Stephen Boyer, IBM Almaden Research Center
Without order, a collection is just a heap of stuff. For centuries, mankind has been working on bringing order to collections in the earliest libraries and archives. In the same way, without order, databases are just random bits and bytes. This keynote will take a look at how we seek to improve order, in databases and in searching. We are continuously challenged by factors that disturb the order, such as the growing amount of data, globalisation, and obfuscation. Integration of big data will demand new search technologies. We’ll take a look at promising developments in patent search with examples from different technical fields. We'll discuss the role of computer curation to enhance the ability to organise, to search content, and ultimately to lead to predictive analysis. These developments will be viewed in the context of the value of intellectual property data to the scientific community at large.
Klaus Baumeister, EPO
The European Patent Office publishes more than 60 000 pages of patent applications and specifications every week. These EP patent documents are made available digitally in two flavours: in PDF/A and in XML. The PDF format is for easy readability and intended to be used for display on computer screens and for printing on paper. The XML format is for machine-to-machine communication, input for search engines and automatic processing purposes like translation systems. The presentation will explain the different types of patent documents, show updated statistics on the publications issued and look into the XML coding of complex structures like tables and mathematical formulas. It will also enlighten some of the difficulties to be solved during the production process of EP patent documents and explain how the EPO is planning to improve the editorial procedures to avoid misinterpretations in textual information of patent specifications.
Sabine Kruspig, EPO
On the one hand, business methods are excluded from patentability if claimed as such (Art. 52 (2,3) EPC; Rules 39 and 67, Regulations under the PCT). On the other hand, technologies in the ICT (information and communication technology) industry become key enablers for new business environments. A short historical overview on “business methods” in the IPC, the requirements of the EPC and PCT with respect to search documentation and the patentability of business method related patent applications will be given. The question of what is searched and where it can be searched remains a fascinating challenge within the legal frameworks, particularly in the area of computer-implemented business methods.
Marjolaine Thulin, AWAPATENT
Finding out the status of European patents after grant has become easier and easier over the years thanks to improvements in the European Patent Register. It is however still time consuming as the number of member and extension states increases and extra checks need to be conducted. And what about European patent with unitary effect, set to arrive on the scene very soon? At the EPO Patent Information Conference in Stockholm in 2008, a wish list concerning the legal status information on European patents after grant was presented by this speaker, as a dream she had. This year the EPO is introducing the longed-for Federated Register for granted EP patents. This presentation will try to answer whether the speaker’s dream has finally come true.
Monika Hanelt (Agfa Graphics), Patent Documentation Group (PDG) liaison to BUSINESSEUROPE
This presentation will look at the passive part of the EPO's new Global Dossier which is integrated into the European Patent Register and is a brilliant resource for tracking the patent applications filed within the five biggest patent offices. Global Dossier came about in the framework of discussions between the IP5 Offices (EPO, USPTO, JPO, KIPO, SIPO) and IP5 industry representatives.
There is also an active aspect to the Global Dossier in planning, which this presentation will describe, highlighting the importance of continued co-operation between the IP5 Offices and IP5 industry in this area. BUSINESSEUROPE (BE), the leading voice for all-sized companies in 35 European countries, officially represents the user community in Europe. PDG and BE have agreed to co-operate when it comes to patent information, PDG companies being member of BE via the national business federations. IP5 industry proposed a prioritised list of users' wishes at the Global Dossier Task Force Meeting in January 2015, including legal status data, alerting and standardisation of patent assignee names.
Marios Sideris, Pierre Held, Frédérique Lequeux, EPO
This talk will present an update on the newest developments in the Cooperative Patent Classification (CPC) system. It will include topics like the expansion of CPC internationally, the format and the presentation of the CPC scheme and definitions data, the frequency of publication, CPC data coverage etc. as well as future developments in the area. Finally, it will take a look at certain special features of CPC such as combination sets, which are of interest for the search experts.
Bettina de Jong, Shell, representing CEPIUG
The idea of certification of patent information professionals was born at the EPO's conference in Copenhagen in 2002. Now, with this conference again being held in Copenhagen it is time to look back and celebrate what has been achieved so far. The draft Articles and Rules have been shared with the community and when these are finalised, the certification organisation can be formally established. We are not there yet, but the examination and certification is coming within reach. On the 13th anniversary of the conception of certification we will present the milestones achieved so far and the next steps to its realisation.
