Www.monash.edu.au IMS5401 Web-based Systems Development Topic 3: Development for the web 3(c) Information Architecture.

www.monash.edu.au

IMS5401Web-based Systems Development

Topic 3: Development for the web

3(c) Information Architecture

www.monash.edu.au

2

Agenda

1. The nature of the problem

2. Organisation schemes for information

3. Organisation structures for information

4. Navigation

5. Labelling

6. Providing search capabilities

7. Developing a site architecture

8. Information professionals and web architecture

www.monash.edu.au

3

1. The nature of the problem

• Problems and needs• Importance• Context• Historical lessons

www.monash.edu.au

4

Information organisation problems: Information architecture

• Sub-dividing to fit page size• Sub-dividing diverse information according to

topic• Structuring information in a certain sequence• Linking to related information within the site• Labelling information so the audience can

quickly identify what is relevant to them• Finding an information layout which solves these

problems is the role of the Information Architect

www.monash.edu.au

5

Needs for site architecture

• To provide a rational framework or structure for the pages in which all site content will be stored

• To organise the framework in a way which helps users easily access the most important or frequently needed parts of the site

• To enable new content to be easily located within the site

• To lead inexperienced users through the site while enabling experienced users to get quickly to what they want

• To enable users to search the site

www.monash.edu.au

6

Importance

• A site URL takes the user only to the home page

• The content of the home page is a tiny fraction of the content of the site

• If you can’t find information in the rest of the site, then it is pointless

• Some comparisons:• A book• A library• A supermarket• A shopping centre

www.monash.edu.au

7

Context: A similar familiar problem

• How do you store information on the hard disk of your computer?• “I have everything in one folder - it means I know

where everything is”• “I put everything in its own separate folder - it means

nothing gets mixed up with the wrong things”• “I make a compromise between these extremes; I

have a number of folders and sub-folders, each of which contains collections of related things”

www.monash.edu.au

8

Your solution to the familiar problem

• How did you choose your system of organisation:• what folders you needed?

• what to call them?

• how to inter- relate them?

• what to put in them? etc

• How well does it work? Do you ever:• lose things?

• Put things in the wrong place?

• Be unsure where to put something?

• Take a lot of clicking to get to what you wanted?

• The test: could other people easily find documents on your computer?

www.monash.edu.au

9

Historical lessons

• Methods for organising, labelling, classifying and cataloguing information have been used in many disciplines throughout history

• These experiences have shown that:• You can’t get it ‘right’• You can get it less wrong

• We (ie in computing and the web) have been slow to learn from them

www.monash.edu.au

10

2. Organisation schemes for information

• Organisation scheme: how we classify items of content and how we combine them into groups of related items

• Simple examples:• Phone book• Supermarket• TV guide• Newspaper

• Organisation schemes are either:• Exact• Ambiguous• Hybrid

www.monash.edu.au

11

Exact organisation schemes

• Identifies groups or categories for all the information which are precisely-defined and mutually exclusive

• The group into which any item of content belongs is objective and easily identifiable

• Simple examples:• Alphabetical• Chronological• Geographical

• Very easy to create and use, but only work if the user can search in the same way the information is organised

www.monash.edu.au

12

Ambiguous organisation schemes

• Identifies groups or categories for all the information which are not exactly defined and are not mutually exclusive

• Deciding which group an item of content belongs is subjective and may be difficult

• Simple examples:• Topic-based• Function-based• Audience-based

• Enable broader and more useful groupings of information, but harder to manage; require judgement and understanding of users

www.monash.edu.au

13

Hybrid organisation schemes

• The items of content are divided into both exact and ambiguous groups

• A simple example:• Look at the groupings on the Monash or SIMS

home pages. Mixture of bases for grouping items

• Hybrid schemes can overcome the limitations of exact and ambiguous schemes, but at the risk of lack of consistency for the user; (how easy is the Monash home page first time around?)

• Help the user by making clear the basis for each type of group

www.monash.edu.au

14

3. Organisation structures for information

• The organisation structure sets the relationships between information:• Between items of content and • Between groups of items of content

• The organisation structure determines how users can navigate through the information - moving from one content item or group of items to another

www.monash.edu.au

15

The basic organisation structure - hierarchy

• Historically the dominant method; was initially considered old-fashioned in the world of hypertext, but is still the foundation of most systems

• Hierarchies are used to relate groups of items of content

• They may be clumsy and slow, but they are easy to understand and navigate

• Simple examples:• The Yahoo site directory • The organisation of information on your hard disk

www.monash.edu.au

16

Designing a hierarchy

• Narrow and deep?• Broad and shallow?• Implications for usability and content

classification• For sites with expanding content, tend towards

broad and shallow

www.monash.edu.au

17

Hypertext

• Non-linear structure which connects individual items of content to one another or to groups of content

• Flexible, can support any linkages you like (including hierarchy)

• Lack of a clear formal structure for information content tends to confuse the user

• Easy to get lost• Therefore, works best when used within a more

formal structure

www.monash.edu.au

18

Relational (database) model

• Link item groups to one another, using some key item as the point of connection

• Compare with Entity-Relationship diagrams• Entities = groups of items• Attributes = individual items• Relationships = connections

• Good for providing flexibility of content display for different audiences

• Suits only specific types of information - narrow tightly-defined data collections (eg product catalogues)

www.monash.edu.au

19

Creating an organisation system

• You can rarely use a single organisation scheme or organisation structure

• Note the benefits of each type and employ whatever blend best suits your site content and user needs

• May use different methods to suit different parts of a big site (sub-sites)

www.monash.edu.au

20

4. Navigation

• “Where am I?”• Types of navigation systems• Navigation structure• Placement of controls• Navigation aids

www.monash.edu.au

21

Where am I?

