Making Content Findable

1
Making Content Findable
gtgt
I N F O S E E K S O F T W A R E

• Andy Feit
• Vice President General Manager
• Infoseek Software

2
Todays Goals
gtgt
I N F O S E E K S O F T W A R E

• Discover ways to make your web-based or intranet
  content more findable by enterprise search
  engines
• Share some tools and techniques for improving
  navigation on your site / intranet
• What to tell your content authors, and ways to
  help them

3
Intranets are wonderful things
gtgt
I N F O S E E K S O F T W A R E

• Corporations typically use intranets for
• information sharing
• information publishing
• document management
• email, workflow, corporate directories, etc.

Source IDC doc 19643, July 1999
4
except when they arent.
gt
I N F O S E E K S O F T W A R E

• Slow access
• Difficulty accessing information
• Hard to find information
• Poor content
• Out of date
• Poor search engine

Source IDC doc 19643, July 1999
5
Frustrations add up...
gt
I N F O S E E K S O F T W A R E


Hard to find information
Source IDC doc 19643, July 1999
6
Frustrations add up...
gt
I N F O S E E K S O F T W A R E


Hard to find information
Too much information thats too hard to find
Information Chaos
Source IDC doc 19643, July 1999
7
Information Chaos How to solve it?
gtgt
I N F O S E E K S O F T W A R E

• Information management
• Publishing standards
• Content-discovery features

8
gt
I N F O S E E K S O F T W A R E
Information Chaos How to solve it?

• Information Management
• Content Management Packages
• Vignette StoryServer
• Allaire Spectra
• NetObjects Authoring Server

9
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Information Management
• Content Management Packages
• Vignette StoryServer
• Allaire Spectra
• NetObjects Authoring Server
• Publishing Workflow

10
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Publishing StandardsFindable Content
• Web-enabled
• Ability to access content from a browser

11
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Publishing StandardsFindable Content
• Web-enabled
• Clickable from other accessible pages
• Allows spiders (crawlers, robots, etc.) to find
  content by following links

12
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Publishing StandardsFindable Content
• Web-enabled
• Clickable from other accessible pages
• Have descriptive titles summaries
• Page titles should be unique and descriptive
• Page summaries should accurately describe page
  content

13
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Publishing StandardsFindable Content
• Web-enabled
• Clickable from other accessible pages
• Have descriptive titles summaries
• Use META tag information
• And a search engine that indexes them (ex.,
  search.state.mn.us)

14
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Publishing StandardsFindable Content
• Web-enabled
• Clickable from other accessible pages
• Have descriptive titles summaries
• Use META tag information
• Use XML
• Create your own tagging schemas for e-commerce,
  content, database records

15
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Content-discovery FeaturesNavigation Schemas
• Site Maps / Directories
• Automatically or manually created
• Forms-based
• Selectable from pull-downs
• Link-based
• Standard, hand-made or content-management
  system-based navigation

16
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Content-discovery FeaturesSearch Engine
  Software
• Full-text indexing
• Natural Language search syntax
• Simplified Boolean syntax
• Ability to natively index META data (and do field
  level search on it!)
• Ability to index XML optimally
• Good relevancy ranking

17
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E
Search Directory Software


Browseable, searchable index of all content
Full-text search Engine
Manageable topic hierarchy
18
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E
Architecture of an Enterprise Search Engine
Web Browser
Navigation User Interface
Search Results
Topic Listings
Document Index
TopicRules
Search Engine Spider
Web-based Content
19
Information Chaos How to solve it?
gt
I N F O S E E K S O F T W A R E

• Search Directory SoftwareChoosing the right
  one
• Administrative factors
• Scalability / Performance
• Control, Flexibility
• Admin UI and Ease of Management
• Good search performance, easy to use
• Real-Time Updates
• Leading edge, not bleeding edge

20
Indexing Challenges
gtgt
I N F O S E E K S O F T W A R E

• But spiders cant do everything
• Some pages include common (and not-so-common)
  features that make spidering and indexing a
  challenge

21
Indexing Challenges Dynamically Generated
Pages
gt
I N F O S E E K S O F T W A R E

• ASP, JSP, CGI, DB-driven, etc.
• Personalized or tracked pages
• Solution
• Make sure you use a search engine that can index
  these types of pages
• Create standard user profiles
• beware of cookies for personalization or
  tracking black holes

22
Indexing Challenges Image Maps
gt
I N F O S E E K S O F T W A R E

• Server-side image maps do not contain HREF links
• Solution
• Use client-side maps, which contain HREF links,
  or
• Place links from a server-side image map on a
  separate html page
• Example
• usgs.gov, clemson.edu

23
Indexing Challenges Javascript-created Pages
gt
I N F O S E E K S O F T W A R E

• HTML pages created by Javascript are hard to find
  by the spider
• Solution
• Include a ltNOSCRIPTgt section in the code,
  providing HREF links inside
• Example
• mylifepath.com

24
Indexing Challenges Javascript Menus Lists
gt
I N F O S E E K S O F T W A R E

• Spiders cant follow menus or lists created using
  Javascript
• Solution
• Include a ltNOSCRIPTgt section in the code,
  duplicating the menu or list items as links
• Example
• zinezone.com

25
Indexing Challenges Re-Directs
gt
I N F O S E E K S O F T W A R E

• Some indexes maintain the original URL as the
  target, not the actual page
• Solution
• Ensure that the actual target URL is listed on a
  site map or other page
• Make use of robots meta tag, no index, no
  follow

26
Indexing Challenges META Refresh
gt
I N F O S E E K S O F T W A R E

• Older browsers some spiders do not recognize
  META refresh URLs
• Solution
• Include a normal HREF link to the new page on the
  refresh page, or
• Ensure that the actual target URL is listed on a
  site map or other page

27
Indexing Challenges Frames Framesets
gt
I N F O S E E K S O F T W A R E

• Some spiders will not follow links contained
  within ltFRAMESETgt tags
• Solution
• Include links within a ltNOFRAMESgt tag section, or
• Include the links within a text listing or site
  map
• Examples
• saic.com, publish.com

28
Final Tips
gtgt
I N F O S E E K S O F T W A R E

• Provide GUIDELINES to your authors
• PICK a good enterprise search engine
• Give the spider good TRAILS to follow

29
For More Information
I N F O S E E K S O F T W A R E
To review this presentation and its information
sources, go to http//software.infoseek.com/what
snew.htm (and by the way, you can download a
really good search engine and topic manager,
FREE! )