US 8239358

US Patent 8239358, titled "System, method, and user interface for a search engine based on multi-document summarization," was invented by Dmitri Soubbotin. The patent was filed on January 30, 2008, and issued on August 7, 2012. The current assignee, as of March 24, 2026, is Semantic Engines LLC.

The abstract describes a method for searching multiple documents on a computer system. It involves sending a query to a system core, which passes it to a search component. The system core receives search results (related documents) and sends a specified number of these to a summarization component. The summarization component then processes these documents to create a multi-document summary. This summary is then received and displayed by the system core.

A search of CAFC 2026 dockets for patent 8239358 did not yield any specific results indicating litigation or scheduled cases for this patent in 2026.

Here is a plain-language overview of the independent claims:

• Independent Claim 1 (Method Claim): This claim describes a method where a user's query is sent to a system. The system uses a search component to find relevant documents and then sends a select number of these documents to a summarization component. This summarization component creates a multi-document summary, or "digest," based on the query's topic by analyzing the original documents and extracting sentences and text fragments. Importantly, each extracted sentence or fragment in the summary includes a clickable link back to its original source document. This summary is then displayed to the user as the main search result, alongside the traditional list of search results. The system also shows key semantic concepts from the documents. The user can extensively interact with this presentation, including refining their search, navigating parts of the summary, accessing the original documents via the embedded links or the search results list, requesting new summaries with specific content exclusions or inclusions, customizing how source links are shown, rearranging summary content, saving the summary, and providing feedback (rating) to the system.

• Independent Claim 5 (Method Claim): This claim outlines a method where a user sends a query and, in response, directly receives a multi-document summary from a system core. This summary is presented as the primary search output and is a "digest" focused on the query's topic. It comprises combined sentences and text fragments extracted from the relevant documents, with each fragment including a reference link to its source. Alongside this summary, the system also displays the conventional search results and semantic concepts. Similar to Claim 1, the user is provided with extensive interactive capabilities, such as refining the query, navigating the summary, viewing original documents through the embedded links or the search results, requesting further summaries with specific content adjustments, choosing display preferences for links, modifying the summary's arrangement, saving the summary, and rating it to send feedback to the system.

• Independent Claim 7 (System Claim): This claim describes a system designed to perform the functions outlined in the method claims. It includes components (means) for:

  • Allowing a user to interact with the system.
  • Searching documents based on a user's query.
  • Generating a multi-document summary (a digest) that contains sentences and text fragments extracted from the search results, with each fragment having an attached reference link to its source document.
  • Presenting this multi-document summary to the user as the primary result of their query.
  • Also presenting the traditional search results from the search component alongside the summary.
  • Extracting and displaying semantic concepts to the user.
  • Providing a wide range of user interaction capabilities, mirroring those described in Claim 1 and Claim 5, such as refining queries, navigating the summary, accessing source documents, requesting modified summaries, customizing link displays, modifying summary content, saving the summary, and rating it for feedback to the summary generation component.