IBM STAIRS
| Company type | Software |
|---|---|
| Industry | Information retrieval, Software |
| Predecessor | Internal IBM development (1969) |
| Founded | 1973 |
| Fate | Replaced by IBM SearchManager/370 and SearchManager/2 in 1991; phased out by 1992, support ended 1994 |
| Headquarters | Armonk , United States |
| Products | STAIRS, STAIRS/VS, STAIRS/CMS |
| Parent | IBM |
The IBM Storage and Information Retrieval System (STAIRS) was a software solution designed to facilitate storage and interactive free-text searching of textual information. Operating on the OS/360 platform, STAIRS utilized the CICS or IMS transaction processing systems and supported interaction through IBM 3270 display terminals.
History
First launched as a commercial product in 1973, STAIRS originated as an internal IBM development during the company's 1969 antitrust litigation. Initially called STAIRS, the product was renamed STAIRS/VS following the release of IBM's "/Virtual Storage" operating systems such as OS/VS1.
STAIRS initially functioned as an application running under IMS and CICS. Additionally, IBM Canada developed a version for VM/CMS in the late 1970s, which was primarily marketed across Europe under the name STAIRS/CMS.
In 1991, IBM replaced STAIRS with IBM SearchManager/370 and SearchManager/2. The original product was phased out by 1992, with official support concluding in 1994.
Description
The querying capabilities of STAIRS allowed users to construct boolean expressions using operators such as AND, OR, and NOT, complemented by modifiers like adjacent to or within the same paragraph as. Documents could incorporate formatted fields to hold structured data elements like dates or geographic locations.
Users had the option to save queries for later reuse or modification, enabling them to build upon prior search results by integrating further search criteria.
Search outputs were presented on 3270 terminals with specific highlighting to draw attention to matched terms.
Uses
STAIRS was primarily employed by large enterprises and governmental bodies managing vast collections of unstructured documents. It was also available through database vendors offering subscription-based access to their information retrieval services.
Technical details
STAIRS achieved efficient searching by indexing all words except for user-defined stop words, which typically consisted of common terms such as "and" or "the."
The system maintained two tiers of indexes: a dictionary containing unique entries for each word, and an inverted text index storing document identifiers along with positional information for every occurrence of each word. The original document content resided separately within a third file designated as the text file.
Database updates for STAIRS were performed offline. In versions other than CMS, data resided in basic direct access method (BDAM) files, which later introduced challenges related to upgrades and cross-platform portability.