Enterprise Search Enhancement Singapore Government

One of the largest ministries of the Government of Singapore chose Pingar for an automatic metadata extraction project. This was part of a program to improve document retrieval in SharePoint 2013. 

Problem

With more than 2500 users directly within the ministry and having multiple child agencies and subsidiaries associated with them, this ministry of The Government of Singapore had close to 10 million documents stored across multiple document repositories, primarily SharePoint 2013. 

Without any content organisation policies and processes in place, users stored documents in disparate locations, created duplicates and struggled to locate documents when in need. It was increasingly difficult to manage such a huge volume of documents and with an influx of new documents getting stored daily, they decided to implement an intelligent content analytics and information management system that will empower SharePoint 2013. 

The ministry identified search efficiency as the primary problem due to the poor content tagging and categorisation process by its users. To solve this issue the ministry wanted to reduce or remove all manual and repetitive tasks related to information management, bring in automation and build intelligent systems and thus enable their skilled workforce to focus on more important and productive tasks. 

Solution

Pingar’s DiscoveryOne Content Enrichment solution and services enabled the ministry to implement automatic metadata extraction both from the existing 10 million documents and all new documents generated. DiscoveryOne extracted metadata such as Key Phrases that describe the main topics of a document and categorised the content against a taxonomy. It automatically detected and tagged content such as employee names, departments, locations and customised fields.

DiscoveryOne was able to match similar terms, misspelling of words, and equivalent spelling in variations of English and other languages such as Chinese. A key part of the solution was enabling faceted search, statistically proven to be the best attribute to improve the user search experience within an ECMS. Faceted search enables users to search and find information in the enterprise as easily as they can on Amazon or Ebay with filtered search results by product, organisation, location etc.

The entire document tagging and classification process was automated with little or no dependency on users.  This resulted in consistent, accurate and relevant tagging for all documents which ensured the documents were found quickly via metadata powered faceted search. After implementation it was found that users were able to search and find the documents in less than 70% of the time they had spent earlier. They were also better able to discover information beyond their previous ability in general search, resulting in more useful information being discovered.

Users were able to use the ECM systems more efficiently – finding it was easier to create, edit and finally search for documents. This resulted in them uploading more and more documents into ECM instead of storing them in file systems and hard drives. The project opened doors for extending the automation process to other systems such as Documentum, beyond SharePoint 2013, which is being explored now.