Abstracts

Innovation Logic for new IP assignments 

Simon Dewulf, PatentInspiration and Imperial College

Increasingly, IP professionals engage in new types of assignments based on patent data. How can clear-cut algorithms in patent data open doors for IP professionals in marketing, innovation, research & development, human resources and even finance? Innovation methods help researchers to open their brainstorming approaches for successful idea generation within innovation projects. The ideas and associations made are, however, based on the knowledge of the individuals. Using the vast amount of data within the patent database can boost the outcomes tremendously. At Imperial College London, PhD research validated the ‘Patent Data Driven Innovation Logic’ approach using patent text mining techniques to provide valuable research input for technology-driven companies.


New Espacenet – advanced features

Heiko Wongel and Roland Feinäugle, EPO

The new Espacenet is not only a redesigned user interface; there is a completely new search engine and infrastructure behind it that offers users a range of new or improved functionalities. The enhanced result list, the dynamic query building and the filtering function, to name a few, enable the users to carry out more sophisticated searches and analysis than would be possible with classic Espacenet. This presentation will elaborate on some of the most prominent improvements of new Espacenet.


Patent insight reports from the EPO – future and emerging technologies

Nigel Clarke, EPO and Björn Jürgens, IDEA CITPIA PATLIB Centre Andalucia

During the last couple of years the EPO has started to produce and publish patent landscape reports. We call them “Patent Insight Reports” and we concentrate on future and emerging technologies.

  • Why do we do them the way we do them?
  • How do we choose the research topics?
  • Who are these reports aimed at?
  • Is there more to it than just the report?
  • What’s lined up for the future?


You’ll get the answers in this presentation.


Using patent information to deliver added value to innovators

Catriona Bruce, IP Australia

At IP Australia, we have been using our new IP Data Platform to produce interactive patent data visualisations and reports for international search applicants. This presentation shows how IP Australia is using its Patent Analytics Hub to add value for patent applicants, with examples of two recent trials of interactive visualisations. Firstly, the delivery of patent data analysis together with international-type novelty and inventive step reports at the time of filing a provisional application. Secondly, support of an Australian government initiative, the Entrepreneurs' Programme, by providing patent information as part of a business assistance package. Finally, the audience will be given a demonstration of the IP data platform, a new cloud data lab, and an opportunity to join its user community.


Latest renewable energy advances based on patent trends

Alessandra Salgado, IRENA Innovation & Technology Centre

This presentation provides an example of a successful service based on the EPO’s PATSTAT data. The online platform INSPIRE developed by the International Renewable Energy Agency (IRENA) provides up-to-date information on standards and patents for renewable energy. The INSPIRE platform uses the EPO’s PATSTAT Y02 Classification data as a source, helping users to easily search through, locate and perform analysis of more than two million patents applicable for renewable energy. The presentation will introduce INSPIRE applications, IRENA’s work on innovation and the role of patents in the renewable energy sector.


Latest information on IoT-related classification: ZIT towards G16Y

Toshiki Yamaguchi, Japan Patent Office

With the dramatic development of emerging technologies including the Internet of Things, IoT, there has been a growing public need to improve access to this patent information. Under such background, in November 2016, the JPO established the world's first patent classification for IoT-related technologies, "ZIT" as a Japanese broad facet classification code, and has been assigning it to patent applications filed to the JPO.

The IPC revision working group, held in Geneva in May 2019, approved the establishment of a new IPC sub-class "G16Y", which will enter into force in January 2020.

In this presentation, the latest information on these classifications will be provided.


Update on the CPC and how to be kept up-to-date with CPC revisions

Pierre Held, EPO

The Cooperative Patent Classification (CPC) entered into force on 1 January 2013. Over the years, the CPC scheme and its definitions have been improved to cater for changes in technology and make sure that searchers around the world can find the best prior art in the most efficient way. In this short presentation we will provide an update on the latest developments in the CPC and show you how you can follow CPC changes, e.g. CPC revisions on public platforms, e.g. the CPC website (www.cpcinfo.org).


CPC International – what's in it for the users?

Pierre Held, EPO

The Cooperative Patent Classification (CPC) entered into force on 1 January 2013. Six years down the line, the CPC has evolved from a bilateral EPO-USPTO initiative to an international classification standard used for classification by 29 offices worldwide. Adaptations to the CPC infrastructure and to the way CPC data from national offices are stored, displayed and searched were needed. This happened in the week-end of 24-25 August 2019 with the launch of the CPC International project. In this short presentation we will provide you with an overview of the major changes impacting users as a result of the switch-over to the new system.


The EPO's experience in machine learning /automation

José Alconchel y Ungría, EPO

Machine learning has been used at the EPO in three different aspects: pre-classification for routing of the patent applications to the right examiner, re-classification of documents after entering into force of a new CPC scheme and classification of patent documents in CPC. The experience in this area with the current and further developments will be presented.


Artificial intelligence and patent classification

Alexander Klenner-Bajaja, EPO

With increasing volumes of patent applications and also increasing volumes of relevant prior art in general, the EPO is looking into artificial intelligence driven automated classification approaches to handle this challenge not only in the future but already today.

