Hot Neuron has released Version 2.3 of its Clustify software, which features content-based e-mail threading to reduce the number of documents that must be reviewed during e-discovery.
The company explains that Clustify groups related documents into clusters and identifies a "representative document" for each cluster, allowing the user to review and categorize related documents together, for greater efficiency and consistency. The user specifies the desired relationship between the documents by selecting a similarity function. The similarity function might indicate that the documents should be conceptually similar, or that they should be near-duplicates.
Hot Neuron reports that Version 2.3 adds a similarity function aimed at grouping e-mails from the same thread together based on an analysis of the body of the e-mail, which is useful when headers aren't available. Clustify labels each cluster with descriptive keywords, providing a uniform interface for navigating the documents regardless of which similarity function the user selects.
Clustify can automatically categorize newly added documents by using the specified similarity function to compare the new documents to the ones that have already been categorized, sometimes referred to as "predictive coding." The new release adds more control over this process. Version 2.3 also integrates with newer versions of third-party tools and offers more options when exporting results to other systems, says the company.