Summary of WolframAlpha & Legal Research

We’ve had a lively discussion of WolframAlpha & its implications for legal research. Thanks to everyone who has contributed. Comments can be viewed in the comments sections here and here. Please add your comments. Here is a summary of the responses as of 5-19-2009:

• Currently, WolframAlpha seems unable to answer legal questions, whether they concern identifying primary law or applying law to facts. Greg Lambert got similar results. By contrast, respecting four legal questions we posed to W/A, for each of the four questions Google returned documents containing correct answers within the top two items of its results lists. Wikipedia returned correct answers to only one of those four questions.
• Currently, WolframAlpha seems unable to answer most factual questions (even statistical questions) about U.S. courts or U.S. judges. See results from Susanna Leers and Stephanie Davidson. But it seems able to answer questions about law-related employment (such as lawyers’ salaries), as well as other labor or employment statistics. Scott Frey notes that W/A designates “Government” as a “future topic.”
• Currently, WolframAlpha can return certain financial data for companies in response to a search by ticker symbol or company name and certain natural language terms, but Ramona Martinez points out that it does not recognize all ticker symbols. Dan Giancaterino offers useful examples of successful corporate finance queries. Theresa Strike notes that currently W/A is best at quantitative queries and is fastidious about syntax.

Tags: Expert systems, Legal research, Search engines, Semantic search engines, WolframAlpha, WolframAlpha and legal research