The EPO has already today successfully implemented their own deep learning technology to automatically classify patents in the CPC system to ultimately enable a more efficient search. In this talk a brief overview will be given on where the EPO stands with its AI based classification, about the differences in manually and machine classified patents and the implied consequences and opportunities, the major challenges that have been identified for automatic patent classification and approaches to overcome them.


News on the European Patent Register - it happened because of you

Yolanda Sánchez García, EPO

Towards end of 2018 we carried out a survey among our users to ascertain their level of satisfaction with the European Patent Register and to identify any needs and wishes. This presentation will highlight the main survey results, and as the EPO's response, will underline some important future improvements, including new developments in the rapidly growing Federated Register service.


Expectations of a European Patent Attorney

Francis Leyder, European Patent Attorney

A frequent task of a European Patent Attorney is to draft a notice of opposition on the basis of prior art retrieved by a searcher. This presentation will share some expectations as to the search, and will provide views as to how prior art searches could better support oppositions at the European Patent Office, in particular for objecting a lack of inventive step.


Searching for perfect patent protection - a very personal story

Eric de Cock, Tata Steel

Generally a patent department mainly consists of patent searchers and patent attorneys. But these two groups don't always understand each other's perspectives. While their goal is the same: a perfect patent protection. In his personal story Eric de Cock takes you on his journey to strive for that perfection. He will take the audience from being a patent searcher to becoming a patent attorney and combining these two skills. His story will shed some light on practical experiences that may be helpful for both searchers and attorneys.


Categorisation of legal event data - how can it help the patent information user?

Christian Soltmann, EPO and Marjolaine Thulin, AWA Sweden AB

Patent information users are confronted with legal event data every day: For example when they assess the legal status of individual applications and granted IP rights. Or when they have to come to terms with large numbers of documents during freedom-to-operate (FTO) and monitoring searches.

Using the example of FTO searches, we will show the essential role high-quality legal event data plays for the efficient implementation of these searches and for the selection of potentially relevant documents. Special attention will be paid to how added value information of commercial service providers and recent achievements such as the WIPO ST.27 standard and the INPADOC classification scheme can help users to increase the efficiency and quality of their searches.


The QPIP certification, now finally going live!?

Frederic Baldour, on behalf of The International Standards Board for Qualified Patent Information Professionals (ISBQPIP)

Over the last year ISBQPIP has realised several major steps with the certification project. This presentation will review last year's achievements and update the audience with the expected next steps and calendar events. In particular addressing the organisation, the PER grant and coming exams.


Legal status information - beyond the patent offices

Stephen Adams, Magister Ltd

In the search for ever-more precise and comprehensive understanding of freedom to operate, patent information professionals are being asked more wide-ranging questions than ever before, about the thicket of intellectual property rights which must be overcome.  Information about actions in civil or administrative tribunals and courts, which may significantly affect the enforceable scope of a patent grant, is generally poorly integrated with the records about the same case held at the national intellectual property office.  This paper will discuss some of the hazards involved in trying to answer the apparently simple question, "Has this patent been challenged in court?".


Patent Early Warning System - more than an alert

Rosanna Lindquist, The Hershey Company

In today's 24-hour news and social media cycle, information is more prolific and viral than ever. This can lead to a long trail of reactionary business responses across business functions. Through the creation of a Patent Early Warning System, the Hershey Company has been able to anticipate industry activity ahead of the media and prepare a pro-active business response. This presentation will address the business case for creating an early warning system and best practices and tips for making this system work for your organisation.


Gold standard collections for training and evaluating machine-learning systems

Tony Trippe, Patinformatics

The use of machine-learning systems for patent searching is becoming more common place, and is being used more frequently in addition to more traditional searching methods. There are systems currently on the market which allow patent searching on different topics based on machine-learning algorithms. As a patent information professional, it can be difficult to evaluate the systems, or to compare them to one another.

One way of addressing this issue is to create gold standard collections for the training and evaluation of machine learning systems designed for searching. These collections will help patent searchers compare the results of different systems and determine if the precision and recall the systems provide will be sufficient for their day-to-day work.

In addition to covering the use of these collections this talk will also discuss the creation of them. If these collections are shared with the wider patent information community we can build a body of data and knowledge that will benefit the community as a whole.


AI, semantic and similarity searches, in the patent domain - performance, comparison, and analysis

Burkhard Schlechter, Austrian Patent Office

The results of patent searches performed using a number of different commercial and free-to-use semantic and similarity search engines were benchmarked against results from corresponding human, intellectually performed searches.

The individual strengths and weaknesses of each semantic or similarity search engine and the variations in search results associated with different technical fields and input requirements are compared with each other and against the benchmarks.

This presentation gives the results from, and conclusions about the performance and utility of the different search platforms.


Insights into performance evaluation and best practice when using semantic AI search tools; pitfalls and tricks

Linus Wretblad, Uppdragshuset / IPscreener

The semantic AI-text-based search methodologies adds a new option to the searcher’s toolbox for helping to explore the prior art. However, opening up the nice wrapping of promises and having this futuristic instrument at my hands, what do I do then? It seems that we don’t have much in the way of instruction manuals teaching us how to use it efficiently.

This presentation will take a glimpse under the hood and look into best practices. It will include a review of performance statistics, quality measurements and, last but not least, usage challenges. Finally, some personal thoughts on the future; what to expect next in the field of information retrieval and how the searcher’s might role evolve onwards.

.  

Quick Navigation