The future of enterprise search technology [Q&A]
Businesses are gathering more data than ever before, but being able to make effective use of it means being able to find information when you need it.
What challenges does dealing with these ever larger amounts of data present? And how can you be certain that your searches are delivering relevant information and insights to drive your business decisions? We spoke to Stéphane Kirchacker, VP EMEA at search specialist Sinequa to find out.
BN: What benefits do enterprises gain from having a good search facility?
SK: Data is increasingly at the core of enterprise operations. Organizations are currently generating more data than ever before, and they expect to be able to leverage that data to make better decisions. The problem is, many organizations lack the tools to make the best use of their data. Unfortunately, existing search capabilities rely on key words and are limited in their scope. They don't enable users to search across scattered data silos or across structured and unstructured data simultaneously.
Implementing an intelligent search platform can make all the difference to a business. With intelligent search, it is possible to gain insight from all your data -- structured and unstructured -- in real-time. This allows employees to obtain more relevant information, faster, and the benefits of this are numerous.
Good search facilities enable individual employees to save valuable time and their cognitive burden is reduced, allowing them to devote more of their efforts towards important tasks such as decision-making and innovation. Cumulatively, the company's workforce spends less time searching for information, leading to efficiency gains and cost savings. Access to richer data also enables better decision-making and more comprehensive research, allowing an enterprise to remain knowledge-driven and competitive in its field.
BN: What are the challenges presented by large volumes of unstructured data?
SK: Unstructured data includes everything from the text in an email to words spoken in a video meeting. Organizations generate vast amounts of this type of data every day and its untapped potential is huge. Yet, while it is relatively straightforward to extract value from structured data, unstructured data is more complex. For example, you might need to find a specific paragraph in a document or a particular sentence in a video meeting transcript. It will be impossible to sift through thousands -- potentially millions -- of other similar documents and transcripts to find it. To make matters worse, your average search tool will be unable to interpret the text in individual documents to find exactly what you’re looking for.
To get value out of unstructured data, a different approach is required -- intelligent search. This draws on a combination of technologies to facilitate the management of unstructured data. For example, it will include a multilingual search engine which means information in different languages can be accessed, while natural language processing capabilities and machine learning algorithms enable the search function to pinpoint the information you need, wherever it may be buried. With intelligent search, unstructured data can be used to its full potential.
BN: How have technologies like natural language processing changed the search process?
SK: Natural language processing (NLP) enables a computer to understand or generate human-like language. In enterprise search, this capability is invaluable. It allows the search platform to accept more general search queries because it can use its human language skills to more precisely determine what the user is looking for. This saves time for the user, who isn't required to work out the right computer language needed to find what they are looking for.
NLP can also help make sense of unstructured data. For example, if a company wanted to understand responses to its social media output, an NLP tool tuned for 'sentiment analysis' would be able to sift through social media comments and other unstructured data to gain insights. Essentially, NLP plays a crucial role in search and increases the relevancy of intelligent search results, making search tools more user friendly.
BN: How can you ensure consumer privacy is protected and regulatory compliance maintained?
SK: It is vital that businesses ensure their information and insights are only made available to legitimate users without compromising security or privacy. Users should only see what they have the permission to access, and business should look at how they can implement access control so that permissions are kept up-to-date dynamically to fit with company policy.
For Sinequa, security is of the upmost importance. As well as having built-in security, our platform can help companies meet regulatory requirements. An intelligent search engine helps you keep better track of your data, and Sinequa’s automated document classification means sensitive data can be immediately identified, tagged, and made visible. Meeting auditors’ requests is much less arduous when such information can be identified quickly and accurately.
BN: What drove you to create a free scientific research repository and portal?
SK: When the pandemic began, we wanted to do what we could to help in the fight against COVID-19. We recognized that there was an ever-growing body of knowledge around the virus, and that scientists and medical professionals would need to access this evolving mass of data to support their research.
Drawing on our expertise in intelligent search, Sinequa developed the COVID-19 Intelligent Insight portal. It is the world's largest repository of curated scientific information specific to COVID-19, including over 170,000 research papers and information from 2,000 clinical trials, with new content being added daily. The portal provides researchers with answers quickly and efficiently, helping them to analyse the latest trends.
Key features enhance the user experience, such as AI-generated article summaries, created using deep learning models, which allow users to browse the contents of an article without having to read it in full. An intelligent document navigator also increases efficiency, guiding the user to relevant passages within a document. We worked closely with users to enhance the tool and to meet their needs. Further features enabling better collaboration and sharing of information were added based on feedback.
It was essential that this portal was free and provided open access to everyone in order to best support the global research community. The application of intelligent search to the fight against COVID-19 demonstrates, for us, the true power of leveraging huge volumes of data. The real-world benefits can be immeasurable.