Automated search engines can parse complex legal information and return more contextually-sensitive and relevant results that can be easily integrated with a law firm’s existing knowledge management systems and workflows.
The Future of Legal Search: Meeting Lawyer Requirements by Delivering More Contextually Sensitive and Relevant Results
1. • Cognizant 20-20 Insights
The Future of Legal Search:
Meeting Lawyer Requirements by Delivering More
Contextually-Sensitive and Relevant Results
Executive Summary Need for Change
Research is one of the most vital and time-con- Legal research is the backbone for building a
suming activities in a lawyer’s workload. While strong case. For decades, lawyers have been
the legal information industry has innovated by dependent on the duopoly of Lexis.com and
improving search efficiency, the effectiveness of WestLaw to provide vast information databases,
search largely relies on the research expertise combined with internal expertise and experience
of lawyers, paralegals and law librarians. The in searching, to uncover relevant results. While the
vast growth of available data adds additional industry has always focused on expanding cover-
challenges to the task of identifying the most age of legal information sources, it is ultimately
critical and relevant information to a case or the relative ease and accuracy of finding relevant
client matter. resources that can generate maximum value for
the legal community. Information service pro-
With access to activities and data that a lawyer viders are providing options to allow lawyers to
manages — including client information, current search more efficiently, but the overall results are
matters and ongoing client development opportu- still largely a measure of the lawyer’s inputs.
nities — next-generation search algorithms should
be able to understand lawyer research require- The industry has operated in a “pay-per-search”
ments and automatically parse legal information or flat-rate subscription model, which has forced
databases to extract relevant resources. The vast lawyers to be judicious in their use of search.
search history and usage patterns available from While large law firms have often negotiated
information service players can be leveraged to flat-rate subscriptions to large comprehensive
build intelligent search systems that can largely databases, smaller firms and sole practitioners
automate much of the tedious research work and have felt the pinch. In many cases, the emphasis
present relevant results directly, improving speed, has been on paying for search rather than the
accuracy and relevance. results. Attorneys’ clients pay for results, not the
process. A similar practice can be adopted in the
This white paper explores a scenario for con- legal information search industry, with revenues
structing an automated search engine that parses derived from returned results — not the search
complex legal information and returns more process.
contextually-sensitive and relevant results that
can be easily integrated with a law firm’s existing Market trends — such as the rise of alternative fee
knowledge management systems and workflows. arrangements and initiatives like the Association
cognizant 20-20 insights | september 2011
2. of Corporate Counsel’s Value Challenge — have to become a one-stop shop for not only all cat-
forced law firms to find efficiency while not sacri- egories of legal content, but also any content that
ficing accuracy and comprehensiveness. As such, can have a bearing on a case.
any ability that lawyers have to quickly obtain the
most accurate and complete set Truly intelligent search has arrived. Examples
Social media, of results is likely to help them include the iSeek capabilities currently in use at
the University of Pennsylvania Medical School
crowdsourced data meet this efficiency goal and and Wharton Business School, as well as the
client needs.
and other sources semantic search capabilities of TotalPatent and
of information Large law firms have built their other similar solutions.1 Information services
own knowledge management consumers expect integration into their workflow,
continue to generate databases and have huge such as Jigsaw Data Fusion’s seamless connection
volume and increase volumes of work products with Salesforce.com and D&B 360’s interoper-
complexity. available. Integrating these ability with Microsoft Dynamics. As consumption
data sources along with the patterns shift, so does the expectation of infor-
public legal sources can maximize the collective mation services providers to provide meaningful
knowledge for members of the firm. insights and contextually-rich content that is
easily accessed, searched, found and applied.
Changing Industry Themes
While numerous developments have impacted With the rise of newer players in the market (i.e.,
information services players, several critical Bloomberg Law) that are seeking to expand their
themes have emerged to truly transform the reach through novel pricing models and growing
industry: price sensitivity among users, a push toward a
pay-per-results model could emerge. Free content
• The explosion of complex data. providers such as Google Scholar have also
• Changing consumption patterns. impacted pricing and revenue models. With new
tools and smart content available, information
• Conditions for business model innovation.
services providers are able to offer micro-seg-
As the sheer volume of data increases, so does mented and localized data on a scale previously
data complexity. Consumers of information now not possible. All of these conditions have created
see the rapid rise of audio and video not only as opportunities and even mandates for innovative
part of multimedia sources but also as a primary business models.
