Making the search engine of knowledge base ANA more robust with the power of AI.
Make the search engine of knowledge base ANA more robust with the help of Artificial Intelligence (AI).
- Speed improvement
- Accuracy improvement
AI capabilities applied to the search engine using Azure Cognitive Search, Bing Spell Check and custom code.
Just imagine the following: you have a giant database with a wealth of valuable content at your disposal, but you cannot use it to its full potential. What a pity, right? But what if there is an opportunity to extract this potential? Just as any seasoned treasure hunter must have powerful tools and resources at his disposal, today’s ‘digital treasure hunter’ is no different.
Technology, and in particular the power of Artificial Intelligence (AI), can help you fulfill more of your organization’s potential. Read on to learn how this works in practice through the ANA case study. You will learn how dev-UP harnessed the power of AI to make ANA’s search engine more accurate, flexible, and intelligent.
In the opening, we mentioned the challenge of this case. The challenge in making ANA’s search engine more powerful consisted of improving the speed and accuracy. Firstly, the speed of the search engine needed considerable improvement. It sometimes took ten seconds (or more) before the user received a result after entering a search term.
Secondly, the search engine had to be more accurate. The accuracy could do with some improvements as typing errors, for instance, often resulted in no or fewer (relevant) results. A fix was necessary because such a problem could result in membership cancellations due to the deteriorating user experience of ANA. Overcoming this challenge required a robust solution. And what better way to create such a solution than with AI as a strong catalyst?
ESOMAR allowed dev-UP to work on developing and implementing a solution. Given the AI expertise of dev-UP, a solution has been found based on intelligent application of various AI capabilities. Part of the solution was the use of Azure Cognitive Search. This cloud search service with built-in AI capabilities helps identify and explore relevant content at scale. Machine learning techniques make it possible to gain insights from various types of content.
In addition, Azure Cognitive Search also provides semantic search capabilities to understand user intent and contextually rank the most relevant search results. Furthermore, we used Azure’s Bing Spell Check to ensure a fast and high-quality spell check when entering search terms. Previously, search terms were not necessarily corrected, which led to fewer or no results. Combining efforts of Azure Cognitive Search and the Bing Spell Check, this is no longer an issue.
In short, this solution has resulted in a much faster, more accurate and more robust search engine. However, we did not stop there. Dev-UP always goes the extra mile but still keeps things simple. How is that possible?
We keep IT simple by identifying the key benefits technology can bring in your specific situation, a step-by-step implementation, and interim demos to keep you up-to-date about the progress.
The capabilities mentioned in this case sound great. However, the most important thing is what the solution brings your organization of value: What’s in it for me (WIIFM)? In addition to the improved intelligence, speed, and accuracy, Azure has set up their infrastructure in a way that supports automatic scaling.
Using the Advanced Search, members can search more specifically for content based on, for example, date, author, or company name. It is possible to tailor the solution to work with data sources already present within the organization using custom code. No new data needs to be acquired, which saves resources. Finally, during the process, we focused on security by only granting full access to the knowledge base to people with the correct authorization, i.e., the members.