From the onset of the digital era, keyword search has been our go-to for finding relevant information in documents, emails and other unstructured text. But let’s face it, keyword searches often miss the mark. They typically return too many irrelevant documents, inflating review time and costs, and they miss a lot of crucial documents, posing a serious issue in discovery.
Enter Sherlock AI, the machine learning algorithms powering DiscoveryPartner and our answer to these problems. By combining keyword and algorithmic search into a single platform, Sherlock AI allows anyone to surface key information without having to use complex keyword search syntax.
What is Search 2.0?
Search 2.0 is a groundbreaking approach to finding information in extensive data collections. By seamlessly integrating AI with keyword search, you can find what you need in large document populations–quickly and easily–without having to master complex keyword search syntax.
What sets Search 2.0 apart is its ability to surface highly relevant documents, even those that might elude traditional keyword searches. At the same time, you can efficiently weed out irrelevant documents with a simple Thumbs Down, saving time and costs looking at unimportant data.
See Search 2.0 in Action
Watch this video to see how Search 2.0 allows you to find great documents quickly and easily.
How Search 2.0 Works: A Closer Look
Let’s take a closer look at how Search 2.0 works. The screenshots below are from our demo site, featuring a collection of over 2.7 million documents, primarily emails from Jeb Bush’s two terms as governor. The topic involves a disputed claim regarding the excessive extraction of water by two commercial bottling companies, leading to the depletion of a crucial resource in Florida.
The request (RFP) was this:
Bottled Water — All documents concerning the extraction of water in Florida for bottling by commercial enterprises.
It was created by NIST (National Institute of Standards and Technology) for their annual TREC (Text Retrieval) Conference.
When approaching this exercise, there are various methods one could employ. In a Search 1.0 world, the most obvious choice would be to utilize Boolean syntax for a keyword search. However, we propose an alternative approach, recognizing that the complexities of Boolean syntax can be daunting for many users.
Use Freeform Search Instead of Complex Boolean Search
DiscoveryPartner offers a unique freeform search capability to get going. Simply type in the text of your request, including punctuation. You can enter every term you can think of that might be associated with your topic. The only requirement is to enclose your request within square brackets.
Sherlock AI will intelligently process your input to uncover the most relevant documents that contain the highest number of terms, overlooking the absence of specific syntax. You can sort by relevance (high to low) to see the most likely relevant documents first.
Gain Rapid Insights in Snippet View
DiscoveryPartner’s unique snippet view of results is useful when sorting the results by relevance. The snippet view presents extracted text alongside relevant fields in a quick overview of your search results, allowing you to gather insights before proceeding further. By examining the extracted text snippets, you can swiftly grasp the context and relevance of each document so you can make informed decisions on the next steps to take. Snippet view streamlines the evaluation of search results, saving you time and effort.
Preview Document Texts to Ensure You're on the Right Track
From the snippet view you can see the text of individual documents. This helps ensure you are on the right track. Click on the record itself to view the individual document and its associated tags and metadata.
Select and Send Multiple Documents to Sherlock AI
Scroll down the page to review and select relevant documents from your search. You can folder the documents or send them directly to SherlockAI to find more. You have the option for Sherlock AI to place relevant (Thumbs Up) documents in a selected folder.
Review New Documents Surfaced by Sherlock AI
In milliseconds, Sherlock AI will analyze the document you sent, build an AI model around key terms and apply that model to the other documents on your site. It will then rank the documents by relevance, bringing back the next most likely relevant document with key terms highlighted.
Click Thumbs Up for a Relevant Document, Thumbs Down If It's Not
Take a look at the document from Sherlock AI. If it is relevant to your inquiry, give it a Thumbs Up. If not, give it a Thumbs Down to help Sherlock AI get on the right track. It will then rank the documents by relevance, bringing back the next most likely relevant document with key terms highlighted.
A simple Thumbs Down weed outs irrelevant documents, saving you time and costs.
See Key Terms Important to Your Search
DiscoveryPartner provides key terms that may be important to your search. You can view them in a Word Cloud. The more important terms are larger in size.
Or view key terms in list format. Boost terms you think are especially important, suppress terms that Sherlock AI should ignore. Sherlock AI takes them into account when you next send a document judgement.
Continue Until You Find What You Need
As you keep judging documents, Sherlock AI gets smarter. Continue with Sherlock AI until you find what you need, or he stops finding relevant documents. Progress is tracked at the top of the screen. You can stop at any time and return to the session later.
Sherlock Keeps Getting Smarter
As the review progresses, Sherlock gets smarter about the information you seek. You can see the key terms evolve as you continue your review. Continue reviewing document by document, or use use our unique Cluster Batching feature.
Enhance Efficiency with Cluster Batching
You can enhance review efficiency by using Sherlock AI’s unique Cluster Batching algorithm. Rather than review documents one by one, ask Sherlock to send the next 50 documents in ranked order for your review.
Sherlock AI will group results in individual clusters of documents with similar content. You can then review documents individually or tag them by cluster.
See DiscoveryPartner and Sherlock AI in Action
Want to see a few more examples? Watch as Sherlock AI tackles two of his toughest cases.
Back in the early 2000’s citizens of Florida were concerned about felon voting rights. Can Search 2.0 help us find documents on this thorny issue? Let’s find out.
Climate change was an important issue during Jeb Bush’s terms as Governor of Florida. Can Search 2.0 help us find documents about this issue? Watch and see.