means of delivering a variety of content types. As
more and more courts digitize, some are going Collectively, these trends make clear the need
a step further. For example, the Illinois State for information services providers to innovate on
Supreme Court has eliminated their core search platforms and provide additional
Leveraging search print records of its cases, value-adds for the legal industry.
history, information fully digitizing them butaudio
providing near-real-time
also
A New Look at Search
search providers and visual feeds of court pro- The current search model involves lawyers who
can start analyzing ceedings. Social media, crowd- use their intellectual capacity to determine appro-
how lawyers actually sourced data and other sources
of information continue to
priate keywords and specifying filters. Results
are determined through strict keyword matching
search to build generate volume and increase and use of synonyms. Leveraging search history,
artificial intelligence complexity. information search providers can start analyzing
how lawyers actually search to build artificial
tools for constructing Consumption patterns con- intelligence tools for constructing queries based
queries based on tinue to evolve, even beyond on cases on which a lawyer is currently working.
cases on which a the obvious impact that mobile
devices have had on the indus- Factors such as the metadata for the case and
lawyer is currently try. Users are increasingly com- the position of the lawyer, along with the stage
working. fortable with content formats of the proceedings, can be used to determine
other than text alone. User- the lawyer’s research objective. Search engines
generated content, audio, video and podcasts are translate this objective into a search query,
now considered relevant sources of information. which can be run against the legal information
Many users expect information services providers database.
cognizant 20-20 insights 2
3. Search Model involves analyzing the pleadings to understand
the legal issue. A case keyword list is generated
Building the Search Framework
by running the legal search ontology against the
By analyzing search queries run by lawyers, as case documents to uncover keywords related to a
well as using standard dictionaries, it is possible specific case. The output will form the backbone
to build a legal search ontology (see Figure 1). The of context. For example, for patent-related cases,
following dictionaries can be utilized to do this: the presence and absence of keywords typically
associated with “patents” will
• Legal terminology dictionary (a dictionary of
be checked to form context
legal terminologies that also shows similar and Deriving context
(i.e., violation of patent,
inappropriate licensing fees, involves analyzing
related legal terms).
• Practice-specific dictionaries. objection to granting of the pleadings to
• Industry-specific ontologies. patent, etc.). Product line
understand the
and litigant involvement can
• Keyword dictionary based on keywords used in
be generalized to enhance legal issue.
search.
context by using attributes
• Corporate and legal directory showing major such as industry and category of litigant (multi-
corporations, legal firms, etc. national, government organization, small firm,
individual, etc.). This will aid in finding resources
Search Query Pattern relevant to this case category.
Search queries of successful searches can be
analyzed to discover patterns for searches across Metadata for each case will contain context and
various case types (see Figure 2). Patterns would case information, such as case type, jurisdiction,
cover the usage of filters like jurisdiction, date industry, litigants, attorney, judges and stage of
range, litigants, Boolean operators used between proceeding, along with keywords derived.
keywords and type of keywords used.
Search Process
Deriving Context Sensitivity Building the Search Query
To understand what a lawyer requires in his The purpose of search entails:
research, the system should be capable of iden- • Identifying resources that are related to the
tifying the context of any case. Deriving context context of the case.
Building a Legal Search Ontology
Common keywords
from search history
Legal terminology
Practice-specific
dictionaries Build legal
search ontology
Industry ontologies
Corporate and law
firm directory
Metadata creation Tag content
Legal content Smart content
Figure 1
cognizant 20-20 insights 3
4. Using Pattern Recognition to Optimize Search Query Results
Smart content Analysis of Analyze usage of filters,
successful Boolean operators Generate search
searches used in search query patterns
Search and results
viewed hsitory
Run automated
search query
Build search query
Smart based on metadata Compare Update search
content and tags for sample result sets query patterns
legal resource
Run searches
in which this
resource featured
as result
Search and results
viewed history
Figure 2
• Uncovering information related to litigants mining which results presented in the search are
involved in the case. selected, etc.).
• Discovering trends with potential ramifications.
