When computer-assisted legal research first became possible, scholars and researchers carefully constructed their searches, knowing that their results relied on the search engine matching their requested search terms to the actual words within the database being searched. As different people used different words for the same concept (e.g. attorney, counsel, lawyer), the search string was critical to results. To help with producing more reliable outcomes, information providers offered options to search using controlled language tools (e.g., subject headings, key number system), bringing consistency across sources.
With supercomputers now able to mine large amounts of uncontrolled data for patterns, the current generation of search engines no longer requires careful word choice or indexing. While they may provide an option for the user to revert to Boolean searching, their default interfaces allow a user to type in questions, phrases, or terms for interpretation by the engines. They then produce search results through proprietary (and largely hidden) algorithms and relevancy definitions. Researchers may or may not understand what they are searching or how the relevant documents were selected. This program will set forth the current state of search algorithms and explore the benefits and concerns surrounding them.
Business Meeting at Program Conclusion.