Other Essential Tools

In the following sections, we explore additional tools that you should consider using (if you are not already) in your eDisclosure workflows today.

Clustering

Clustering technology is useful when you do not know what your data set contains. It groups similar documents together based on their content and is a form of unsupervised machine learning, which means it does not require manual review before documents are clustered. Clustering essentially provides the legal teams with a bird’s-eye view of the data, no matter how large.

According to Deloitte, it is estimated that clustering can lead to a 15 to 20 percent increase in review speed. This is significant, considering the extensive duration often associated with traditional linear document reviews.

Communications Analysis

Communications analysis uses AI to bring structure to the vast and complex world of communications. It identifies individuals, typically referred to as ‘entities’, and maps their communication patterns, grouping them based on their most frequent contacts. This process unveils intricate connections and relationships, a critical aspect in investigations and case management. Communication analysis is a handy tool for making sense of large amounts of communication and can help investigators understand the relationships and interactions between different entities, which identify potential leads or suspicious activities in their investigation.

Email Threading

Typically, emails include multiple replies or forwards when they are collected for review. Email threading is a method that identifies and groups together emails that form part of the same conversation. This in turn eliminates the need to review multiple emails conveying the same information. However, unique attachments that get dropped off the chain or communications that branch out to different people will still be made available for review with email threading. This is because the technology analyses conversations that diverge and attachments which change to ensure nothing important is missed during the review.

Entity Recognition

Entity Recognition is another useful tool in eDisclosure. It is a technique used to identify and classify named entities (such as names and organisations) within the text of documents. This helps to quickly locate documents involving certain people or organisations within your data set. For example, Entity Recognition can be used to automatically identify documents referring to the company “AI eDisclosure” within your data set which can then be prioritised for review.

Sentiment Analysis

Sentiment analysis is a tool that scores documents on the likelihood that they contain negativity, anger, desire, or other emotions. Through this analysis, you can quickly and easily locate documents that show unusual or highly charged interactions between participants. By detecting unusual communications between key individuals, you can locate communications that need further investigation and build deeper context around the conversations and ideas that are important to the case.

Machine Translation

Machine translation is a tool that automatically translates documents in foreign languages. This makes it easier to review documents from different jurisdictions without the need for human translators, saving both time and resources. This technology is especially beneficial in today’s world where cross-language communication is common as it helps to bridge the language gap and makes information accessible to all, regardless of the language they speak.

Anonymisation

Anonymisation tools are used to automatically detect and redact personal or sensitive information from documents. This is especially useful in complying with privacy laws/regulations and safeguarding sensitive data during the disclosure process.

Privilege Review

Privilege review is a process that uses AI to identify potentially privileged information in documents. Lighthouse claims that using AI in Privilege Review is 6.6 times more accurate than search terms alone. This significantly reduces the risk of mistakenly disclosing privileged information which is critical to maintaining the integrity of a case.

Incorporating Multiple AI Tools

You can incorporate more than one AI tool in your workflows. This should provide further efficiency on your project and improve accuracy. For example, clustering can be used to identify a selection of documents based on a specific concept or idea. Email threading can then be used on this cluster to thread the results. Finally, by setting up an Active Learning project on the threaded document population you can get to the documents which matter the most first. This approach has the potential to strengthen the power of each tool in combination and choosing AI tools that complement each other well can significantly enhance your document review workflows.