“The MCC is pleased with the results of this project. It has been able to leverage its SharePoint investment and use the native features provided with the product. The conceptClassifier for SharePoint platform, working natively within the SharePoint environment, has improved search for the business users. Overall, the project has achieved, if not exceeded its objectives.”
- Effectively retrieve information
- Efficiently retrieve information
- Visibly improve the accuracy and relevance of search
The Medical Council of Canada (MCC) strives to achieve the highest level of medical care for Canadians through excellence in evaluation of physicians. It assesses over 11,000 medical students and graduates each year through its examinations.
The Medical Council of Canada (MCC) had been using Microsoft SharePoint 2010 for three years. To improve on this investment, the MCC recognized that search is a critical business function and wanted to improve the search experience for its business users. The primary issue was the inability to utilize one search interface and retrieve highly relevant information from multiple repositories. Important requirements were a tool that was easy-to-use, and accurate, relevant search results.
Overall considerations included avoiding end user adoption issues. The MCC wanted an intuitive interface, without business users having to learn advanced search techniques. As with all business applications, the justification and return on investment also needed to be achieved. Finally, the technology decision had to be forward compatible with any new changes to the existing IT infrastructure, repositories, and applications.
The MCC had deployed SharePoint Search, but had not customized or tuned the product using the native features to improve search experience.
The MCC is focused on business user satisfaction, and providing technology tools to enable users to accomplish their day-to-day activities more efficiently. To understand the needs of the business users, the IT management ran focus groups and delivered an internal survey to understand and gain insight into the main issues.
The summary of the results of the survey indicated that:
- Business users wanted a single search interface across diverse repositories.
- Business users wanted a single view of content when searching diverse repositories.
- Improved ease of use needed.
- Improved accuracy of search results required.
In addition to addressing the business users’ concerns, IT management had its own priorities in developing a search solution that included the following:
- A transparent solution with little or no business user training required.
- A way to leverage the enterprise’s investment in SharePoint.
- Adherence to stringent security permissions in place.
- Use of search as an ‘enabling’ technology, so business users could manually change the auto-classified semantic metadata tags.
The project objectives were defined to address the needs of internal business users, satisfy IT objectives, and integrate with SharePoint. Once these objectives were in place, the MCC began crafting a solution to address the challenges of search.
After planning the objectives and scope of the project, the MCC focused on leveraging its existing investment in SharePoint, and reconfigured a number of the native features available for search. As it had initially deployed SharePoint Search out of the box, this first step was to ensure it was taking advantage of inherent SharePoint features. Some of the customization and deployment activities included:
- SharePoint Web Analytics Search Reports that provided the SharePoint Search project team with information on what the business users were searching for, and showed if the team was succeeding or failing. The reports were modified, to gain a better understanding of what the business users were typically seeking during the search experience.
- Best Bets, although a manual process, was implemented to highlight the most often used documents, such as expense claims, to reduce time when searching.
- Search Scopes and Tabs in the Search Center are now being used, and have been customized to reflect how the organization organizes and utilizes information. For example, one of the specific Scopes created was for presentations only.
- Property Mapping was deployed to improve indexing and results.
- Query Suggestions proved manually intensive. The limitation is that SharePoint creates the rules after multiple manual click-throughs. The search results were noticeably improved through this task, but the time and effort was considerable, so it was eventually abandoned.
The MCC had started with a focus on using SharePoint to the fullest, to address the needs of business users. Since search is a critical business requirement in the organization, it felt that adding the conceptClassifier for SharePoint platform would further enhance the environment.
Staffing for the project included a resource responsible for planning and project management, a senior systems administrator, and a junior information management specialist. Supporting the internal team was a senior level Concept Searching technical resource.
The initial installation of conceptClassifier for SharePoint included setting up the conceptClassifier for SharePoint server and the SQL server. The next step included configuring conceptClassifier for SharePoint. This was achieved in less than one day. Additional time was spent creating content types and site columns, to enable conceptClassifier for SharePoint to populate the SharePoint environment. It also allocated one week of trouble shooting.
The final deployment task was to set up the taxonomy manager component in conceptClassifier for SharePoint. The initial taxonomy framework was created in approximately one week by an information management specialist. Although taxonomy management is an ongoing activity, the speed at which this was able to be accomplished, with no experience of using the tool, was quite remarkable, and illustrates how easy the product is to use.
The MCC is working towards assigning departmental site administrators to maintain the taxonomy, based on their subject matter expertise. Although the platform eliminates end user tagging, the MCC felt that authorized business users needed the ability to manually change, add, or edit the automatically generated metadata.
Goals Achieved and Results:
The initial objectives were originally classified into Functional and Non-Functional Requirements. The Functional Requirements are summarized below:
- Visibly improve the accuracy and relevance of search results – Achieved: Immediately.
- Include a relevance model that provides the ability to tune search results – Achieved: Uses custom metadata that appear as refinements next to the search results. Relevance model includes existing controlled vocabularies.
- Allow the search administrator to collect data on search-related end user behavior – Achieved: Uses modified out-of-the-box reporting capabilities provided in SharePoint.
- Support secure mobile access to enterprise systems and information – In Progress.
- Leverage rich, semantic metadata – Achieved.
- Minimize (or remove) end user involvement in the application of document metadata – Achieved: conceptClassifier for SharePoint automatically generates semantic metadata. The MCC is in the process of creating additional content types.
- Utilize a distributed search model rather than search based on a singular, central repository – Achieved.
The Non-Functional Requirements are summarized below:
- Easy for end users with basic search skills to efficiently retrieve information without significant training, i.e. no more than a couple of training sessions – Achieved: Solution has currently not required training.
- Easy for end users with basic search skills to effectively retrieve information without significant training – Achieved: Solution has currently not required training.
- Security, including tight integration with Active Directory for authentication – Achieved.
- Integrated with the MCC’s existing repositories, with particular emphasis on the Windows environment – Achieved: Integration with distributed files shares and SharePoint.
- Relatively easy to customize – Achieved.
The MCC is pleased with the results of this project. It has been able to leverage its SharePoint investment and use the native features provided with the product. The conceptClassifier for SharePoint platform, working natively within the SharePoint environment, has improved search for the business users. Overall, the project has achieved, if not exceeded, its objectives.
Future Business Applications:
Although the MCC selected conceptClassifier for SharePoint to specifically enhance the accuracy, relevancy, and the business user experience of search, the product also provides capabilities that can be used in the future, to improve other business processes that require the use of metadata. The MCC is evaluating two activities for the future. One is automatic metadata tagging and identification of records, for its records management application. The other is data loss prevention, to automatically, in real time, identify and secure any organizationally defined data privacy or confidential content, and prevent its unauthorized use and portability.