Whether you are running a preemption check or reviewing
legal scholarship to support your own research, HeinOnline’s Law Journal
Library is an important resource for accessing legal scholarship published in
law journals around the world. In an effort to make the contents of this
resource more accessible and customizable for researchers, HeinOnline
introduced two new tools: (1) Keyword Search Builder and (2) More Like This.
These tools are available in the Advanced
Search features of the Law Journal Library, the ABA Law Library Collection
Periodicals, and the Bar Journals Library.
Search Builder is a natural language processing and machine learning
tool that enables the user to adjust the “boost factors” given to individual keywords.
In layman’s terms, this means that you get to tell the system how important (or
less important) keywords are in your search; basically, you get to have more
control over your search results.
Below is an example of a string of keywords that I added to
identify articles addressing the ethical responsibilities attorneys have to
safeguard client data against a data breach.
HeinOnline's Keyword Search Builder
You can see from the image above that the “weights” can be
adjusted on a sliding scale. The farther to the right you move the weight, the greater the "boost" this term will receive in the algorithm performing your search.
In addition, once you’ve created an initial set of keywords and run a search,
you can continue to add, subtract, or adjust the weights assigned to the
keywords to further refine your search.
“More Like This” tool is another way for researchers to refine their search results by starting with an article that they already feel is
relevant. When you are viewing an article in HeinOnline, the “More Like This” option
is available near the top of the page, alongside the regular options for viewing,
printing and downloading:
HeinOnline's More Like This Feature
By selecting this tool, you are automatically taken to a
Keyword Search Builder results page. However, you will notice that keywords and
weights will have already been created for you based on “interesting words” the
system identified in your original article. Once at this page, you
can alter the keywords selected by the system and thus create your own unique
Machine learning and natural language processing continue to
provide users with new and exciting ways to conduct their research. Be sure to
give these two HeinOnline tools a try as supplements to your traditional
Boolean and natural language searches. They might help you discover a new line
of relevant scholarship that your initial searches missed!
Posted by Melissa M. Hyland on Mon. February 4, 2019 3:03 PM