Driverless cars Pat Base tutorial Making the most
Driverless cars: Pat. Base tutorial Making the most of Pat. Base functionalities for an effective Patent Landscape study
Summary This tutorial uses the topic of driverless cars to show the advanced features in Pat. Base can be used to create clean and reliable patent landscapes in an efficient manner, and to help users increase the power of their searching to gain actionable intelligence. The step-by-step approach taken in this analysis is applicable for general patent landscape searches • Prepare by studying some basic literature and patent searching with the emphasis of expanding query results • Expand general keywords and classification codes used for the specific technology • Generate statistical graphs and analyse the results Main Pat. Base features highlighted in this tutorial: • • Non-patent literature searching Non-Latin searching Pat. Base Thesaurus Citation searching Classification finder Optimise Pat. Base Analytics V 2 • • • Advanced keyword highlighting Semantic searching Use of the ”Similar” function to find similar documents
Step 1: Building a comprehensive search strategy This tutorial guide assumes some initial searching has been conducted and we pick up from that point. This guide aims to help information professionals expand on basic searches to form more expansive result sets, capable of mining as landscapes to provide trending and other forms of patent analysis. The start of this tutorial is having basic results in hand. Both query expansion and query refinement techniques are further discussed in this step. The Pat. Base features listed in the next slide allow information users to both narrow and expand their result sets. Expansion from existing initial searches to include more relevant recall, as well as narrowing overly broad results and eliminating erroneous hits. These functionalities help the user to input a high quality, comprehensive data set before proceeding to analysis, ensuring that a quality output is achieved.
Snapshot The Snapshot module is useful to drill down to specific areas within the search, or spot trends and run searches across all Pat. Base data on them. The user can click on an area of interest within Snapshot, here we use Chinese documents as an example. This is then combined with the original search query and run in Pat. Base. As shown below: This functionality assists information professionals in attaining the best data set with as few clicks as possible – quickly eliminating irrelevant information.
Snapshot • • Streamline your charts by grouping together assignees, inventors, jurisdictions & classification codes Ideal for cleaning up results sets before sharing charts/graphs with clients & decision makers Choose between bar chart, pie chart or table vew Select how many results you want to view Group jurisdictions, assignees, inventors or class codes Choose to display results in graph as number or percentage
Narrowing down your result set To make the first set as relevant as possible, the searcher limits the raw result set. By using the “Optimise” function to limit the result set to certain technical areas based on IPC codes, the set is limited by almost 40%.
Narrowing down your result set Pat. Base Analytics and the newly released Snapshot feature also provides the ability to visualise the top assignees, classifications, inventors and more. All items are capable of drilldown and bring back into the Pat. Base search history to further refine result sets. These can be noted and included in your final search strategy. Therefore, starting with a basic keyword search, Pat. Base Analytics can be utilised to identify additional keywords, classification codes and active assignees/inventors, thereby further empowering the searchability of worldwide patent data.
Keyword Analysis The searcher generates a more exhaustive overview of keywords by selecting the Keyword menu option within Pat. Base Analytics V 2. Using this approach, keywords identified can be brought back into the search history to be used in conjunction with initial strings. Even large sets (in this case almost 50, 000 records) are easily clustered and are visible within minutes.
Finding classification codes The Classification finder function offers a way to search classification definitions within Pat. Base using keywords. Single terms and phrases can be used as well as operators like: and, or, near. Classification codes can be taken into the Snapshot module for a quick analysis For each class, a detailed description is available, and can be used to decide whether it is relevant
Finding classification codes Pat. Base Analytics V 2 and Snapshot both have the functionality to identify relevant classification codes to supplement your search strategy. Easily switch to CPC classifications.
Non-Patent Literature An important source to find relevant background art is the use of the non-patent literature module. This can be a good source of additional keywords or to broaden the user’s knowledge of the subject area. Users can search multiple sources of publicly available non-patent literature.
Similar & semantic searching When a particularly relevant patent is identified, the user can take advantage of similar and semantic searching to find more patents with similar concepts. The user can copy sections from the abstract or full text of the document and paste in the semantic searching module. This will identify relevant concepts and suggest associated technology areas for the patent families identified; complimenting keyword and patent classification searches by locating additional pertinent documents. The “Similar” option can be used to identify patent families that are similar to the patent of interest based on their title & abstract, description, claims etc.
Advanced highlighting Keywords can be highlighted in the text in ten different colours and truncations (right and left hand), and Boolean operators can be used. An example is shown below where “autonomous”, “vehicle*” , “driverless*” and “sensors” are highlighted in the text. A corresponding hit map in full text documents allows the user to jump to sections of a document where each highlighted term is present. To enable a comprehensive and efficient review of a Pat. Base answer set, the advanced highlighting option is activated by clicking on the pen in the top right of the screen.
