Glossary · 2 · Topics and information design
Bottom-up information architecture
Also known as: Bottom-up IA
German: Bottom-up-Informationsarchitektur
In technical communication, a bottom-up information architecture is one that emerges from self-contained pages, their metadata and the links between them, so that readers who arrive at any page through search or a link can orient themselves and move on along related subjects, rather than from a single hierarchy imposed from the top, such as the table of contents of a book.
- Information architecture
- Every Page is Page One
- Topics
In one sentence
A bottom-up information architecture emerges from self-contained pages, their metadata and links, not from one imposed hierarchy (Mark Baker).
Example
A service portal for a filling machine has no master outline. Each page on a fault names the machine, the station and the operating mode, and links to the procedure that clears the fault and to the reference data it needs; readers arriving from search start wherever the fault description led them.
How it applies
- Origin: Mark Baker sets out the idea in Every Page is Page One (XML Press, 2013), which contrasts “Information Architecture Top Down” with “Information Architecture Bottom Up” in two chapters. In his blog he summarized the argument as: there is no single “right place” for content that ensures readers find it; content has to be found by the reader’s search and work once the reader arrives. He names the web as the extreme case of a bottom-up architecture and a short booklet as content too small to be navigated that way.
- Requirements on the page: A bottom-up architecture only works when each page can stand on its own. This is why it is tied to Every page is page one: every topic establishes its own context and links along the relations a reader may follow.
- Means: Metadata and a taxonomy or faceted classification let pages be filtered and listed by product, audience or task; links connect pages by subject affinity; search ranks them. In DITA, relationship tables and subject scheme maps can supply links and controlled values; in iiRDS, metadata on each iiRDS Topic serves the same purpose for delivery.
- Machinery documentation: Content delivery portals and service apps that present instructions by product, component and situation apply the idea. The delivered instructions for use remain a complete, ordered information set where they have to be supplied as such; see Instructions for use (IFU).
Bottom-up vs. top-down information architecture
A top-down architecture defines a hierarchy first and places each piece of content in its “right place”, as in a book with a table of contents. A bottom-up architecture lets structure arise from the pages, their metadata and their links; lists, hubs and tables of contents can still exist, but as one way into the content among several. Baker’s argument concerns where the organization comes from, not the absence of organization.
External references
- Mark Baker, Every Page is Page One (XML Press, 2013), publisher page with the table of contents: https://xmlpress.net/publications/eppo/
- Mark Baker, “No ‘Right Place’ for Content” (blog post, March 31, 2014): https://everypageispageone.com/2014/03/31/no-right-place-for-content/
- Mark Baker, “Bottom-Up Information Architecture Behind the Firewall” (blog post, February 12, 2015): https://everypageispageone.com/2015/02/12/bottom-up-information-architecture-behind-the-firewall/