Easy discovery of suitable patents using semantic search and 100+ searchable fields inlcluding prosecution/appeal histories in Japan, making full use of machine learning technology without the need for expert knowledge.
Search results contain aggregate results of patent attribute information such as applicants and patent classifications on sidebar, and you can easily add filter conditions simply by clicking on them.
You can search with logical set operations as with general search engines.
In the main search field, you can :
AND search by separating the queries with a space such as 'neural network',
OR search by connecting "OR" between queries such as 'neural OR network',
NOT search by connecting "-" between queries such as 'neural -network',
Phrease search by enclosing multiple words separated by space with double quotation marks(") such as '"neural network"'.
In Advanced Search's Filters(Details), by default the space separator operates as the OR search (including any). You can change the behavior of the space separator with the right select box.
You can search documents where multiple words appear close together.
In the main search field, you can
Near search by inputting '*Nx"word1 word2 ... word N"' (x is numerical value) such as '*N5"neural network"'.
You can filter by how many additional words can appear in the search results.
In Advanced Search's Filters(Details),
you can Near search
to select "Including within x" of the right select box and enter a numerical value in the x field and input multiple words separated by blank.
Near search can not be used for any search fields that do not support near search, such as application number and filing date.
Fuzzy search can return results where some characters may be different from the search query. For example, if you enter "GOOGLE" for applicants of US, you can also see results with misspellings such as GOOGEL OR GOOGOLE.
Fuzzy search is available for some search fields in the Advanced Search's Filters(Details). adding a filter condition, to select "Fuzzy match" of the right select box and entering a numerical value up to 5 in the x field you can hit attributes whose changed within x characters from search query.
Similarity search with the potential semantic meaning based on machine learning technology. For example, the keywords "SOLAR PANELS" and "PHOTOVOLTAIC PANELS" have similar potential semantic meanings. Searching with a keyword "SOLAR PANELS" or sentence including the keyword will return similar concept documents including "SOLAR PANELS" or "PHOTOVOLTAIC PANELS".
By specifying the patent or publication number, it is possible to search similar concept documents. You can also semantic search by clicking the link of the search result list or details page.
If your search is not returning relevant results, please input exact keywords appearing in documents to second field of Advanced Search.
You can extract similar concept keywords by clicking button in Advanced Search after inputting the keyword in the main search field.
When you click on the similar keyword, the selected keyword is added to the end of the main search field.
When you select a cross tabulation field next to the graph displayed in the search result, the top 10 items of the two selection items are extracted. Easily make various charts such as bubble charts and bar charts with cross tabulation between the top 10 results
It is also possible to compare with specific values instead of the top 10 by specifying the search conditions in the Advanced Search -> Cross Tabulation
Patents can be evaluated for strength and value depending on the number of times cited in other patent applications.
Evaluation can be customized with a choice of 40+ conditions including prosecution/appeal histories (JPO only) from the score tab in Advanced Search.
By visualizing citing/cited quotations of individual patents, it is easy to find basic patents and technologies.
Just by clicking a node, you can trace citing/cited relationships with children -> grandchildren more and more. Double click to jump to the detail page of the corresponding node.
Applicants/Assignee, Agents/Law firms and Inventors Ranking. An objective analysis of innovator/practitioner performance based on complex big-data analytics.
It can be reflected patent scoring and various search methods as the above-mentioned to the rankings.