Government of Ontario

October 2, 2010

The Provincial Government of Ontario is an Enterprise that has a vast array of information assets that are managed by the various ministries and public service agencies. These information assets are stored in multitudes of internal shared folders, databases, content management systems, internal and public facing websites.

An “industrial strength” Enterprise search platform that was flexible and programmable was required to enable unified access to all of these data sources and present search results in a secure environment that allowed role based access to information.

FAST Enterprise Search Platform (ESP) was selected as the strategic search platform and ACIS Consulting as the provider of consulting services to assist with deployment of enterprise search services across various Ontario Government departments.

This deployment process was conducted using a series of discretely defined projects, each department presenting its own business (ROI) justification.

The following is a brief synopsis of some example projects that were implemented by ACIS Consulting in partnership with the various service owner departments at Ontario Government.

Public Facing Search Portal

The Ontario Government was interested in providing a single access point to all publicly available information from all of the government websites. A search based information access portal was viewed as the quickest way to achieve unified and common access to all this information scattered across disparate websites, document formats and service providers.

The public facing search portal, which only had around 50,000 documents crawled from 30 or so websites; has now grown to handle more than 5,000,000 documents crawled from over 200 public facing Ontario Public Service (OPS) websites.

Internal Facing Search Portal

Another objective was the provision a single access point for searching all internal websites that were of interest to all public service employees. A search portal that is only accessible to employees who have authorized access to the OPS network was quickly deployed.

The OPS internal search portal, has now grown to handle more than 2,000,000 documents crawled from over 50 internal Ontario Public Service (OPS) and a selected list of partner websites.

Ministry of Health Shared Folders Content Search

One of the key challenges faced by many internal search projects is the dilemma of providing access to sensitive information while protecting security.

The Ontario Ministry of Health has several departments each with thousands of valuable documents contained in internal shared folder structure. The objective of the search project was to improve access to information stored in shared folders by enabling search and validated access to these documents across various departments.

A department wide review and segregation of the content was carried out to classify content based on sensitivity . FAST ESP was then configured to automatically tag and classify indexed content based on this classification scheme.

This project was used to validate that FAST ESP can be configured for role based access to secure content stored in various shared folders. Content source information can be used to generate rules based classification metadata that can then be leveraged for faceted browsing and navigation.

Improved Access To Govenment Services

The website is the primary gateway for accessing information and services provided by the Ontario Government. A key component of the website upgrade was to improve the search experience by providing more precise results and offering dynamic user interaction and guided navigation.

The Service Location Finder search capability was significantly enhanced by integrating native FAST Search and capability to type-ahead and suggest search results related to database of services accessible via the website.



October 1, 2010

Canwest was interested in improving the search experience to its subscribers of newswire services. Newswire feeds from over 3800 sources around the world are indexed by FAST ESP in near real time basis.
The FAST entity extraction and content categorization capability on massive volumes of content was leveraged to achieve the desired improvement in search experience and customer satisfaction.
Improvement of the search response time performance was among the key metrics used to measure success of this project.


Canada Post

September 30, 2010

Canadapost selected FAST ESP as the core search platform to power its Public Facing as well as its Intranet search needs.

Internet search results use the automatic classification capabilities of FAST ESP to enable dynamic drill down within search results automatically grouped into content targeted to business, retailers, general public or by specific keywords within the content.

September 2, 2010 is the Boston Globe online news and information portal that is intended to present “all things Boston”. The information portal included content from websites of government, education and business sites around Boston.

The key business objective was to make the website “sticky” to visitors by offering engaging federated content provided by businesses and partners in the Boston area.


The New York Times

September 2, 2010

New York Times Archived News Search

The New York Times has a vast collection of archived news assets dating back to early 19th century. The Archived News project was intended to leverage search technology to monetize these assets by making archived news content readily accessible to historians and researchers looking for archived news stories.

In addition to categorizing news articles by publication date and news column names, FAST ESP was used to process this vast volume of content and automatically extract searchable entities within the news articles such as topic keywords, person names, company names, location names etc.

Users can now search this archived news collection by any given keyword and obtain a chronologically organized list of past headlines that are now part of history.

An additional technical objective of this project was to be able to service over 120 queries per second (430,000 queries per hour) while utilizing the smallest possible server footprint.