West editorial, indexing expertise surpercharges law firm documents
CaRE for West km 2.0 has been pretrained on 20 million headnotes written and classified to the West Key Number System by West’s highly experienced attorney-editors – to sort documents against these legal topics. CaRE incorporates a suite of classification programs capable of learning how to classify enormous document collections to complex category systems based on fine distinctions. For example, CaRE can distinguish between “search and seizure of a vehicle” versus “search and seizure of a computer.” CaRE also can classify documents to more than one category or to no category, if the document is not actually classifiable to the taxonomy.
Together, KeySearch and CaRE are said to offer search results with unprecedented accuracy; achieved without the need for law firms to invest large amounts of attorney time or staff resources to code and catalog documents. “This technology literally exports West’s 125 years of expertise at indexing legal information to a legal organization’s document management system,” said Peter Jackson, vice president of Research and Development at Thomson Legal & Regulatory. Jackson is a world-renowned expert on artificial intelligence and information architecture. “No one else can provide such a capability, because no one else has the data to train the system. Furthermore, no one else has scaled automatic text categorization to cope with the task of routing millions of documents to hundreds of thousands of categories.”
To help technologists, legal professionals and the public better understand the underlying technologies underlying CaRE, West has released West km 2.0, Classifying Document Collections with CaRE, a white paper written by Daniel Dabney, Peter Jackson and Kingsley Martin. Also available is A Brief Practical Introduction to Taxonomies, a white paper written by Daniel Dabney. The white papers also can be found at west.thomson.com/westkm.
Contact West. Web: west.thomson.com
Copyright Research Information Ltd. 2003
Provided by ProQuest Information and Learning Company. All rights Reserved