Monday 22 October 2007

At the dawn of the 21st century, we face the realisation that the tangible economy and its attendant modes can no longer be represented or financed using the convention of the “balance sheet.” Presaged with the collapse of the 2001 WTO debates in Doha, heralded with the recent stock market multiple personality disorder, and against the growing drumbeat of currency strain, global finance is about to experience the collapse of the tangible, industrial market paradigms and the emergence of the intangible, asymmetric risk capital era.

• In 2005, the US and international governments cancelled or curtailed sovereign immunity from infringement liability previously afforded to government contractors. In 2006, the US Justice Department argued to encourage patent infringement when it feared the loss of Blackberry, while the Commerce Department decried intellectual property abuses in China.
• In 2008, all banks and financial institutions will have to test their loss reserves for their exposure to “intangible economy risks” under the Basel II Accords. Neither their internal systems nor external business practices are prepared -- anywhere.
• Over $1.5 trillion in uninvested capital in the Islamic world -- estimated to grow to as much as $3 trillion by year’s end -- directly or indirectly stands ready to move into diversified currencies further destabilising weakened dollars, yens and euros.
• Infrastructures designed to arbitrate to whom proprietary limited monopolies are granted have been shown to be unsustainable and overly provincial giving rise to the emergence of “sovereign-backed” monopolies.
• The presumed hegemony of the Trilateral countries in terms of invention and innovation is increasingly challenged by the “Silk Road Axis”.

A discussion of these events (and their consequences) will be presented together with a series of anticipatory scenarios to explore the strategies to adapt in this financial transformation.

Throughout 2004 and early 2005, the management team of 3M’s worldwide information organisation undertook a project to envision the staff, services and resources needed to support 3M’s information requirements in 2010. Using several intersecting strategic planning processes, a map for the future was created along with a plan for closing the gaps to meet future organisational needs.
This presentation discusses the strategic planning processes used, the outcomes, and recent adjustments required to maintain alignment between the information group and 3M’s changing corporate strategies. New models necessary to stay on the path to 2010 are described including mass customisation, prioritised intermediated information research, and libraries as innovation spaces.

The development of biomaterials is accelerating in many areas of medical and surgical devices. There are an increasing number of biomedical application areas, from cardiovascular and dental, to ophthalmologic, orthopaedic and urological treatments. There is also an increasingly complex patent landscape emerging that covers fundamental areas such as surface modification and characterisation of biomaterials which are important for drug delivery systems, tissue engineering and implants. The paper describes an approach using advanced text mining with concept maps, together with large scale patent analysis that provides a path for discovering new patent white spaces. We also discuss a methodology for mapping new potential innovations.

Bio-pharmaceutical industry is challenged to replace drugs that come off patent with novel approved medical treatments that meet increasing FDA safety and efficacy demands. Therefore, the status of Competitive Intelligence has risen from a support function to a strategically important discipline at the core of investment decisions, mergers, acquisitions and licensing activities. In addition, there is an increasing need to identify new drug candidates and other Intellectual Property for licensing.

We propose a solution that complements the search and mining of IP / Patents and biomedical literature with an innovative approach toward the annotation of domain specific entities, including chemical names. Those chemical names of arbitrary complexity are linked to chemical structures, represented by Smiles strings. Our approach will render the scientific literature and patents searchable by web-enabled chemical structure/substructure search applications.

Our solution includes a data warehouse that can contain both structured and unstructured data, along with chosen annotations. IBM’s Blue Gene supercomputer is used to increase search and annotation performance. By means of Blue Gene we are able to populate and query the data warehouse rapidly, and optionally to add computed properties to given chemical entities. The presentation will include a brief demonstration of the solution.

A fully effective search tool for chemists requires access to both unstructured textual sources as well as structured chemical information. This presentation looks at two of the challenges this presents. First, how to extract information from unstructured text such as patents or the academic literature. Second, how to conduct searches across disparate sources of information -- that is, how both document similarity and chemical similarity can be combined in a single similarity search. Such a search might start with a portfolio of patents together with some keywords and chemical structures. The search would find those patents which both use similar language in the text and which reference the chemical structures. The search can then be refined iteratively by including or excluding any of the items returned by the search: documents, words, phrases, structures or fragments; and then repeating the similarity search. Consideration will also be given to the challenging environment for such searches where there are very large numbers of documents (tens of millions of patents) containing many structures (millions of chemical structures).

