Read through our latest webinar held in conjunction with Google Enterprise all about Enterprise Best Practices and creating successful search applications using the Google Search Appliance 7.0. Search Technologies provides implementation and consulting services to Google search Appliance Customers. For further information, see http://www.searchtechnologies.com/google-search-appliance-services.html
http://searchtechnologies.com
Deploy with confidence: VMware Cloud Foundation 5.1 on next gen Dell PowerEdg...
Enterprise Search Best Practices Webinar 4.2013
1. ENTERPRISE SEARCH BEST PRACTICES
Iain Fletcher
ifletcher@searchtechnologies.com
Bill Fowler
bfowler@searchtechnologies.com
Greg Gomez
ggomez@google.com
Maria Lundahl
mlundahl@google.com
2. Agenda
• Search Technologies Overview
• Enterprise Search Best Practices
• Three pillars of enterprise search success
• Google Demo Presentation
• Q&A
2
3. Search Technologies Overview
• The largest IT services company focused on search
engines
• Consulting
• Implementation
• Managed services
• Technology independent, working with most of the leading
search engine vendors
• An increasing proportion of customers using the GSA
3
6. Introduction
• The GSA is used for a wide range of search solutions
• Many initial GSA implementations address “plug-in-and-go”
• We are working with an increasing number of GSA
customers to extend to enterprise-scale
6
7. Enterprise Search Applications
Better search improves a
wide range of business processes
7
• Compliance & Risk
• Customer Service Support
• Research & Development
• Legal & Contracts
• Competitive Intelligence
• People/Expert Search
• General Intranet…
Internal
• Partner Extranet
• Online Publishing
• Ecommerce
• Customer Self-service
• Field Maintenance
• Website Search
• Tech Sales Support…
External
8. The Platform Approach
Key Advantages
• Cost sharing provides lower TCO
• Agility is improved, as it is much easier to create new
applications or repurpose content via search, to meet
emerging needs
• Improved user productivity
• Users benefit from a consistent and instantly available, repository
independent search experience
8
10. Relevancy & Scalability
• Relevancy is easy over small data sets
• Most modern search engines scale well in terms of
document count and query load
• Maintaining relevancy at scale is the challenge
• Poor relevancy is a common complaint with existing enterprise
search systems
• This is a sweet spot for the GSA
10
11. Connectivity & Security
• What is the #1 reason to not find information?
• Because it hasn’t been indexed
• Data connectors enable the search engine to access
content sources
• A growing range of connectors are available for the GSA
• There is also a connector framework for custom developments
• Every company has its own combination of repositories
11
12. Connectivity & Security
• Connectivity occasionally creates “issues”
• These are mostly caused by the characteristics of the repository or
the general IT environment, rather than by faults with the connector
software
• Security causes the most issues
• In larger organizations, security can be complex
• Multiple LDAP/AD servers
• Nested permissions, etc.
12
13. Connectivity & Security
Best Practices:
• Don’t expect connectors to always “plug-and-go”
• Usually they will, but sometimes it will take a little effort
• Secure connectivity can almost always be achieved
• Take security seriously
• Make friends not enemies of repository owners by
showing them that you take security seriously – you
will need their cooperation!
13
14. Metadata Capture & Creation
Metadata is the foundation of important search functions.
For example:
• It supports relevancy
• It drives dynamic navigation
• It enables infographical results display
14
17. Creating & Capturing Metadata
Capture
• Be diligent about taking what is available from the source
• This includes metadata encoded into file paths, or
available from complimentary sources
Creation
• Automate the creation of metadata based on
extraction techniques
• The GSA 7 provides new capabilities for this
“Entity Recognition”
17
18. New with GSA 7: Entity Recognition
• Provides both “regex” and dictionary-based approaches to
automated metadata creation
• Regex – identifies patterns in the content to match names, emails,
phone numbers, etc.
• List-based, enables a focus on industry or company specific
terminology
• Open & Customizable
18
19. Summary
• The GSA 7 has a full set of capabilities to deliver enterprise
search excellence
• Key issues are:
• Relevancy at scale
• Security-compliant connectivity capabilities
• Metadata capture and auto-generation capabilities
• Plus the administrative simplicity you’d expect from a
Google application
19
20. The Google Approach
• Google receives >1B queries per
day
• Algorithm is made up of >200
signals
• >500 changes to the algorithm
each year
• Each change is analyzed through
1% testing of actual Google traffic
• Over 10,000 such experiments
each year
Signals
Signals
21. Evolution of the Google Search Appliance
2002 2009 2012
6.
1
4
Apr Oct Dec
2011
7.
0
Sep
2004
6.10 6.12
2006 2013
50X Increase in Capacity between 2002 and 2013
2M to 100M documents
We’re an IT services company dedicated to search engines Search is all we do We believe that we are the largest company of our kindWe provide strategic consulting, implementation work, and a range of managed servicesWe are technology independent and work with all of the leading search engine vendors.Established in 2005, we now have about 110 staff and more than 400 customers. A growing proportion of our customers now use the Google Search Appliance.
We have worked with more than 130 customers who have chosen to use the GSA, here are a few of them
Why not stick with the embedded search functionality that most repositories provide?READ SLIDEIn the case of the GSA, it will look familiar too…
Scale within the enterprise also implies a wide variety of document types, formats and lengths, which further complications relevancy calculation.I’ll leave it to Google to mention the technical aspects of achieving relevance at scale during their presentation.
The number 1 reason for not finding something through using search is simple – because it has not been indexed.To index content, of course you need data connectors.In the GSA environment, there’s a growing range of connectors available, both from Google and from third parties. There’s also an API for developing new connectorsEvery company has its own combination of repositories to connect to
Connectors almost always need configuration. Even where they will literally plug-and-play, and with the with GSA this is often the case, it is still worth taking the time to understand the full range of configuration possibilities. Connectors frequently have issues.We’ve seen many companies over the years who have approached enterprise search with strong expectations of plug-and-play when it comes to connectors, only to be discouraged when things don’t run entirely smoothly.When issues occur, it is usually not because of the connector software. Most issues are caused by the characteristics of the repository, or by the IT environment more generally.The enterprise search implementation team typically has no control over the configuration of document repositories, so you have to “go with the flow” and work with what is available. Security causes the most connector issues, I’ll come back to that shortly.But the good news is that with planning, and with a willingness to configure or customize, pretty much any repository can be successfully connected to, and this gets us to base camp, getting the content indexed and made searchable
The point is, you need metadata to do this.The GSA provides a RESTful / xml results delivery services that can be used to create user interfaces such as these
Thanks for your attention.If you have any questions that you’d prefer weren’t discussed in public, please feel free to drop me an email.