A few posts back, I blogged about InfoRapid Knowledge Portal and I promised I’d soon write about a related product. This builds very similar maps, but has many more functions and is not read-only as ‘Portal’ is.
So … I have just added InfoRapid KnowledgeBase Builder 2.0 (“Builder”) to Mind-Mapping.Org. The maps it makes are in almost the same visual form as my recent post about the browser-based ‘Portal’ but with this software, you’re in charge. This is a Windows desktop application, and you can choose the source or build a map from scratch … you can edit the map freely. Unlike Portal, Builder does not show the extracted text below the map, but like Portal, text extracted from the source appears as a pop-up when you hover the cursor over a node as you can see here:
Clicking on one of the small white arrows in the top left corner of each topic node opens a built-in browser window and loads that topic’s web page, if it is connected to one.
Maps from wikis
If you’ve ever tried to produce a mind map of an existing web site you’ll know that it is only easy for the very simplest of sites. And even for simple sites, concept map topology (see Concept maps or mind maps? the choice) is better than a tree-oriented mind map, because it allows any node to connect to any other node, just as any web page of a site can connect to any other on the site. An example of the kind of problem that turns up is the repetition resulting from navigation, terms and conditions, and privacy notice links on every page.
I’ll discuss its ability to crawl* a website later, but first the example above. It represents part of WikIT, the mind map wiki. That is built with MediaWiki software that includes a tool to make an XML dump of the wiki database. KB Builder can import this, and you can see (part of) the result.
Maps from data files
Other data formats that it can import are CSV (Comma-Separated Values), RDF (Resource Description Framework), OWL (Web Ontology Language [A A Milne – geddit?]), XMI (XML Metadata Interchange), XSD (XML Schema Definition) and GED (Gedcom File for Family Trees).
The User Interface
The four panels on each side of the map list the topics, the relationships, properties and an archive. The archive is an intermediate area that allows you to make attachments to topics. The line items in the panels allow the map to be manipulated in a way that Knowledge Portal does not – topics can be removed, added and modified. These panels can be folded into a small tab, if you wish, so that the map can use almost the whole window.
I tried its sitemap capability on my site topicscape.com, a mature and complex site that at first seems to result in a fairly confusing map with a lot of details off-screen (you’ll see I’ve folded the side panels away to make this image):
I felt that a zoom control was needed, but couldn’t find one on screen or in the menus, but it turned out that if you hold the Ctrl key down, the mouse wheel zooms the map. Then the overall shape of the site became clear with its several centers-of-connection.
The Ctrl+mousewheel trick works, because the map part of the window is rendered in a browser – Internet Explorer (it needs IE8 or later).
As I said, it’s challenging to produce a graphic sitemap with software, and I find this result impressive and useful.
Attempting to show all the connections in a website gets out of hand very quickly for reasons that I mentioned. If you’re doing it manually, you can use your judgement about what to exclude, but it’s more of a challenge with software. KnowledgeBase Builder makes the best fist of this that I’ve seen so far. It can follow and map the links on a website directly or it can import from one of a variety of data file types to build a map.
One clue to how it does this so well is a control in the View menu: “Show Cross-Connections”. Allow Builder to map out a good-sized web site, then select Show Cross-Connections. The center of a clear map, may well turn to something like the picture on the right. Plenty of connections are good in a web site, but such repetitions get confusing on a map of the site.
When starting to crawl a site, you get to specify number of web pages, number of link levels, links per page, timeout and other controls typical of web-crawling programs.
Builder places a thumbnail image of the web page under each node. Click on that, and you’ll be taken to a larger image. This can be useful if the page later disappears and you would like to refer to it. The size of thumbnail and image can also be specified before you initiate crawling. Builder adds a description on each connecting line, derived from the anchor text of the link it followed.
Building maps from scratch
So far I’ve talked mostly about Builder making maps automatically from a variety of data sources. You can modify these maps as you wish once they’re made but you can also start with a blank map, as with most mapping software.
For each topic that you add manually, you provide a name and can add a description. The name appears on the node, and the description appears when you hover over it. The description is similar to ‘Notes’ attachments in other mapping software, but is more closely integrated because it’s there by default, it is not necessary to click on the topic to see it, and it appears in a pop-up by the node, rather than in a window away from the node. The map adjusts dynamically to an elegant layout as you build it.
It also allows you to select any topic to make the central focus: Click a topic and everything rearranges around it, much like 3D Topicscape, Personal Brain or Cayra.
In true concept mapping style, you can add descriptive ‘linking phrases’ on relationship lines. You can also select colors, dotted or solid lines, and style of the ends, with or without arrows by selecting a relationship line and amending the items in the Properties pane.
You can connect two existing nodes by dragging one onto another in the diagram. If one is not in the visible area of the map, you can click on one of the nodes to be connected (to center it), find the other node to be connected in the Items pane, and drag it to the Relations pane, which at that point will be showing details of the selected, centered node.
Files, images and web pages can be attached to nodes. This is the first point at which I found a downside: The steps for doing this are, for example, select a file in Windows Explorer; Copy it (by Ctrl+C, say) … it appears in the ‘Archive’ pane, top right; then drag it from there to the Items pane top left, the Relations pane, bottom left or a node label in the map. Dragging directly from Explorer to the map would be so much easier and more in line with most other mapping software.
Making a map from scratch with Builder involves frequent movement between mouse and keyboard – it is not possible to move around the map with the keyboard alone – so this may not be the ideal brainstorming tool for fast idea capture.
There is a neat presentation mode, accessed from the view menu or toolbar. This hides the four side panels, and even the small tabs used to re-open them, so that the whole window is used for the map, menu bar and toolbar. A button appears in the toolbar for when you’re ready to revert to the editing view.
There is also a free read-only viewer so that people other than the originator of the map can explore it without being able to make changes.
Everything I’ve described so far is free for private, (i.e. non-commercial) use. There is a professional version for 99 Euros, and this permits multi-user editing of a map. There is also an Enterprise version allowing for unlimited users within the enterprise and embedding maps on a web site. For this you must request a quote, which I suppose will be based on the company size.
I feel this is a tool that has received too little attention and is well worth a look. Here it is in the Master List.
* A Web crawler is software that systematically browses the World Wide Web or a specified site, typically for the purpose of Web indexing or downloading a site. A Web crawler starts with a list of URLs to visit. As the crawler visits these URLs, it identifies all the hyperlinks in the page and adds them to the list of URLs to visit. URLs are recursively visited according to parameters that you specify to Builder. This process is also known as spidering. [Adapted from Wikipedia].