Algorithm Comparison

PrintEmail
Susan Nevelow Mart
William A. Wise Law Library
University of Colorado Law School
2450 2450 Kittredge Loop Drive
Boulder, CO 80309
susan.mart@colorado.edu

Algorithm Comparisons: The Truth About Easy Search

The following PowerPoint has both illustrations and notes. The first slide can be used to illustrate the limits of algorithmic searching and the fact, startling to students, that every algorithm produces different, incomplete, and not always relevant results. The second slide can be used to illustrate the improvement in results with terms and connectors.

WestlawNext and Google can be remarkably good - especially at getting a researcher started. They are incomplete, but so is every search. But Casemaker, Fastcase and even Lexis Advance are vastly improved by terms and connectors and Bloomberg ONLY uses terms and connectors.

Students still need to know how to craft a decent search. I use the comparison charts of the same searches in four different databases using natural language to show the wildly varying results with algorithms, and the same searches using terms and connectors. I also use the slides to illustrate to students how database design is influencing their searches.

I tell my (advanced) students to use both algorithms and terms and connectors, if there is no financial penalty for doing so. They each perform very different functions!

The first slide shows the top ten results for the natural language search right to receive information in Lexis Advance, Casemaker, WestlawNext, and Google Scholar.  The second slide shows the top ten results for the terms and connectors search “right to receive information” in Lexis Advance, Casemaker, Bloomberg Law, and WestlawNext.  Cases highlighted in grey are relevant and unique, those highlighted in orange are relevant and appear in multiple databases, and those in white are irrelevant.