Pat. Base Thesaurus The Pat. Base Thesaurus identifies related keywords and synonyms which can be selected and, with or without modification, can be searched directly in Patbase. Users can take a term, whether it’s a gene or protein, chemical or general. Search string can be edited before being run in Pat. Base. For the driverless cars search, Pat. Base Thesaurus is used to find synonyms across multiple languages. This helps ensure that the search strategy is comprehensive and that patents written in non-English languages are identified.
Non-Latin searching The search query can be copied into the search screen of the Non-Latin search interface to automatically search the machine translations of the Non-Latin data in Pat. Base. Each term is translated separately in order to maximise the comprehensiveness of the translated search query. An example is given for “driver less car”. The translation tool automatically generates synonyms which can be added to your search strategy. The user can search machine translations of non-Latin data from China, Japan, Korea, Thailand Russia. Pat. Base uses self-learning systems for machine translations in cooperation with our partner RWS. In addition, using RWS’s worldwide expertise in the field of scientific and patent documents we are able to accurately translate names of patent assignees from non-Latin languages. Recently doubling the number of indexed Chinese patent assignee names to 40, 000!
Citation Explorer Citation Tree now displays forward and backward citations at the same time • Citations can be viewed side-by-side, visualised on a timeline or tree and relevant statistics revealed • Analysing patent citations is a great source of competitive & business intelligence, to see who is involved in inventing the critical technology and how the knowledge in the sector is building up – and where
Step 2: Visualisation of the Landscape Once compressive datasets are in hand, visualisation and analytical framework – included with all Pat. Base subscriptions – can be employed to show board-level ready presentations to a variety of audiences. The visualisations are only as good as the underlying data and can be useful in answering business questions. Critical business questions to address through Landscaping: • What are the growth trend and rates of the industry? • How do individual sub-categories of the industry differ? • Who are the key players and emerging and how do they compare? • Where are the key patent owners filing and protecting their IP? • What is the legal status of the patents in the landscape?
From Results to Analysis Incorporated Analytics module is accessible directly from the Search History page and from the user’s search results. The printable Analytics dashboard provides a summary of the analysis, with the ability to drill down further into any of the analysis options shown.
Is this a growing area of interest? Analysis can be done by year to analyse the rate of patent filings and grants. An effective method to see if the area has been becoming more crowded over the past couple of years or decades. Is this a growing area of interest? We can see in these graphs that the number of granted patents in this technology area has drastically increased in the past few years.
What are the current fields of interest? Analysing classification codes will identify which technology sectors are most common in your result set. These charts show that the technology area “Instruments” is most common, followed closely by ”Mechanical Engineering”. The bubble chart shows that the IPC code B 60 R has received a great deal of investment.
Content Analysis Various forms of text-mining creates interactive ways to drill into specific areas of the data, irrespective of classification systems, allowing other slices of the data for empowering information analysis. Such analysis is also useful in presenting data characteristics to various audiences.
Who are the key players? The raw data can be drilled down further. Here, the user will search families for the assignee Magna Electronics Inc. in Pat. Base. Multiple analysis by assignee options are available. In this instance, unsurprisingly Waymo (Google’s self-driving car development company) is featured in the top 3 assignees.
How are the key players trending? Overall increasing trends and grant success rates show investment in driver less cars to be strong and growing. R&D expenditure continues to flow into the area and the number of jurisdictions and assignees becoming engaged is further evidence of growth.
Where are the key players filing? Analyse which jurisdictions are being filed in most frequently to help direct your own patenting decisions. In this example, the user identifies the US, Germany, WIPO, EPO, Japan and China as the most popular jurisdictions.
Drill into raw data instantly All data can be easily exported from the Analytics module and brought into Excel for further analysis. Chart Title 160 140 120 100 80 60 40 2018 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 Series 1 Series 2 Series 3 Series 4 Series 5
Conclusion • Using a high quality patent database and an effective means of expanding and narrowing results sets helps users to build landscapes consisting of highly relevant visual output capable of answering business critical questions Coverage • With integrated landscaping tools available for no additional cost* in every professional version of Pat. Base, users can interact with data in more ways than ever • Data Quality processed in house • Ongoing work on and projects on Assignee names • Standardised and de-duplicated • The experienced Minesoft Customer Support teams are ready to help your landscaping needs • 105 Issuing Authorities & 50 Full Text Countries • National Patent Office Legal Status Feeds • Complementary Data Quality Sources • Over 30 different sources • Exclusive to Pat. Base • New sources being processed every year *All tools, analysis functionalities and data shown in this tutorial guide are included in every full Pat. Base subscription
Further steps… If an assignee of interest is discovered, the user can dig deeper by generating an Assignee Snapshot ‘report’. When viewing a family from your result set, click the question mark icon next to the Probable Assignee. A new window will pop up with assignee, agent, inventor and examiner details for each publication within the selected family. Click on the bar chart icon to load an Assignee Snapshot ‘report’.
Assignee Snapshot A printable report is loaded, with summary, jurisdiction, and application/grant information about the selected assignee. All the graphs can be quickly expanded and/or downloaded. A quick and easy way to find more information about a commonly recurring assignee or a notably absent one.
- Slides: 28