Goals and scope

Goals and scope

Galicia is intended as an integrated software platform including components for the key operations on lattices that might be required in practical applications or in more theoretically-oriented studies.

Thus, the basic configuration of the platform includes the following major functions: input of contexts, lattice construction and lattice visualization. Various auxiliary functions are also included in its architecture such as interactive data inputs, result export to various formats, etc. To sum up, \textsc{Galicia} should at best cover the whole range of tasks that make up the complete life-cycle of a lattice (see Figure 1).

Figure 1: lattice life-cycle

FCA and new lattice-based techniques

The intended impact of Galicia is two-fold since it should support both applications of FCA and development of new lattice-based techniques.

As a FCA-tool, Galicia offers an open architecture and generic implementations, which eases its adaptation to a particular application domain and problem settings. Moreover, Galicia presents a set of features, both functional and non-functional, that facilitate the analysis task.

For example, a wide range of data formats (context types) are allowed in the platform. Besides classical, i.e., binary, contexts, multi-valued contexts are admitted as well as more complex data descriptions such as the relational context families.

Flexibility

In addition, a rich set of algorithmic methods for lattice construction and maintenance is included in the system's architecture, so that the largest possible variety of analysis tasks could be covered by available services, while simplifying the implementation of uncovered tasks. Thus, the lattices and related structures can be constructed in both batch and incremental mode, while a set of maintenance methods allows small changes in the input data to be efficiently propagated to the already available analysis results. Consequently, the platform shows a high degree of flexibility upon user decisions on analysis parameters and therefore fits exploratory analysis tasks.

Moreover, the platform performs complex manipulations on lattices, such as assembly and slicing, which, even yet poorly studied, have a large potential for practical applications. For example, lattice assembly is suitable for the integration of the results from partial analyses whereas, lattice slicing models the decomposition of a complex global result into a set of smaller and therefore more intelligible parts.

Visualization

Finally, the platform offers highly sophisticated visualization mechanisms whose prime aim is to enable interpretation of analysis results. Except for the classical 2D graph drawing, Galicia features three-dimensional lattice diagrams and should soon provide a nested-line-diagram framework, a feature which extends the possibilities for lattice inspection and navigation.

In order to enhance the capacity of Galicia to support practical applications of FCA and Galois lattices, some frequently occurring lattice sub-structures are built in, together with the respective manipulation mechanisms. Thus, Galois sub-hierarchies, frequently used in re-engineering of object-oriented applications, and iceberg lattices, helpful for association rule mining, can be manipulated as first-class entities within the platform.