Roger J Burt
Who owns the world’s patents? The answer is that no-one can be sure. This gives rise to an issue of significant economic importance. Innovation lies at the heart of technological progress, and patents sit at the epicentre of the global reward and recognition system. Without unambiguous ownership information, the full value of intellectual property – an emerging asset class - will not be realised. ORoPO is a not-for-profit open register of patent ownership information. It sets out to complement the work of Patent Offices worldwide, and is a practical solution that addresses the fundamental problem of accuracy and accessibility of data around patent ownership.
Julie Callaert, KU Leuven – ECOOM (Centre of Expertise for R&D Monitoring)
Anyone who has ever extracted patent portfolios from single organisations has likely been confronted with problems arising from the presence of multiple name variants in patent databases. To facilitate extraction of patent data at the institutional level and to avoid the prospect of creating a puzzle with missing pieces due to undiscovered name variants, different methods have been developed for automated harmonisation of applicant names in large patent databases. This presentation will provide an overview of existing methods: their similarities and their differences. The effect of applicant name harmonisation in improving search results and producing more reliable statistics will be illustrated with concrete examples. Several potential applications will be outlined.
Frank Tietze, Institute for Manufacturing (IfM), University of Cambridge
This talk shares our learnings from a recently completed regional patent analysis conducted to examine the innovativeness of Schleswig-Holstein, Germany's most northern state. Despite the continuously improving patent-data quality, we encountered problems arising from incomplete data on inventor and applicant addresses, which appears to particularly important for any regional patent study. After intense discussions about the choice of a suitable database that may lead to reliable results, we decided to extract data from the Thomson Reuters database. In order to validate the data reliability, we compared our data with data from the OECD operated RegPat dataset, which however provides only aggregated regional data.
Stephen Adams, Magister Ltd
The average patentability searcher faces the proverbial task of locating a needle in a haystack; the single answer or very small answer set selected from amongst the ever-increasing volume of documents, with ever-diminishing differences between each one. At the other extreme, patent information is also utilised for large-scale statistical reviews, technical state-of-the-art surveys and other projects which demand the collection, analysis and presentation of very large answer sets. The techniques for retrieving and handling these large sets are quite distinct from the familiar territory of the patentability search. The use of software to render a visual impression of trends in bulk data is a popular solution to the problem of how to handle these large results sets. However, the end-user of these tools is rarely an expert in human-computer interaction or visual psychology. Is it possible to be misled by the technology, into trains of thought which could lead to bad commercial decisions? Are these tools are revealing true phenomena or merely artefacts of the software algorithms? This presentation will attempt (probably without success) to display some answers.
Nigel Clarke, Alain Materne, Gershom Sleightholme, EPO
Patent thickets are dense webs of overlapping intellectual property rights. Firms who wish to be active in a particular area of technology need to ensure that their products will not infringe existing patents or existing patent applications which might later be granted. This requires an adequate patent search to uncover all interrelated patents and applications. Frequently, interrelated patents and applications are from a single applicant, but often with multiple inventors, or sets of inventors, and concern similar, but not identical subject matter. The challenge is to identify such relationships between patents and applications, as distinct from other relationships such as families, citations and divisionals. This lecture presents simple-to-use but effective techniques which anyone may apply, using freely available data, to retrieve such related applications. The techniques are based on the applicant name, inventors and titles. Additional possibilities derive from the application number, priority dates and classes. These techniques are based on those used at the European Patent Office to find similar applications at the pre-search stage and which have proven their value in widespread systematic use for over a decade. Compared to existing techniques such as patent landscaping the techniques proposed here are simpler to use, quicker and less expensive. Examples from various fields of technology will be presented using Espacenet.
Emmeline Marttin, EPO
This talk is about searching Espacenet for patent applications, taking the pharmaceutical area as an example. After a brief introduction of to how to use the advanced search interface in Espacenet, a practical example will use the CPC to showcase how to make a search table using classification. Based on the results of the search, there will be a discussion on how it can be refined and re-focussed – using the problem-solution approach – to arrive at an even better overview of the closest prior art.