‘Hello, is it me you’re looking for?’ – Legal Search, Relevance and Credibility

Better legal search resources will inevitably result in a better service for clients; less time spent looking for answers and more time advising and acting.

At its core the criterion for a good search tool is: a product that can produce increasingly relevant and personalised results from credible and trusted sources. As outlined in our article Legal search: what does the future hold?, the difficulty for many within the legal field is the ability to find a single search source that can direct them to the information they need and provide credibility as well as relevance.

Whilst Google is able to meet the demand for relevant and personalised results (see the Knowledge graph) it can fall short of consistently delivering verifiably credible sources; relevance does not always equal trustworthiness. Conversely, many legal databases have hundreds of years of experience and credibility behind them; what they often lack is the technology to provide satisfactory personalised search results, hence a view that sometimes Google is ‘easier’ to use. Using Lexis®Library as an example, products such as Halsbury’s Laws of England and Simon’s Taxes are well known and trusted products within the legal and tax fields, what customers need in addition to this credibility is a search experience that is relevant to them; an increasingly personalised experience.

There can be challenges to developing a resource that is able to deliver both relevant and credible search results. Looking at ‘Google-bombing’ for instance; this is an example of the limitation of developers to control the credibility of their search results whilst still prioritising relevance. Skilled users are able to manipulate the Google algorithm to promote their preferred search result to a higher ranking position on the results page, the most notable example being the top search result for ‘miserable failure’ which for a period of time was a biography of George W. Bush. Whilst instances of Google-bombing have been few and far between in recent years due to developments such as Google Penguin, it still demonstrates that Google’s algorithm is not impenetrable and credibity is still something that Google is working hard to achieve and maintain.

Arguably what Google has in its favour is what Scott Liewehr, CEO of Digital Clarity Group calls ‘the benefit of the crowd’. Google uses PageRank to determine the credibility and relevance of its results by the number of links to that document, with additional weighting for known credible sources. Essentially the more links there are to a website, the more likely it will rank highly within the search results. In Google’s world there are millions of content creators linking to documents and websites, thereby creating a more relevant pool of results. This ‘crowd’ is not necessarily something legal databases have access to, and the number of links generated by smaller pools of content creators may not be enough to determine the relevance of search results for a user. However, this challenge does mean that if legal search tools are to compete with Google, they have to utilise the other resources at their disposal.

What are legal databases doing to improve their search tools?

The introduction of Big data is proving useful for advancing personalisation and relevance within legal search tools. Mark Tolram, Project Manager within the LexisNexisUK Search team acknowledges that whilst ‘everyone uses Google and knows that they do personalisation really well’ the danger with personalisation in a legal and corporate context is that you do not necessarily want a personalised set of results for each user; you are more likely to need a personalised set of results for groups of users within similar fields or sectors. For example, it is not desirable for a number of lawyers within the same law firm and practice area to use identical search terms and get different search results. Of course there are instances where such differences in results could be useful and this is where Big Data comes in; it would enable teams to analyse trends and identify the search needs of certain groups and users based on particular ‘signals’ such as location, practice area and possibly even employer. In essence it could allow for personalisation that is tailored to the needs of the corporate world. Tolram further notes that ‘the future is machine-learning’ and using methods such as Learning to Rank to combine human data entry with machine learning will play a key part in developing more personalised legal searches.

Additionally, many have argued that until recently businesses have not invested enough into their search mechanisms and teams; the idea being that to have an effective search tool the right people need to be in place and senior colleagues need to be made aware of the intricacies of search engine development. David Espley's comments in 'Legal Search: what does the future hold?' point to a commitment by businesses such as LexisNexisUK to developing search solutions not just by utilising the ‘energy’ and skills of its staff but also by investing the necessary ‘time’. Mark Tolram credits LexisNexisUK with providing search teams with the ‘time and space’ to look into the various search concepts in an effort to continually improve the LexisNexisUK search experience. The idea of being given the space to investigate solutions and experiment with search tools is the key to progress and success in any development process and is one which businesses, and in particular legal databases are becoming increasingly aware of.

The focus on ‘customer intent’, or ensuring search results are relevant to the customer, has a different meaning in a legal context than it does more generally: legal search tools have to take into account a greater need for a uniformity of results within fields and practice areas. The strength of legal databases is their credibility; and that is half the search battle. In a world where there is a visible example of search being done well, it is important that legal databases leverage their strengths whilst also exploring the breadth of relevance engineering information available.

Article written by Esther Adetoba

 

 

Filed Under: Technology

Relevant Articles
Area of Interest