Similar Results
When a lawyer indicates the case being worked By leveraging user history, it is possible to identify
on, the system generates multiple search queries results that are similar in nature but do not meet
to address various categories of search. Metadata the search criteria. When two or more results from
from the case, including context, will form the a search are viewed, there is a high possibility
building blocks of the search query. The search that the results are related. By combing through
query will be formed by arranging these blocks the result view history, it is possible to identify
in a suitable pattern, with a relevance score groups of resources that reveal a high degree
formula set based on search pattern history. This of similarity based on user perception. If one of
search query will be run against the database to the results matches the search criteria, there is a
determine results. The scoring formula will assign strong likelihood that other similar resources may
weights to various parameters based on how be relevant. By presenting these resources as a,
relevant they are to the search. “You may be interested in these results” section,
the lawyer is provided a broader range of poten-
Resources matching the case context are tially relevant search findings.
considered most relevant. “Keyword matching”
search results will be ranked at a lower priority. Fine-tuning for Optimal Results
Keywords still need to be considered to ensure no In case of searches that generate very few or no
relevant results are missed even though they may results, the search query may need to be modified
not match the exact context. A wider interpreta- to generate results that are of lower relevance.
tion is required for trends. Search results will be Options include broadening date ranges, jurisdic-
classified into various categories during display tions, industries involved and more lenient con-
to allow the lawyer to quickly drill down to what sideration of keywords.
is required.
Display of Results
This system learns how best to search by continu- Display results clustered around various views
ously updating patterns based on user behavior (filters) — source, industry, litigant, jurisdiction,
(e.g., modification of the query by the user, deter- date range, etc. — will lead to better organiza-
cognizant 20-20 insights 4
5. tion of results. Selecting a particular view will Modification of Search by the Lawyer
allow the lawyer to drill down into the results. If a lawyer is not satisfied with the search run
Additional relevant resources that cover profiles, by the system, he can view the query run by the
analytics, public records, etc. can be shown in a system and modify it. These
sidebar to deliver complete coverage. Integration modifications will be stored Color-coding
of “similar results” and relevant work products by the system and used to
from the firm’s knowledge management system, refine its search algorithm. schemes can be
when possible, will deliver comprehensive results Lawyers can also view options used to indicate
and ensure tighter integration of research with to set preferences for searches whether a result has
internal work products. Leveraging other online generated. They can specify
litigation products can display filings made by the which parameters to strictly already been viewed
firm, enabling additional productivity and effec- ensure, which to ignore, as by the lawyer or
tiveness. well as relative relevancies and another colleague,
threshold relevancies. They
can also customize the views as well as previous
Color coding of results can further help lawyers
select what is most important to them. Color-
coding schemes can be used to indicate whether
to match their priorities. involvement of
a result has already been viewed by the lawyer Extension of Search
the law firm or its
or another colleague, as well as previous involve-
Proactive Search
client and whether
ment of the law firm or its client and whether the
result has been recently modified. Lawyers are willing to pay the result has been
a premium for timeliness of recently modified.
Lawyers can also set alerts for any view displayed critical information and access
in the results page to allow for more timely and to high-value expert content. Legal informa-
effective change tracking. tion services providers typically tend to provide
Automated Search Process
Anticipate lawyer’s interest from
client list, current matters and
client development opportunities
Dockets Determine
research context
Publications Retrieve user
preferences
Industry News Search
Build search queries
Content enrichment,
Articles including tagging and Law firm
metadata creation Legal knowledge Score results
information management on relevance
Citations repository system
Generate clusters
from results
Profiles Analysis
Additional resources
like analysis,
Social Media similar resources
Reports/profiles
Court recordings,
including audio/video Display results
Log results viewed
Usage database
Fine-tune search algorithm
Figure 3
cognizant 20-20 insights 5
6. alerts on a daily basis, summarizing all the key Litigation Workflow Platform
changes that a lawyer wants to keep abreast of. The search platform can be extended beyond the
Newer players, such as Law360, are addressing traditional research space to cover other areas of
this gap by providing near-real-time alerts and the legal value chain. Integration with online filing
expert opinion.2 However, they still rely on lawyers systems can help lawyers quickly retrieve work
explicitly configuring the system to follow alerts. products from their firm with a strong bearing to
the motion currently being filed. The platform can
The intelligent search system can be used to also be extended to help people who are searching
proactively find information that is relevant to for a lawyer. Based on the keywords and other
the lawyer. By knowing which cases a lawyer is details entered, the system can retrieve a list
currently working on, as well of lawyers who have worked on similar types of
Proactive search is as the lawyer’s long-term issues by searching for cases with similar context
an ideal opportunity clients, the system can search and extracting lawyer names from the metadata.
for new and interesting devel- It can also show resources such as lawyer profiles,
to highlight the value opments and deliver them in
which can help users make smarter decisions on
of paid content. By the form of real-time alerts. hiring lawyers and verdicts for similar cases to
providing relevant A customizable home page
analyze the possibility of winning cases.
