5.3 Result list

The result list in INTERGATOR Patent Search is a central tool for the structured analysis and evaluation of patent information. It not only displays the semantically most relevant results for a search query, but also provides numerous functions for interaction, sorting, filtering and further processing.

The result list is therefore not merely a results display. It also serves as a starting point for further AI analyses, relevance assessments, re-ranking steps and search reports. From the result list, relevant documents can be identified, saved to lists, combined with additional filters or prepared for later evaluations.

Each result is displayed with the following information:

• Title of the patent and a short technical context from the description (TECHNICAL FIELD)
• Publication numbers (e.g. CN113308548B), linked to the detail view or the original document
• Application and priority dates, including legal status (e.g. GRANTED, ALIVE, PENDING)
• IPC main class for thematic classification, including technology domain where assigned
• Applicant and inventor for quick identification
• Thumbnail preview image for technical drawings where available
• Position in the result list as a sequential number

By default, the results are ranked by relevance, which is calculated using an AI-supported semantic model. A numerical relevance score is not displayed – instead, relevance is indicated by the position in the list and the content proximity to the search context.

The result position is therefore an initial indication of semantic proximity to the search query. For a technical evaluation, however, a result should always be reviewed in the context of the specific question, the relevant technical features, the text references and, where applicable, the reference document used as the starting point of the search.

Documents can be selected from the result list for further work steps. These include in particular:

• the detailed review of individual patent documents in the document view,
• the creation of semantic document filters,
• the preparation of a relevance analysis,
• support for a validity analysis,
• the gradual refinement and prioritization of results through re-ranking,
• as well as the compilation of relevant documents for a search report.

A reference document is not created by simply marking a document in the result list. In typical assistant-supported workflows, the search is started via an assistant, for example by entering a publication number. The corresponding patent document then serves as the source or reference document. On this basis, INTERGATOR Patent Search generates a more complex search query with semantic filters, document reference and additional search criteria.

Certain functions are only available if the required context exists. For example, some analysis or re-ranking functions require a source or reference document, a query generated by an assistant or an active document filter. If this context is missing, the function cannot be executed meaningfully or must first be prepared by selecting a suitable search workflow.

Various icons are available to the right of each result:

• Add to watchlist – adds the document to the personal watchlist
• Add to favorites – marks the result for later analyses and comparisons
• Hide – temporarily removes the result from the current list (can be restored via the “Hidden results” area)
• More options – opens a context menu with:
  • similarity search (“Similar Documents”)
  • filter creation based on the document
  • PDF download (where available)
  • copying the publication number

Depending on the configuration and current search context, additional functions may be available in the context menu, for example to start specialized search assistants, use the document as the basis for a semantic filter or prepare an AI-supported analysis. Context-dependent functions usually refer to the currently selected result or to an already active document filter.

Tip: Many icons provide mouse-over texts (tooltips) that explain the respective function.

In assistant-supported workflows, a reference document is typically defined at the start of the search. For example, in a validity search, opposition search, FTO search or monitoring search, a publication number can be entered. The corresponding patent document then becomes the source or reference document for the generated query.

Independently of this, relevant results in the result list can be used as the basis for further searching or analysis. If a result is added as a document filter or used for a similarity search, its technical content serves as a reference point for further processing.

This is particularly helpful when:

• a result is technically very close to the original question,
• similar documents are to be found,
• the result list is to be focused more strongly on a technical core,
• a relevance analysis is to be performed with reference to a specific document,
• or a validity analysis requires a specific IP right or reference document.

A document filter can be created directly from a result. It can be added actively to the search or initially created in a disabled state in order to collect several potentially relevant results and later include them selectively in the search strategy.

The initial sorting of the result list provides a first semantic prioritization. During the further search process, this prioritization can be refined using additional information.

Relevance analysis and re-ranking help to gradually condense a larger result set down to the technically most important documents. The following aspects can be taken into account, among others:

• semantic proximity to the search query,
• similarity to a source or reference document,
• technical features and feature groups,
• overlaps with an initial description or reference document,
• relevant text passages and references,
• bibliographic criteria such as priority dates, applicants, IPC classes or status,
• as well as manually set filters and the technical assessment of the searcher.

Re-ranking does not replace the technical review by the user. However, it helps to identify promising results more quickly, hide irrelevant documents and focus the further analysis more precisely.

The result list can also serve as the basis for later documentation of search results. Relevant results can be marked, saved, organized in lists and prepared for reports.

A search report can combine, for example, the following information from the result evaluation:

• the initial question and search strategy,
• filters and reference documents used,
• relevant results and their bibliographic information,
• technical summaries,
• text references and reasons for relevance,
• feature overlaps,
• as well as notes on further review or analysis steps.

This means that the result list supports not only the review of individual results, but also the traceable documentation of the entire search process.

The result set can be narrowed down in a targeted way using the left and right sidebars:

• filters by country of application, status, applicant, IPC class
• multiple filters can be selected
• clicking on a filter value immediately activates it

For better clarity, the result list displays only one main document per patent family by default. The option “Show duplicates” can be used to additionally display related family documents, such as international publications.

Filtering is an important part of the step-by-step result evaluation. It can be used to narrow down the result list before an analysis or report is created, for example by relevant countries, periods, applicants, technical classes or legal statuses.

By clicking bibliographic fields such as IPC class or applicant name, a filter can be created directly. A right-click provides the option to create a disabled filter that can be activated later if needed.

This interactive filter creation supports an iterative search process. During review, relevant patterns can be identified and immediately transferred into the search strategy without rebuilding the search from scratch.

• Clicking the publication number opens the document in the detail view
• Displaying the map (semantic context visualization) is optional using the “Show map” switch
• The page navigation above the result list allows browsing through larger result sets

This display helps you quickly understand relevant results, select them in a targeted way and continue processing them in a structured manner. The result list therefore forms the transition from search to analysis: it makes relevant documents visible, enables their technical prioritization and creates the basis for relevance analysis, re-ranking and documentable search results.