There is a large amount of chemical information contained in commercially available patent databases. Uncovering that information is relatively straightforward, but converting it into intelligence is often more challenging. This presentation examines several methods for turning chemical information from patents into actionable intelligence. At one end of the spectrum are tools that are readily available to most information professionals, such as Microsoft Excel and simple analysis tools that are built into commercial search engines. At the other end are sophisticated text and data mining tools that can uncover hidden intelligence from lists, matrices and relationship maps. The presentation shows that it is often possible to uncover intelligence with the simpler tools, while discussing situations where the more sophisticated tools add value.

This introductory keynote will look at the range of business models extant in professional and business information markets, trends in business models change and implications of these changes. The review will cover subscription/licence, advertising/sponsorship and the emerging feasibility of more "open" and network/community models. A range of examples will be covered across a number of information markets.

Traditional search engines require that “matching” documents contain nearly the exact words or phrases used in the query. Such search engines cannot identify relevant documents which use synonymous, but not identical, terminology. As a result, finding important documents can be difficult due to the need to guess what terms an author would have used when describing a given concept. Conceptual Search addresses this problem. Sophisticated statistical techniques are used to group documents into topical categories, and to build a database of synonymous terms. Then, when a search is performed, all synonymous terms that are related to the searcher’s input are considered – not just the literal terms in the query. At the same time, the topical groupings permit the conceptual search engine to ignore documents that contain the literal words in the query, but which are nonetheless irrelevant to the searcher’s needs. The result is a search engine which has a high retrieval rate (recall) and excellent ability to discriminate relevant from irrelevant documents (precision). Using such a search engine, searchers can search faster, and have greater confidence that all relevant documents have been located.

Google caused great debate amongst the patent searching community at the beginning of 2006 by unveiling a beta-test search engine for US patents. But how good are free patent databases on the internet? The new patent databases are certainly opening up the world of patent searching to more people; inventors, laboratory scientists and engineers are carrying out patent searches. But is there a place for them in the professional patent searcher’s tool-kit?

This presentation starts by looking at databases from the patenting authorities such as EPO, JPO and USPTO and then beyond to the more recent developments in free patent search engines from free patents online, fresh patents and Google. These sources are compared for functionality, usability and also coverage and timeliness; all important criteria for a comprehensive patent search whether for freedom to operate, prior art, research or validity purposes. The presentation is illustrated with examples from ultrasonic devices, micropumps and other typical searches from The Technology Partnership which support our developments of next generation physics-based products.

The EPO's free web-based esp@cenet service provides an entry level to patent information novices or to more experienced searchers when beginning a new search before moving on to more sophisticated searches. Search term input is via search masks, and allowed search inputs may be conveniently grouped into "parameters": eg, publication date, publication number etc, "classification symbols": IPC or ECLA codes, "proper names": inventor name, applicant name, or "keyword". The keyword input in esp@cenet can be used to search in the title field or the title and abstract field. Many technical fields such as mechanical engineering are tractable by keyword searching, However a recent profiling survey of esp@cenet users showed that two of the largest user groups are from the chemical and pharmaceutical sectors. Much chemical patent and journal literature searching in commercial and public domain databases is carried out using chemical structures, ie, the allowed search input is a chemical structure, written, drawn or selected from a library, by the searcher. This input is then used by search engines to interrogate databases, and retrieve the relevant documents, patents or journal articles for example. Keyword searching on its own is not the most effective way to search chemical or pharmaceutical literature.

The current esp@cenet search engine does not support chemical structures as input, however it does support keyword searching. In view of the utility of chemical structure searching to chemists or pharmacists, but with the restriction of searching to keywords (chemical names) in esp@cenet, it was decided to investigate the possibility of deriving chemical names from structures identified by the user, and using those derived chemical names as search input for esp@cenet. This paper reviews the R&D work which has identified the necessary building blocks to facilitate the process for structure searching in esp@cenet; Draw/select structure => convert structure to unambiguous code => convert unambiguous code to chemical name => identify synonyms => present user with synonyms => search with synonyms in esp@cenet.

An overview of the possible public domain software which could be used to build such a process will be presented, and from this the way to a possible chemical structure "translation engine" adapted for esp@cenet will be indicated. The anatomy of an idealised chemical structure search incorporating classifications will be demonstrated. Some of the remaining difficulties which present obstacles to progress will be discussed.