Tuesday 16 Oct

Markush structures are a representation of the chemical compounds covered by the claims of patent documents. Due to their highly generic nature they are more difficult to search than specific chemical structures since they require special indexing and search features to make them searchable. Few databases exist that offer such features. The presentation will discuss various aspects of Markush structure searching and indexing from a chemical company’s perspective like importance of Markush searching, difficulties associated with it, current status of Markush databases, recent developments and our expectations towards data providers.

For more than 40 years and long before the Internet was invented, traditional online hosts have been providing efficient online services. In the meantime, access to a continuously growing amount of electronic information has become a commodity. More and more scientific literature has become accessible via Open Access and in particular patent information is available free of charge – a big challenge for commercial information providers.

Nevertheless the number of information providers has grown; they offer a broad spectrum of information packages and serve the diverse needs of various user communities.

In many cases, basic or incomplete information is considered to be good enough. However, advancing innovation, driving R&D, and protecting intellectual property demands on-going high-quality and trustworthy content and retrieval tools.  Value-added and aggregated content, precise searching, and reliable expert support are still lasting values provided by online hosts.

How can information providers reach their objective to provide high-quality services under increasing cost pressure? What is the proper strategy to cope with rapid technological changes, increased competition, a changing user base, and the geographical shift of content and users? Who are the users of tomorrow, what are their requirements, which benefits can they expect in the future? Which business models are appropriate?

The numerous challenges to information providers will be discussed and an outlook will be given.
 

 Chair: Randall Marcinko

 For decades, scholarly scientific information has been accessed via powerful online hosts.  The information has been curated by subject specialists and made available in private walled gardens where substantial fees are charged for access.  Best practices for storage and retrieval have included subject indexing with extracted entities, powerful searching that includes boolean and proximity operations, subject and entity alerting to the enduser, continually increasing precision and recall in some of the most challenging subject areas and a host of additional added value features.  The past decade has witnessed unprecedented growth in internet access to a plethora of content of all types.  While sometimes advertising-supported, fees are not charged for access to information via search engines and other internet search tools.  The enduser of today is tempted by Wikipedia, by Google by others to rely on fast, free access to vast storehouses of internet content.  Information professionals continue to tout the merits of fee-based and well curated databases of information provided by trusted online hosts.

We have assembled a panel of some of the best online hosts and our discussion will focus on the future of information delivered by these hosts and others in the information industry.  We will ask the tough questions about the future of this industry and involve the audience in an interactive exploration of the future of the Online Host.  Each of our speakers will make a brief statement of their views, followed by a hard-hitting question period to delve into the issues.  How will the enduser of tomorrow justify paying significant fees to access online content when vast amounts of non-curated content can be retrieved for free on Google?  How will fee-based information be differentiated from free? How will the enduser of tomorrow be trained to recognize the quality of information delivered by online hosts? Will the publishers of databases of scholarly information be tempted to make currently deep-web information available for free to search engines?   These are just a few of the many topics that will be posed to our venerable panel of industry experts by the session moderator augmented by interactive audience participation.

The benefits of having a widespread patent portfolio are well known: Patents provide you an exclusive right with respect to a certain technology and put you in a position to dominate your competitors. On the other hand, in a situation of strategic balance patents may even offer a deterrent effect.

During such an “arms race” for patents, competitors are often not aware of the economic impact: Patents are expensive, no matter what their value is, especially due to their hidden costs. Therefore many companies consider getting their portfolios rated to find useless patents aiming at reducing costs. Rating methods are also used to prevent futile inventions from being patented.

The presentation shows approaches to qualitative as well as quantitative ratings of even big patent portfolios and their economic outcome.

 

Inventive activity is spread all over the world and the number of patents increases significantly. Increasingly the innovation process termed “open innovation” relies on series/bundles of patents rather than single inventions. Therefore it becomes highly necesarry to improve the circulation and the commercialization of patents. But if the necessity to create an IPR market is certain, the central question is now:

•    How to implement concretely such a market?
•    How could it become a patent exchange?
•    Under which conditions such a market will benefit the core actors of invention: pro and SME?
 

Different representations of chemical information are common place in patents. They influence the production of searchable data repositories, and also have a direct impact on the searcher's ability to formulate a reasonable search strategy. With a view to overcoming the difficulties in comprehending, translating and transforming visual representations of chemical compounds, options for using software assisted working methods are explored. The information-seeking behaviour of the patent searcher will be analysed in the context of a prior art search for new chemical entities. The performance of existing software, such as state of the art name-to-structure converters and chemical annotation tools, is not always satisfactory when applied to the uncontrolled representation of chemicals in patents. Possible approaches to improve the efficiency and reliability of the search task will be discussed. Common patterns in search tactics seem to be guided by domain knowledge, the structure of searchable data repositories and by the application of the patentability requirements for chemical inventions. 

The information and analysis provided by the presentation may be of relevance to designers in developing systems that will support the searcher in the cognitive processes used to formulate technical database queries and to adapt to the chemistry information retrieved at any stage during the search process.

Text mining worldwide full-text patent documents is now possible using cloud computing. This presentation looks at some of the challenges involved, and some of the benefits. We contrast with keyword search, and show how text mining can improve recall of relevant patents, reduce noise, and provide a methodology to better inform search strategies. We also contrast with layered approaches, where keyword or substructure search is performed first on a patent database, and text mining is only applied to the subset of documents brought back from the initial search.

The presentation concludes by overviewing other applications of patent text mining, including prioritising  patents for review,  synthesising  information for competitive intelligence, and extracting facts  and relationships in order to connect to structured data or to relationships extracted from other unstructured text. We will also look at developments in the recognition of novel chemical structures, the extraction of data from tables, and linking information from different parts of a patent.
 

International patent filings account today for about 50 percent in total patent applications (WIPO, 2011b), as any business involved in global operations must deal with the fact that whereas economic activity tends to be increasingly borderless and a high-tech global technology company, such as Microsoft, earns more than 50 percent of its revenue in overseas markets (Phelps, 2005).

Given that there is not a global patent system and significant differences between national patent systems exist (Lerner, 2002), a firm faces a dilemma with respect to the potential benefits of protecting its invention or technology in a foreign market, on the one hand, and the costs to bear to obtain patent protection and enforcing it under a foreign jurisdiction, on the other hand (Moy, 1993; Pagano, 2007). Although a home-bias exists (Dernis & Khan, 2004), the rapidly growing international commerce intensifies cross-border competition and forces firms from technology intensive sectors such as in information and communications technologies (ICT) to protect their inventions in foreign markets (De Prato & Nepelski, 2012). We deal with the issue of international patent filings in ICT, and aim at providing answers to the questions concerning the drivers behind this phenomenon. In particular, our analysis focuses on the questions of how technological proximity and sophistication, a proxy for technological competiveness, and product market competition affect the incentives to file a patent application in a foreign country. To address the above questions, we construct bilateral measures of foreign patent applications for the global population of countries active as both a source of patents and a destination of foreign applicants seeking for patent protection abroad. In order to cast new light on the factors driving the international patenting in ICT, we apply the gravity model in which we include as explanatory variables of international patenting  geographical proximity and common language, measures of GDP and foreign direct investment, Innovation capacities in the field of ICT, technological specialization patterns, technological sophistication, and a variable concerning the quality of the patent system in a country in which non-resident applicant seeks to protect her inventions.