free content and that makes relevant content New Social Sources of Content
abstracts of paid collections easily accessible
The proliferation of user-generated content,
provides a valuable service
content, the legal to lawyers by keeping them
videos and social media is challenging the current
information industry up to date with the latest
legal research system. Most of these new sources
of content can help lawyers gauge public percep-
can target upgrading happenings. Based on user-
tions that can help them win over juries, such as
stated information and
of customers. usage patterns, the page
video proceedings of court activities that can
help them analyze nuances such as non-verbal
could emphasize different
communications. Given that Twitter, Facebook
resources to solo practitioners, for example, than
and other social media postings may become vital
for a litigator at a large firm. The legal informa-
evidence in many categories of cases (e.g. dis-
tion industry divides content into “paid” and
crimination actions), lawyers need to be aware of
“free” categories. For long-term growth, infor-
such postings made by their clients or opposing
mation services providers need to increase the
parties.
conversion rate of free content consumers to paid
content consumers. Proactive search is an ideal The current challenge is how to help legal profes-
opportunity to highlight the value of paid content. sionals easily discover such information. Much
By providing relevant free content and abstracts of this content is fragmented across the Internet
of paid content, the legal information industry with very little value-add in terms of making it
can target upgrading of customers. discoverable and easier to analyze. Feeding these
sources of information into the context-based
Educational publishers give students the option to search framework will enable the co-mingling of
buy portions of content that are tailored to their new, relevant multimedia content alongside tra-
needs. Licensing rights also include access to ditional sources of legal information, providing a
content that they own in perpetuity or for a specific one-stop research solution.
duration of time (i.e., a semester). Similarly, rights
to content assets can be made available through Collaborative Search
bundles or collections that are convenient to the Search results and alerts can be used to populate
end user. Proactive search can enable creation research repositories for various contexts. These
of such bundles that will be available for specific repositories can be shared with colleagues,
periods of time (e.g., duration of a trial). allowing them to view results, suggest new sources
for the repository and add comments. Reposito-
Proactive search is also an ideal way to consume
ries allow for the tracking of who has viewed and
research when using smartphones and tablets.
downloaded results and also partitioning results
This reduces the need for users to physically type
into subsets that can be reviewed by different
keywords and search; instead, they only need to
members of the team. A colleague can also follow
navigate through returned research content.
a repository via alerts that are sent to subscribers
cognizant 20-20 insights 6
7. when new results are added. Repositories can be performed. A second shift is now on the horizon
stored in the firm’s own knowledge management in which the medium of consuming research
systems or service provider’s cloud infrastruc- will expand from PCs to encompass alternate
ture, allowing easy access from any device. electronic devices, such as smartphones, tablets
and e-book readers. The legal search industry
Lawyers can share abstracts of user-generated needs to address content dis-
content and other results from the repository tribution and accessibility lim- Better value
with acquaintances in their professional network itations of their products to
to get advice on source reliability. Unreliable ensure an optimal experience
propositions such
results can be discarded from the repository. across electronic devices. as pay-per-result
User-generated ratings can be used to guide the
relevance of results. Integration Across
and assistance in
Litigation Products discovery of relevant
Future of Work results can improve
Large legal information ser-
For a long time, legal research products have
been targeted for legal professionals, government
vices players like LexisNexis conversion rates.
and WestLaw also develop
institutions and law schools, and the products other products targeted at lawyers (e.g. case man-
have been developed in silos, away from other agement and analysis, CRM and practice manage-
tools used by the legal profession. Legal search ment tools). However, development is often con-
needs to be better integrated with other aspects ducted in silos and provides limited opportunities
of lawyers’ and other legal professionals’ daily for collaboration among lawyers.
tasks and also address the needs of non-profes-
sional users. Ideally, a single-sign-in, cloud-based solution that
provides access to various tools and ensures
The Non-Legal Community maximum integration of research and case data
The “do-it-yourself” trend is invading the legal with litigation tools will benefit lawyers the most
space. Rather than approaching a lawyer at the and also help to attract users and keep them loyal
first instance of trouble, many people now conduct to one platform.
initial research themselves before contacting a Ideally, a single-sign-in,
Current examples of collabo-
lawyer. To cut costs, many individuals often draft
ration services — such as Lexis
cloud-based solution
their own contracts, wills, etc. rather than using
lawyers. for Microsoft Office, which that provides access
combines Outlook, Word and to various tools and
Many of these individuals rely on free sources SharePoint — that provide col-
of legal information such as Google Scholar and laboration tools and include
ensures maximum
Fastcase. These players offer little value-add but search functionality for integration of research
are often seen as the only affordable sources of desktop and internal docu- and case data with
information. Incumbent legal information service ment management systems
players such as LexisNexis and WestLaw now need to be extended to cover
litigation tools will
offer free versions of their search products with the entire litigation platform.3 benefit lawyers the
limited functionality and content. However, they
Personalization
most and also help
have had limited success in converting free users
to their paid products, which is their ultimate aim. Legal search providers need
to attract users and
Better value propositions such as pay-per-result to personalize content based keep them loyal to one
and assistance in discovery of relevant results can on stated interests (registra- platform.
