BRSSearch and Markup

1
BRS/Search and Mark-up

• Nico Noorlander

2
Programm

• Introduction Textinfo Technologies BV
• Introduction TT/Search
• Demo TT/Search
• Introduction BRS/Agra
• Demo Agra
• Q A (at least the Q)

3
Textinfo Technologies b.v.

• Founded in 1994
• Mission Being successful in selling solutions
  for problems focused on large amounts,
  strategically important, UnStructured, mostly
  textual information

4
Textinfo Technologies b.v.

• Initially just a sales office of Dataware.
• A slight move from software sellers to solution
  providers (a cliché).
• Still highly dedicated towards DW / BRS/Search.

5
TT and Marketing

• In Holland Hit and Run sales not acceptable.
• Customer Relation is most important.
• We solve his problems.
• We read Principles of Marketing from Phillip
  Kottler about 15 years ago.
• We thought American SW companies did it as well.

6
Strategy

• Strategy is focussed on DW independency.
• Just based on experience never bed on one horse

7
Strategy

• Fulfilment of the Strategy is quit complex.
• 1 Need to develop own interfaces
• 2 Those interfaces need to be focussed on
  BRS/Search or/and/what??

8
Strategy

• Strategy is focussed on BRS/Search independence
• Holland is to small for local engines while
  language is so complex that feature rich engines
  from USA are use-less

9
Back to the basics

• Customer is looking for Information!.
• Information not necessarily text-only
• Text in the past was Collection of letters in a
  terminal a like font.
• Because of Bills success the customer expects
  WYSIWYG.

10
Mark-up, a search for!

• RTF,
• SGML,
• HTML,
• XML,
• ML

11
RTF

• MS compatible, or not?

12
SGML

• Highly academic
• Knowledge depended (Knowledge gives power, DWII
  KMS, other subject)
• Professional Big publishers love it

13
HTML

• Most popular Mark-up language of the moment.
• To dynamic / un-controlled.
• To less standardised or is that caused by the
  browser???.

14
XML

• The standard of the future??????
• At least MS supports it.
• What bill believes seems to be ok.

15
Textinfo Technologies b.v.

• Http//home.textinfo.nl/freddy/Mac/BrsHtml/Web/A/b
  rsHtml.htm
• We developed a client ML independed environment.
• Standard client environment for Search and
  Retrieval called TT/Search.
• Browser connectivity called BRS/Agra.

16
TT/Search

• Search is nice but how do we add or modify data??
• After a search live starts being complex??
• Finally i like to include images etc.
• Word supports that as well!?!
• One-stream publishing on the WEB.

17
TT/Search

• Supports RTF, HTML, ASCII!, xml
• Supports Asci in HTML out
• Supports HTML in out
• Support combined ML per paragraph

18
(No Transcript)
19
TT/Search demo

• Demo

20
TT/Agra

• Developed to build on the existing customer base.
  
• Follows the characteristics of the WEB
• Needs to be useable for Web masters with CGI/HTML
  knowledge. (no BRS experts)

21
TT/Agra

• Needs to be Fast.
• Needs to be Flexible and platform independed like
  BRS/Search.
• Needs to support PTF.
• Needs to support diacrits
• Needs to be ???? (managerable by Textinfo)

22
Features of Agra (1)

• Uses the CGI-protocol
• POST and GET can be used (POST can handle a lot
  more data because its a stream and you dont get
  long URLs)
• Use of BRS Backreferencing
• Use of usernames/passwords through standard
  HTTP-authentication

23
Features of Agra (2)

• Short connection time between the browser and the
  webserver
• Fast
• Easy all search-parameters can be defined in the
  search form, so anyone with HTML-knowledge can
  create an interface

24
Features of Agra (3)

• Flexible output through standard Print-Time
  formatting with HTML
• HTML in databases (also complete documents)
• Diacritical characters are being supported
  (indexed and searchable as the same characters
  without diacritics)

25
Data-transmission
Input data (stream or fixed string)
Webbrowser with form
Webserver
Agra
Query database
Format data through PTF
Webbrowser
Webserver
Search results/documents
26
Agra demo

• Demo