Similarity Density Scatter Plot Landscape
Date: 2023-10-24
How to use this report:
Below is the Similarity Landscape for the patent document domain collection. Click HERE to jump directly to the interactive Similarity Landscape analysis display.
This Similarity Landscape presents an interactive environment for visualizing and analyzing the relative similarities of the patent documents in the domain collection.
The interactive environment enables capabilities to browse the landscape with mouse-over displays of patent document publication numbers and titles. These publication numbers are hyperlinked to the complete patent documents available on Google Patents .
The Similarity Landscape can be analyzed in more detail via keyword and key phrase queries as well as via the identification of patent documents by publication number and by assignees. Patent documents matching queries are highlighted on the Similarity Landscape display. These type of queries can be instrumental for various IP research tasks, such as gaining an overall understanding of the crowded or sparse areas of the landscape, or in determining similar nearby patent documents to a specific patent or application, or even identifying patent documents from a specific assignee. In addition, these capabilities can also be used to gather intelligence for Freedom-to-Operate (FTO) research, claims drafting, and R&D strategy. For more information see the report entitled “Leveraging Data Science to Capitalize on the Strategic Value of Intellectual Property.”
FTO research, in particular, can leverage the Similarity Landscape and keyword / key phrase queries to identify similar patent documents that may contain overlapping or equivalent claims. Queries can be executed specifically on the claims of U.S. patent documents contained in the domain collection. These documents can then be examined for overlapping or equivalent claim context. In addition, the Similarity Landscape can then be used to explore the documents that are in close proximity, but maybe not identified as keyword / key phrase matches, to those documents that are keyword / key phrases matches. These documents may also contain potentially overlapping or equivalent claims. This search strategy can be broadened via keyword / key phrase queries into titles and abstracts in addition to claims. This will likely identify a larger set of patent documents, that includes more international filings, that could contain comparable claims. Again, the Similarity Landscape can then be used to examine the documents in close proximity to these identified patent documents, as these close-proximity documents may also have overlapping or equivalent claims. This FTO research strategy thus results in the ability to identify and analyze the patent documents and claims that are the most relevant to the inquiry at hand, be that a focused FTO search, gaining insight into drafting claims to avoid existing IP, or gathering intelligence to craft R&D strategy.
The patent document similarity data used in this analysis comes from Google’s BigQuery “google_patents_research.publications” dataset. This Similarity Landscape analysis tool is designed to accompany the Alpha Data IQ Patent Landscape Analytics Report that covers the same patent document domain collection.
The application of data science and analytics to patent data enables these reports to be produced quickly and economically. Most initial analyses can be executed for under $2,500. Contact me if you need a custom report. https://www.alphadataiq.com/index.php/contact-alphadataiq
This similarity scatter plot shows the patent documents of the domain collection on a landscape with relative position representing the similarity of the patent documents. Dots that are closer together represent documents that are similar, whereas dots further apart are less similar. The similarity data used for this display is retrieved from Google’s BigQuery “google_patents_research.publications” dataset. This tool is designed to accompany the Alpha Data IQ patent landscape analytics report.
Mouse-over functionality shows the hyperlinked patent document publication number and the patent document title. Clicking the hyperlinked publication number will open a new browser window that shows the entire patent document on the Google Patents website. Zoom functionality is executed via “pinch-to-zoom” on mobile devices and the scroll wheel on desktop environments. Pan functionality is executed via simple drag and drop techniques. In crowded similarity areas, it helps to zoom prior to accessing the hyperlink to the full patent document.
Scatter plot display settings are available on the Settings tab. The Similarity Landscape can be queried via keywords and key phrases into documents’ titles, abstracts, and claims via the Keyword Search tab. Patent documents can be further identified by publication number and assignees via the ID Patent Docs tab.
This adjusts the granularity of the density lines display on the landscape. A higher density granularity will help identify the areas representing the macro topics of the similarity landscape. Whereas a lower density granularity will better highlight more detailed similarity clusters.
Adjusting the density granularity will also reset the zoom and pan. So, the best practice is to adjust density granularity before executing zoom or pan.
Enter keywords and/or phrases:
- Keywords and phrases must be separated by commas.
- Enclose phrases in double quotes.
- If a comma is needed along with a keyword or phrase, enclose the word or phrase and the comma all together in double quotes.
- If a double quote is needed along with a keyword or phrase, represent the double quote character as \\", i.e. double quote preceeded by two back slashes.
- Example: keyword1, "multi-word phrase", "inclusion, of a comma", "inclusion of a double \\"quote\\""
- 2000 characters max input.
- Claims text search is only available for USPTO patent documents.
- Matching documents must include all keywords and keyphrases in the selected section(s).
- Matching documents are indicated as black-colored document indicator dots on landscape display.
- Used to identify specific patent documents via publication number(s) and assignee name(s).
- These color indicators take priority over the black colored keyword or key phrase matching indicators.
- Enter full or partial patent document publication numbers separated by commas. 2000 character max input.
- Indicated as orange-colored document indicator dots on landscape display.
Patent document assignee(s):
- Enter full or partial assignee name(s) separated by commas. 2000 character max input.
- The inventor's names' will often also be indicated as initial assignees. So, some discrepancies with Google may exist.
- See the Patent Landscape Analytics Report for more info on assignee names.
- Indicated as purple-colored document indicator dots on landscape display.