Tuesday 21 April 2015

This talk looks at how news, media and publishers are using language technology and the semantic web to go beyond search and data. It will show how industry leaders are using technology to improve their institutional memory and the discoverability of their content.

The European  Patent Office (EPO) faces the unique challenge of providing a consistently high quality patent search with an ever increasing number of patent and non-patent literature to search in. The Information Management Department of the EPO accepted this challenge and is now in the process of implementing new search technologies together with representatives of its business and external partners to provide its customers with the best possible, high quality search that finds the best prior art with less human cognitive effort. In our talk we introduce our foreseen search eco-system and how new technologies like knowledge-bases, federated- and semantic search together with classic information retrieval techniques will lead to the solution of the former described increasingly challenging search-problem.  We introduce qualitative benchmarking based on patents and search reports that guide us in our efforts and we shed light on the special way of how technology supported searches can exploit prior knowledge from the search behaviour of human searchers.  

Exploring the cited and citing patents associated with a certain technology or assignee forms an essential part of any thorough prior art search or patent analysis/landscaping exercise. Citations searching is not dependent on words, and can reveal connections (determined by expert patent office examiners) between patents that would not have been obtained in a subject search alone.

This presentation will discuss the benefits and challenges of using citations to enhance patent searches, and the most effective ways of retrieving and handling patent citations data including browsing tools, tags and analytics.

Patent landscaping is a growing field of patent searching.  Standards and guidelines are being developed to steer this growing field, and these will be highlighted and reviewed.   Some examples of search strategy for landscaping , relevancy analysis , categorisation, selecting visualisations, will be discussed.  Unlike landscapes prepared for policy decision making, industry landscapes need to link to the underlying data – the patent documents themselves must still be read, and the landscape itself will often be updated. Whatever the size of your landscape – continental to regional, you need to ensure you are navigating the right documents, the aim of this talk is to outline best practices.      

TOD system is an expert system for SMEs to recommend them next promising business items based on 1) their pre-owned technologies and experiences, and 2) industrial information about product entities and their relations extracted from massive technical data, such as Patents of United States (USPTO) and Trademarks.

TOD system, fully funded by the Ministry of Science, ICT and Future Planning of Korea, has been developing for three years by Korea Institute of Science and Technology Information (KISTI), and is about to be opened to public at the end of this year.

In this presentation, our experiences and findings obtained during developing the TOD system, collecting and analyzing the data, and designing the service models will be included.

• Followings are some of them which will be covered in this presentation
• Introduction of TOD system
• Source Data: Trademarks and patents of United States

Analysis Data:

• product names
• product value chains and business diversification chains
• relations of product-product, product-applicant, applicant-applicant, etc.

Employed Techniques:

• product name entity and part-whole relation extraction
• patent index analyses
• collaborative filtering
• S-curve fitting
• recommendations based on co-occurrence relation
• recommendations based on chains
• graph and chart visualization
• attractiveness estimation, etc.

Service Methods:

• product evaluation indices
• identification of own products
• promising-product recommendations composed of 10 analytic models
• graphs about competing firms and chains
• attractiveness chart to evaluate candidate products, etc.

Academic and industrial researchers nowadays expect an interactive, easy to use access to the scientific literature and supplementary information in order to satisfy their information needs quickly and comprehensively.

To address these challenges major scientific publishers have begun to chemically enhance/enrich their scientific articles, reference works and eBooks with information based on chemical compounds and reactions. Compound indexing, highlighting of chemical terms, on-click display of additional compound attributes and most important semantic linking to other relevant resources or information have been implemented. First and foremost enabling chemical structure searching in text sources offers benefits to the user never seen before by providing an additional, precise entry point to the full text. Sophisticated cutting edge technologies like chemical named entity extraction, chemical image recognition and automatic work-up of ChemDraw files are applied to fulfill these requirements.

This lecture will give an overview of current implementations of the concepts listed above and will illustrate the benefits as well as the challenges and pitfalls by one concrete example project.

Currently, analysis of large patent sets is a tedious and cumbersome work. In order to improve and speed up this process we developed a patent curation-workflow, in which relevant chemical information, such as Markush structures and chemical compound collections (e.g. exemplified structures), are extracted from a patent set and successively enriched with text-mining retrieved data in semi-automatic manner. The outputs of cheminformatic, OCR/OSR and text-mining tools are combined by means of KNIME and the joined data are finally visualized side by side with the original documents using the ChemCurator application. As well as advanced visualization capabilities ChemCurator offers essential functions for validation and manual refining of the automatically extracted chemical information. The created project specific content gives a solid information base of value to any phase of a drug discovery project.

Biotech patent searchers are masters in the art of knowing which tools to use and when to use them to accomplish a wide range of research tasks. In order to complete comprehensive searches, analysts must perform biological sequence searching in one application, full text searching in another, and report production in yet a third. Furthermore, searching text in the life sciences is fraught with challenges such as the overloaded use of words to name biological concepts (for example, the slingshot gene), and the complexity of biological concepts which forces biotech searchers to build and manage large lists of synonyms. We present here a series of real life search cases using a unified framework for the comprehensive searching of both life science text and biological sequences in a single environment. The framework utilizes life science ontologies to aid the user in developing conceptual queries that avoid the typical challenges in biotech patent text searching combined with a user-centric design to aid in rapidly examining merged results from text and sequence searches.

Figuring out how to identify who, what, when and where in scientific, technical and patent databases is now a relatively well understood process. Given the current technical and computational advances, can we now turn our attention to forecasting the future? This presentation explores some of the issues related to technical forecasting using bibliographic data and outlines some of the current research on identifying technical emergence. Included in the discussion are practical indicators for identifying emergent researchers, organizations and technologies.

Many companies feel challenged when it comes to the patent prosecution processes at the US patent offices. Limited understanding on the time it takes from application to grant and what steps may be encountered along the process, and thus, the investment required to get a patent granted. Upon analyzing this corpus of Big Data from the USPTO, the LexisNexis® PatentAdvisor suiteSM of patent analytics tools is built. PatentAdvisor brings transparency to an opaque process by  providing corporations and law firms with actionable intelligence from the USPTO. It brings valuable insight and efficiency to the patent prosecution process. By understanding how individual patent examiners and art units have responded to previous applications and the general workflow and timelines for those examiners IP professionals are able to reduce the number of office action to obtain a patent, quickly see the status of each patent and compare the efficiency of the outside counsel. And thus develop higher quality IP in a more cost effective and timely manner.