This information is also available in the README file included with the latest release.

1. Installation

Unzip BiQL into a directory of your choice.

BiQL requires YAP Prolog, which is available for all common operating systems.

2. Starting BiQL

As of version 1.1, BiQL has a command line interface that runs on top of YAP Prolog.

Linux or Mac OSX: Open a terminal and start ./biql from within the BiQL installation directory. (This assumes that 'yap' can be found in the PATH.)

You should now see:

% Restoring file /usr/local/lib/Yap/startup.yss
YAP 6.0.5 (i386-darwin10.3.0): di 11 mei 2010 09:03:55 CEST
Welcome to ...
    _/_/_/    _/    _/_/      _/         
   _/    _/      _/    _/    _/          
  _/_/_/    _/  _/  _/_/    _/           
 _/    _/  _/  _/    _/    _/            
_/_/_/    _/    _/_/  _/  _/_/_/_/       
Type HELP for more information
| 

3. Using BiQL

Type HELP to get a list of available commands:

Available commands:
  HELP                  Show this message
  LOAD <filename>    Load a dataset
  LIST domains          List the domains present in the database
  SELECT ...            Run a query and show the result
  CREATE ...            Run a query and store the result in a new domain
  EXIT  or  CTRL-D      Exit the BiQL command line

An example data file is included containing a network of authors and publications. It can be loaded with

| LOAD example.biql

 % consulting example.biql...
 % consulted example.biql in module data, 566 msec 7972992 bytes
Loaded database"example.biql"
This network only contains two domains: nodes and edges

To create a more useful version one can create separate author, publication and author_of domains by typing the following queries:
CREATE author AS SELECT a FROM nodes a WHERE a.type = 'author'
CREATE publication AS SELECT a FROM nodes a WHERE a.type = 'publication'
CREATE author_of AS SELECT <a,p> { a->, p <-} FROM author a -> edges -> publication p
Now the queries used in our MLG paper can be run on this network. (Note however that some optimizations are missing from this version, and some of the queries might be slow.)

4. Contact Information

The BiQL implementation is maintained by Anton Dries of the Katholieke Universiteit Leuven.

5. References

A. Dries, L. De Raedt, S. Nijssen - A Query Language for Analyzing Networks - ACM Conference on Information and Knowledge Management (CIKM 2009).

A. Dries, S. Nijssen - Analyzing Graph Databases by Aggregate Queries - Workshop on Mining and Learning with Graphs (MLG 2010).