improve conversion rates. A new intelligent search tion), electronically recorded
framework can help them in this regard. Further habits and aggregated data. News needs to be
integration of search with other products — such as available in real-time for lawyers to be aware of
legal form downloads, blogs for expert legal opinion, the latest happenings that can affect their clients.
community forums for discussing legal topics and This includes fully customizable U.S., local and
lawyer leads — can provide a one-stop solution for other news, along with the latest business and
addressing the needs of this community. professional news covering practice-specific arti-
cles, client news, verdicts and noteworthy cases.
Legal Community
Device Agnosticism Original content
The shift in research from print to electronic has Search providers need to make available original
drastically altered the way legal research is being content, apart from acting as information aggre-
cognizant 20-20 insights 7
8. gators, to maintain their relevance to lawyers, be an effective way to acquire customers for lim-
as well as tackle the growing threat of free legal ited-use sites such as EZ Law that are generally
information aggregators. The combination of used only as needs arise.
in-house journalism, licensed
Addressing Lawyer Concerns
Ultimately, search content contentand user-gen-
erated
feeds
can ensure There are always misapprehensions regarding
providers must that better information is new technology whenever it is introduced. There
engender loyalty by delivered to users. Certified is a high probability that lawyers and librarians
offering more utility experts can provide no- or
low-cost blog and video blog
may be skeptical of the system’s ability to judge
what is relevant. To build initial accuracy of
in one place, deriving content. Search providers can system while not compromising the quality of
value through social also build online communities results provided to lawyers, such systems should
media resources and of legal to each other who
can link
professionals
and
debut first in law colleges, where law students
can rate the results provided and suggest modi-
creating a captive also share discussions, best fications to the search engine. These recommen-
base to cross-sell practice tips and network. dations can be used to tune the search system
before it is rolled out to a wider audience. This
other products and Ultimately, search providers approach may lead to quicker acceptance by the
services. must engender loyalty by legal fraternity.
offering more utility in
one place, deriving value through social media Another issue that can hinder adoption is
resources and creating a captive base to cross-sell the legal staff’s perception that their years
other products and services. of expertise in research are being rendered
obsolete, which might make them resist adoption
Free Users of such platforms. Librarians and other legal staff
Legal search providers need to convert free users may also feel threatened by automation in their
into a loyal and captive audience that becomes traditional research domain. To address these
familiar with their brand and frequently returns concerns, search automation can be positioned as
through e-mail alerts, newsletters and social a tool that can help legal staff reduce time spent
media updates. This well-indexed and constantly in mundane search activities and focus more on
refreshed content creates a larger Web presence understanding and analyzing results to generate
and search engine optimiza- maximum value.
Search automation tion advantage to further Final Arguments
promote usage and minimize
can be positioned as the reliance on advertis- With the growing advancement of search
a tool that can help ing spend to build brand. technology, it is possible to automate large parts
of legal research. Leaving the tedious work of
legal staff reduce time Revenue can be derived searching to computer systems allows lawyers to
through advertising, pay-
spent in mundane per-results and conversion focus on building a strong case. The explosion of
search activities to paid services. Frequent complex data makes this capability not only more
relevant but also more critical to ensure that
and focus more on trial offers can be used as a lawyers arrive at the right answers quickly and
promotional tool to provide
understanding and partial functionality. efficiently.
analyzing results to Leveraging search and usage history can result in
Similar user experience for
generate maximum paid and non-paid users the construction of more effective search systems
value. with more functionality and that anticipate what a lawyer requires and quietly
additional access to content work in the background, delivering relevant infor-
for paid users needs to be provided. For example, mation directly.
a certain number of placeholder content boxes
In the future, the entire process of search may,
might always be persistent on the screen, even
in fact, disappear from a lawyer’s workload. Infor-
though the services have not been activated. If
mation systems will continuously keep track of
users select a content module to which they’re not
changes and download relevant information
subscribed, there will be a call to action on how to
directly to the lawyer’s work folders.
subscribe and activate it. Similarly, this may also
cognizant 20-20 insights 8