Artefact is a fulltext information retrieval system capable of searching through databases containing documents written in a natural language. Databases dealt with by Artefact can reside both on hard disks and on CD-ROMs, there being no signigicant difference in efficiency. The documents can be either unstructured or divided into fields. In the latter case, the layout of a document has to be specified by the user: each field must be preceeded by its name, a unique character sequence.
There are no restrictions on the format of unstructured documents: they may be arbitrary ASCII files containing texts written in Russian and/or English.
An Artefact database can contain large documents, each document being up to 1.5 Mb in size.
Artefact is primarily meant for processing texts written in Russian and English, irrespective of their subject matter.
The search algorithms used by Artefact provide context-sensitive search taking into account the structure of the document: its constituent fields, paragraphs, and sentences, as well as the word order within sentences.
Artefact allows the user to formulate sophisticated queries by combining words and phrases (to be searched for through the database) by means of search operators, which may specify some context restrictions on the occurrences of the words and phrases in terms of fields, paragraphs, sentences, and words.
Multibase search allows the user to find the required documents in up to 30 databases simultaneously.
While browsing through the documents that have been selected during the search, you may make Artefact show only those parts of the documents that match the query, rather than the whole documents. Those parts are highlighted (as well as the keywords appearing in them) by changes in colour.
The documents (or their marked fragments) selected during the search can be saved either in the search log or in separate files.
All the ideas, algorithms and programs (including the tools used for implementing the user interface) upon which the system is based are original.