• Users get lost • Users may not enter your site via the home

page• Therefore:

• Every page should identify itself with the site (via a logo or organisation name)

• Every page should show the user where they are in the site

• Every page should have a link back to the home page

www.monash.edu.au

22

Types of navigation systems

• Global navigation - home; search; main topics or functions

• Local navigation - sub-topics or sub-sections; top and bottom of pages, etc

• Breadcrumb navigation - where have I come from?

• Contextual navigation - linked content: pages or sites

www.monash.edu.au

23

The navigation structure

• Controlled hierarchical (to fit structure) - rigid but safe and simple

• Go anywhere from anywhere - flexible but messy - too many navigation options needed on each page

• Compromise based on where the audience is likely to want to go - use standard user scenarios

www.monash.edu.au

24

Placement of navigation controls

• Top of page?• Bottom of page?• Sides of page?• Two of these?• Think about the size of the page, when the user

will want to move and place the controls accordingly

www.monash.edu.au

25

Other navigation aids

• Site maps• Site indexes• Pull-down menus• The guided tour• Personalised navigation paths

www.monash.edu.au

26

5. Labelling

• Labelling your navigation items may seem trivial, but it is important to get it right

• No matter how well-designed is your architecture, if it is poorly labelled, users won’t be any better off

• (Imagine a building with well-designed building layout, but no labels on the doors to tell you what is behind each one)

www.monash.edu.au

27

Labelling with words

• Be aware of web-standard terms and use them when appropriate (Home, Search, Site map, Contact us, About us, etc)

• What will your reader will understand by your labels?

• Be careful of jargon and acronyms which external users may not know (FAQ?)

• Don’t be too wordy - makes for a messy page

• Remember that users hate:• not knowing what something means• Going to it to find out only to find it is irrelevant to

them

www.monash.edu.au

28

Labelling with icons

• Be aware of web-standard icons and use them when appropriate

• Can be a useful graphical shorthand, but use sparingly - lots of room for mis-interpretation or bewilderment

• Check out www.iconbazaar.com to see how far iconography can go!

www.monash.edu.au

29

6. Providing search capabilities

• The need for a search engine• Getting a search engine• Understanding search engines

www.monash.edu.au

30

The need for a search engine

• How big is your site?• How easy is it to browse?• How well will your audience know it?• How much will your site grow and the content

change?

… And most importantly …

• Have you developed a good site architecture yet?!

www.monash.edu.au

31

Getting a search engine

• Buy one of the commercial ones• Get one for free off the web• Functionality and price will be driven by the

level of sophistication you need

www.monash.edu.au

32

Understanding search engine capabilities

• Remember the limitations and capabilities of search engine technology (se earlier lecture and references)

• Library catalogues and on-line databases can be useful but have limitations

• Search engines (without intelligent human cataloguing) are significantly more limited

• Search engines are no substitute for a good architecture

www.monash.edu.au

33

7. Methods for developing site architectures

• Top-down and bottom-up approaches• Personae• Card sorting methods• Cluster analysis• Search logs • Developing the architecture - initial and

revision?

www.monash.edu.au

34

Top-down and bottom-up approaches

• Top-down: Identify key topics and then break down into sub-categories and specific items of content

• Bottom-up: Identify specific items of content, then group and

• Approach depends on circumstances; maybe use a mixture of both?

www.monash.edu.au

35

Personae and scenarios

• Who are your ‘typical’ users? • Invent a persona for each • Role-play what they want and how they will

approach the site• Note differences and conflicts in their needs• Requires imagination and understanding of

what your users are like

www.monash.edu.au

36

Card sorting methods

• Commonly used method of eliciting user ideas on site architecture

• Use index cards with site topics, content items and user needs

• Ask users to sort and organise cards according to their needs and understanding

• Collate results for many users and use cluster analysis to identify patterns

www.monash.edu.au

37

Cluster analysis

• Common statistical technique for identifying patterns in data

• Varying levels of sophistication possible • Standard statistical packages can do it;

specialised web content analysis software now available

www.monash.edu.au

38

Search logs

• Search engines keep logs of user queries• (see www.monash.edu.au/servers/stats/infoseek for an

example)

• Analyse for most common searches• Re-configure site accordingly• Monash pathways analogy?

www.monash.edu.au

39

Parallels for selecting an architecture

• To what can constructing a web architecture be compared?

• Library catalogue analogy?• Database analogy?• Shopping analogy?• ??• Methods will develop and improve

www.monash.edu.au

40

Architecture development and revision

• Development - initial architectural design• Personae and scenarios• User focus groups

• Post-implementation - evaluation and re-design• Monitoring site usage• User reaction• User surveys

• Getting it ‘right’ - how likely and for how long?

www.monash.edu.au

41

8. Information professionals and web architecture

• What are the tasks in developing site architecture and navigation?

• What are the knowledge and skills required?• What can information professionals offer?

www.monash.edu.au

42

Tasks

• Classifying content• Deciding on an information structure• Developing profiles of user browsing

and search needs and behaviours• Developing a navigation structure• Deciding on a content and navigation

labelling system• Installing and maintaining a search

capability

www.monash.edu.au

43

Knowledge and skills

• Understanding the site content• Knowing how to classify and organise

information• Understanding the audience and their

needs• Knowing what works on the web• Understanding the psychology of

browsing and searching• Understanding the technology of

searching

www.monash.edu.au

44

What role should information professionals play?

• Knowledge of information (IM)• Knowledge of functionality (IS)• Knowledge of user behaviour (IS/IM)• Knowledge of search technology (IT)