Home  »  AI-SDV  »  IC-SDV 2018  »  Programme  »  Tuesday 24 April 2018

Tuesday 24 April 2018

09:00 - 10:30

Chair: Harald Jenny, CENTREDOC, Switzerland

Machine Translation for patents. A User Centric Review

This talk will provide a brief history of machine translation (MT) technologies, from rules based, to statistical, through human assisted MT (HuMT) and neural MT.   Are machine translations capable of producing a set of documents suitable for search purposes?      Can readers of patent documents reach an understanding of the invention and the scope of protection using machine translation of patent documents?   Does this vary by technology?  By language (pair)?.   How has the existence of bulk machine translated collections influenced our approach to search and analysis of patent documents?   How are developers of MT collections managing their collections as MT itself improves?   How should we report results when MT documents are cited?  What is allowable for legal and prosecution purposes?.  This talk aims to provide a review of how users view machine translations of patents.   By presenting the current situation it will conclude by looking at future options for retrieval and review of patents in any language for legal, business and learning.

Jane List runs her own business, Extract Information Limited, which she founded in 2013.  Extract Information provides consultancy, search services, and training, all with a focus on commercial uses of patent information.  She is currently helping her clients manage their IP portfolios, build IP strategies and policies, file strong patents, understand competitor patents, understand the technology landscape, understand the market(s) for their technology, track their industry, find new customers and markets.  Jane has BSc. in Chemistry, MSc. in Information Science and Cert. in Intellectual Property Law. She is a Member of the Royal Society of Chemistry, BPIP, WON and PIUG.  In July 2014 she became co-Editor in Chief of World Patent Information Journal, and since September 2016 Editor in Chief.   In 2015 Jane co-founded the Cambridge Information and IP Meeting (CIIPM)  to host an annual meeting and to provide training in IP information and IP commercialisation matters.  Prior to Extract Information, Jane held product development roles in ProQuest Dialog, and Lighthouse IP Group, and she previously worked at The Technology Partnership, where she spent 12 years providing IP and information consulting for TTP’s clients in medical devices, printing, food industry, consumer products, drug delivery, clean energy, displays and more.  Prior to TTP Jane worked at DataStar, European Molecular Biology Laboratory, Wellcome PLC),  The British Library  and Thames Water.


Asian Patent Documentation: How EPO Examiners meet the Challenge

As we are all aware, Asian Patent filings are growing drastically.
This presentation will give a brief overview of the methodology (tools and trainings) used by the Examiners at the European Patent Office to meet the challenge of Asian Prior art.


10:30 - 11:00

Exhibition and Networking Break

11:00 - 12:30

Chair: Nils Newman, Search Technology/VantagePoint, USA

Do Mergers Stifle Innovation? – European Commission’s use of the Patent Asset Index™ in the Dow/DuPont Antitrust Case

The European Commission recently investigated the anti-competitive effects of the Dow/DuPont merger. Key to their analysis was how innovation competition might be affected if two important innovators come together in a concentrated market. A thorough patent analysis was carried out on the PatentSight business intelligence platform. The most valuable patents were identified using the Patent Asset Index™. The results showed a significant combined patent share of Dow and DuPont for the most valuable patents in crop protection. The European Commission thus decided that the merger of both companies be conditional on the divestiture of major parts of DuPont's global crop protection business.


Tale of patents filed in France in 1999

The patent applications filed in 1999 at the INPI can be comprehensively analysed for illustrating the most recent French patent story.

In this cohort, more than 7 out of 10 patent applications are granted. 9 years after their filing date (in 2008), half of these granted patents were still in force, and 16 years after their filing date (in 2015), there were about 20%.

Two main factors have a positive influence on the maintenance of these patents: patents that survived after litigation before French Courts and patents encompassing numerous claims.



A Manager’s Dream: End User Search in the Frontend but Expert Search in the Backend

One of the more recent management bandwagons is that because there is Google, decision makers don’t need expert searchers anymore.

But we can indeed provide decision makers with Google-like search engines, as long as we prequalify search results using the knowledge of expert searchers.

The ingredients? A platform to operate custom web search engines, some semantic technology and a business process.




12:30 - 14:00

Lunch, Exhibition and Networking

14:00 - 15:30

Chair: John Willmore, BizInt, USA

Linked Open EP data – a new Product from the EPO

The EPO has recently launched a new product: Linked open EP data. This open and free data set contains bibliographic information of EP publications and the Cooperative Patent Classification (CPC) hierarchy. Linked data , also known as  Semantic Web, facilitates combining a particular data set with other linked data sets in any domain including patents. Given its URI, data about a resource, e. g. a patent publication, can be retrieved in a variety of formats over the web. For occasional use there is a simple data browser, an API and a SPARQL query interface. For heavier use, bulk data is available for download.


In this presentation we will introduce this new EPO product and illustrate the different ways this data can be inspected and retrieved. We will explore the content and point out potential use scenarios.


Google’s BigQuery offers a new Way to access, explore and analyze public and private Patent Data

BigQuery allows researcher to easily combine data from multiple sources to gain new insights. A large collection of public patent data is available. Data is provided by Google, the USPTO, and EBI ChEMBL. Private collections are available through subscriptions from IFI CLAIMS and CPA Global. Private company information can be combined with public data.  A direct connection to Tableau enables visualization of BigQuery results. Examples will be provided that illustrate the unique nature of this new research tool. 



15:00 - 15:30

Exhibition and Networking Break

15:30 - 16:45

Chair: Christoph Haxel, Dr. Haxel CEM, Austria

Global Competitive Technology Intelligence – A case study

As access to the internet grows globally, the number of information sources are growing exponentially.  In addition, innovations driven by technology platforms such as IoT, AI, MEMS, etc. are bringing new competitors into spaces that were never considered before.  Whoever thought that an automotive OEM would feel threatened by Apple or Google.  This period of expansion in both information sources and technologies presents a tremendous opportunity and challenge for global corporations.  Information needs to be tracked on a regular and consistent basis from various (previously ignored) sources to identify new opportunities, innovations and threats. SciTech Patent Art will share its learnings from the establishment of one such well-thought through Competitive Technology Intelligence program for a global corporation.  The presenter will discuss how scope was established, approach used, learnings/refinements to make the program more useful and at the end, summarize key elements of establishing and making such a program successful.  Use of tools such as Artificial Intelligence, which is playing a critical role in program sustenance, will be reviewed in the context of establishing such a cost-effective global CTI program.



Rating of the technical quality of patented inventions opens the door to quality ranking of companies’ patent portfolios

For years, all the professionals in charge of valuing companies' patent portfolios have tried, with varying degrees of success, to identify patents that were based on inventions of high technical quality. Contrary to what is written in many publications, it is not the companies that file the most patent applications that are the most innovative. The challenge is to objectively measure the technical quality of a portfolio of patented inventions based on public information. To this end, the INPI has joined forces with the OECD, the OST and industry in different sectors to improve and standardize methods of using indicators predicting the technical quality of inventions. Our ultimate goal is to create an international standard. This methodology allows a very fine comparison of the quality of the patent portfolios of companies operating in the same sector and the identification of the most innovative companies. If a ranking is now possible, our objective is to allow the most innovative to be better developed especially in their communication, attract investors, reassure shareholders, validate the quality of the head of the patent department or to borrow money using their patents as collateral. Quality must take precedence over quantity.