Search: it isn’t dead – it’s evolving at speed

Graham Digby explains how intelligent search is crucial for the future of legal research.

“The word 'search' is a negative word. It fairly reeks of loss and effort” says James Whittaker, former Googler and technology executive, in his blog Why I hate search. Over the past few years, LexisNexis has had a huge focus on search and has been looking at ways to reduce the effort on the part of our users; specifically looking into how our users’ search behaviour has changed. Historically legal research was conducted by a few trained individuals in a law firm, searching through legal “databases”. All of the “search intelligence” resided in the user creating the Boolean queries that would return the best results.

Now our users are far higher in number, more demanding and they need the applications and content they use to be smarter. It is on us, not the user now, to make search as quick and as painless as possible – to get them the information and answers they need. To meet this challenge we have an ever-growing suite of tools and techniques that we use to add value.

It starts with the user

No task that involves searching starts with a set of search terms and ends with a set of results. If a user wants to find out if a case cited in a document has had any negative treatment rather than searching for the case and its status, why not run the check in MS Word? If the input for a trainee’s research task is an emailed request for information, the intelligence can start in Outlook picking out relevant information and starting the research process for the user without them having to think what the search terms are. Understanding the workflows, inputs and desired outputs that form a user task creates the space for innovation and increased value.

Find the intelligence (wherever it is)

To best serve the user, search needs to be intelligent. That intelligence can come from many sources. In the legal sphere we are lucky that the content is often written with a user and answer in mind; by analysing, combining and deduplicating content we can create curated answer setson specific topics.

Everyday 1000s of users are providing data on what is important and relevant to them. I’ve been pleasantly surprised by the requests from our users asking that we make use of the wisdom and intelligence that they give us to deliver a better experience. By disaggregating the intelligence from the products we are able to provide value to clients searching their own content through the synonyms, acronyms, citation recognition and content classification services that we have developed.

One search algorithm to rule them all?

Just as our users require different answers and content depending on their tasks, roles and areas of practice, it’s no longer appropriate to provide a one size fits all search algorithm. By using different algorithms for different personas, practice areas and content the results can be more tuned to what the user is looking for. This increases engagement as the experience is designed for them. Using analytics those algorithms can then evolve with the users. It’s important to make sure that increased customisation does not occur at the expense of the repeatability and certainty that users expect from a product.

Technological change

These are exciting times for search as the technologies available enable us to provide more than just results but also answers and analysis. Our US colleagues are combining massive computing power, best of breed technologies and developing new patents to analyse both structured and unstructured content to find relationships that would have been difficult to surface until very recently. We are now seeing both documents and data being delivered to the user. Want to find out if an expert witness has had negative or positive treatment from a judge? Want to find out the trend for damages for a particular injury is? Want to find out how a particular type of agreement is being used in different industries? The combination of technologies now allows us to provide these.

James Whittaker argues some interesting points about the desire (or lack of) to innovate in search. “The problem with Internet search is that being stupid about it is profitable. The more ugly blue links you serve up, the more time users have to click on ads” and“Is it any surprise that technology such as Siri came from a company that doesn't specialize in search? (Where do you place an ad in a Siri use case?)” Google and Microsoft would doubtless disagree with this sentiment pointing to knowledge graphs that can provide answers without opening documents and integrations into the users’ workflow tools such as MS Office.

In legal search we don’t have the same business model constraints as web search providers. We also have a smaller set of workflows to support and we know an awful lot about them. This creates a great opportunity for innovation. The value to the user should be measured by the quality of the answer and the least amount of time and effort required to get it – even if that means that at some point the search itself becomes invisible.

*Many thanks to my esteemed colleagues Mark Wasson and Harry Silver for their insights.