Tuesday 20 Oct

Computerization has changed the working environment dramatically and has a significant impact on jobs. In this presentation, the influence of technological progress is shown on the work of the searcher and knowledge manager in the industry over the last 20 years. Changes in customer needs, new products and technical progress are contrasted and compared. Mutual interactions of these 3 criteria will be listed together with their impact on the requirements for an industrial knowledge center.  Knowledge Managers must develop their  job profile accordingly to keep up with these changes.

Consumerization of IT in the enterprise has been defined as the specific impact that consumer-originated technologies can have on enterprises. It reflects how enterprises will be affected by, and take advantage of the technologies and models that originate and develop in the consumer space rather than in the enterprise IT sector. At least one study has shown that over the next 12-18 months, organizations expect to see consumerization of IT in the enterprise creating a positive impact in user satisfaction, user productivity, process efficiency and collaboration, and business agility. IBM claims that a “bring your own device” (BYOD) policy can result in increased productivity and innovation, in increased employee satisfaction, and sometimes in cost savings. From the point of view of vendors of niche cheminformatics software, “monetization” is a real problem: since the market for mobile apps is essentially a consumer one, buyers expect apps to be free or very cheap. Probably many hundreds of “chemistry apps” have been developed, but few are truly useful. I will give some illustrative examples. Technical approaches are either based on HTML5, or on platform-specific development; the debate on the pros and cons of each has an element of religious wars about it. Cloud computing cannot be ignored. Regardless of whether desktop, web or mobile wins, or an entirely new platform becomes the standard, it is clear that change is afoot in the way that chemists access the software that they need.

Searching for Markush structures has been a rather difficult task especially since it was necessary in the past to work with different retrieval systems. With the new implementation of the DWPI Markush database from Thomson Reuters on STN it is now possible to search for Markush structures using a single structure query for all structure databases.  In this system the structure and bibliographic databases are integrated within a content domain which allows easy and fast projections between the databases. It will be shown that the DWPI Markush concept of superatoms can be integrated in the STN query system, allowing users to exploit the full potential of the DWPI Markush data. To enable complete and high precision searches it was necessary to develop a new Markush search engine. Improved evaluation of Markush structures is possible with hit structure display, highlighting, and assembled structures.  Based on this implementation it will be possible to develop further innovative features in the future.

Markush claims are widely used in chemical patents to define large chemical spaces, however understanding these structures, their formalized representation, search and analysis is complex and poorly supported. The problem becomes more acute with the need for comparison of these potentially vast patent Markush structures with classic structure and combinatorial virtual libraries. At this time, all of the existing Markush representations are incomplete, search systems are proprietary and not available for in-house integration and with many key use cases there are no existing tools and often no algorithmic background to begin the solve. This presentation briefly summarizes the current state of Markush claims, the existing challenges from the perspective of both IP professionals and computational chemists and highlights some promising new developments directly relevant to today's discovery processes.

The Web of Linked Open Data, or LOD, is the most relevant achievement of the Semantic Web. Initially proposed by Tim Berners-Lee in a seminal paper published in Scientific American in 2001, the Semantic Web envisions a web where software agents can interact with large volumes of structured, easy to process data. It is now when users have at our disposal the first, mature results of this vision. Among them, and probably the most significant ones, are the different LOD initiatives and projects that publish open data in standard formats like RDF.

This presentation provides an overview and comparison of different LOD initiatives in the area of patent information, and analyses potential opportunities for building new information services based on largely available datasets of patent information. Information is based on different interviews conducted with innovation agents and on the analysis of professional bibliography and current implementations.

LOD opportunities are not only restricted to information aggregators, but also to end-users and innovation agents that need to face with the difficulties of dealing with large amounts of data. In both cases, the opportunities offered by LOD need to be assessed, as LOD has just become a standard, universal method to distribute, share and access data.  

Pharmaceutical companies are highly dependent on access to high quality information retrieval. Insufficient gathering and selection of scientific information could potentially impact corporate decision-making in a wrong direction.

To assess the value of external information retrieval services a number of third party information providers were contacted with two information research requests (within inflammatory diseases). The providers were asked to return with search results and search methodologies used. In the first search the interaction with the providers were kept at a minimal level, whereas in the second search the contact, direction, and interaction were increased.  

It is concluded that information research results from different providers are variable. The expected increase in inter-homogeneity of results from the different providers could not be confirmed after the second search. The overall overlap of results was 38% for the first search and 33% for the second search, and surprisingly none of the references were found by all providers.

To fully cover the area of interest and to avoid bias it is recommended to perform exhaustive scientific literature searches. Researchers and decision-makers should accept large amounts of results from literature searches and promote initiatives to analyse these results in detail.

With more than one million patent applications filed every year, searching for prior-art has become a daunting task. This constitutes an important challenge for technology companies and their legal representatives, as the value of its assets depends on their ability to demonstrate the novelty of their inventive efforts. Failing to identify prior-art makes it difficult for companies to patent their inventions and exposes them to costly litigation. The breadth and complexity of the IP space makes it all but impossible to search for prior-art without the help of machine-based intelligence that identifies relationships between a new invention and those described in millions of patent documents. Existing solutions fail to take into account that companies often use (and are strongly motivated to) different words to describe similar inventions. This makes search efforts based on the similarities between words prone to miss relevant prior-art. What is more, existing techniques do not account for temporal changes in the terminology used to describe particular inventions. This is not a trivial omission as, by definition, the search for prior-art requires comparing an invention with other produced at different points in time. AIP developed an advance search engine that addresses these shortcomings. AIP uses thousands of examination reports to learn about textual relationship that describe scientific concepts and applies this learning to compare inventions. That is, instead of comparing document on the basis that these contain similar words, our algorithm compares document on the basis that these describe similar ideas. We present a number of cases were AIP’s solution helped companies patent their inventions, license technologies, and address litigation challenges.