Outils et bases de données
Tools and databases
The EPO provides the public with free or inexpensive access to tools and databases with which they can retrieve and analyse its patent data. In this section you can find out more about them.
On average, one out of five search reports issued by the European Patent Office (EPO) contains a nonpatent literature (NPL) document citation, mainly from scholarly papers. More than half of all search reports in chemistry and biotechnology cite NPL. Hence, consideration of NPL is essential for ensuring the quality of the prior art searches. In this paper, the EPO resources for searching and retrieving academic papers are presented and an overview of dedicated internet search engines for academic publications is given. Some specifics of NPL search strategies are illustrated and a short note on some legal issues concerning the publication date is also included.
Authors: Ives Verbandt; Els Vadot
- Download "Non-patent literature search at the European Patent Office" (PDF, 2 MB)
- Watch the video (1:11:21)
This paper describes the evolution of the EPO's search tools over time and their envisaged revolution towards supporting a more or even fully automated search process. Regardless of whether the goal of fully automated search is achieved completely or only partially, the chosen approach will in any case bring about major improvements for both the EPO's examiners and the prior art search community at large.
Author: Dominique Andlauer
- Download "Automatic pre-search – an overview" (PDF, 2.1 MB)
The public and private sectors have both invested heavily in recent years in the production of machinesearchable full texts of patent documents. Access to these databases enables even unskilled searchers to locate at least some relevant material for practically any search task. But full text provides only one viewpoint; there are circumstances in which neither the original text nor a machine translation provides sufficient keys to retrieve the best prior art. This article illustrates that comprehensive prior-art search results will only be obtained if the searcher understands the inherent quality (timeliness, accuracy,completeness) of the databases available, including any known weaknesses.
Author: Stephen Adams
The search report established by a patent office mentions documents which may be taken into consideration when deciding whether the invention to which a patent application relates is patentable. This report contains a set of valuable information which, once public, can be accessed through different services on the websites of some of the major patent offices. In the current international IP landscape, patent applications tend to belong to large families where a search report is established for each family member and made accessible by the patent offices where the search was performed.
The Common Citation Document (CCD) service, resulting from an IP5 cooperation project, proposes a single point of access to all prior-art documents cited by each patent office for each patent application and related patent applications belonging to the same family.
Through the CCD, this information is accessible in a standardised format thanks to a system making citation data from patent offices available to the public. This access is made simple thanks to several features of the CCD web application, in particular the graphic display of the timespan for a collection of citations, such as relevant priority, filing and publication dates of a patent application.
The purpose of this presentation lecture is hence to explain how the retrieval of citations, simplified by the easy access to the CCD web application, can help patent searchers during the search phase. A number of special cases and examples will be given special attention. Some examples of the combined use of the CCD web application and DI+ will further stress the importance of the retrieved citations when combined with File Wrapper data.
Author: Pietro Rini
From its inception Espacenet was intended for small business entities, researchers and users that do not need, or cannot afford a subscription service but who could benefit from free, searchable patent data. Today, eighteen years after its launch, Espacenet is still considered by independent commentators as an extraordinary free benefit to the patent information user community.
Our research tells us that Espacenet is used not only by those who are new to patents, the original target group, but also by IP professionals. However, in recent consultations with a range of types of users, we were surprised at how little even the “expert” users know about Espacenet’s features and functions. This presentation is about setting the record straight. What Espacenet can do that you didn’t know it could do. You might be surprised!
Author: Nigel Clarke
The concept of an automated search for relevant prior-art is a hot topic among patent professionals. This workshop gives an overview of how the EPO has implemented automated searching within the pre-search framework, presenting the different approaches currently used and their efficiency. In addition, the impact of the automated pre-search is discussed: How should the results from an automated search is handled to achieve maximal search efficiency, speed and quality?
- Download "EPO pre-search framework" (PDF, 3.5 MB)
Author: Johan Bengtsson
With the ever-increasing volume of data to be searched, the techniques of semantic search will be key to successful prior art searching. These techniques consist of methods for understanding the searcher`s intent, disambiguating the contextual meaning of (search) terms and ultimately improving search accuracy by generating more relevant results. This presentation explores how far the EPO has come in enabling some of those key elements through projects such as Annotated Patent Literature and Enhanced Ranking. It also introduces even more sophisticated models based on machined-learned algorithms that might help to shape the future of search at the EPO.
Author: Alexander Klenner-Bajaja
- Download "Towards semantic search at the EPO" (PDF, 1.5 MB)