Visualisation boosts power of STM databases

Share this on social media:

Topic tags: 

Leading STM database providers are starting to offer visualisation tools to help interpret search results. David Mort reveals how this could help existing users to get more from their information and encourage new users

At the Online Exhibition in London at the end of last year, I was struck by the growing number of providers introducing visualisation products into their mainstream offerings. Amongst STM providers, for example, there were products from Thomson, FIZ Karlsruhe/STN International, and Questel Orbit. And Factiva Search 2.0, Factiva's latest search tool, which launched in beta format at the start of 2006, also strongly features content visualisation technologies.

Visualisation aims to add value by showing a large amount of data on an information map so that it can be assimilated at a glance. Visualisation technologies rapidly recognise associations among textual elements and then present them graphically. Usually organised through charts, graphs or other visual tools, visualisation enables the user to quickly grasp trends and patterns that would not be easily discernable in traditional text formats. Supporters suggest that, without these technologies, information might be buried in a mountain of information that would be impossible to navigate. As CAS president Robert Masse stated at the launch of visualisation product STN AnaVist in 2005: 'The information challenge of the 21st century is not information access, but information utilisation'.

What it offers

Thomson's new visualisation tool can show how authors within an organisation are linked and which external researchers would be good candidates for jobs or further collaboration.

Visualisation tools have been around for some time and they are well used in certain areas. Patent analytics specialists such as CPA and Metrics Group, for example, use the technology to track trends in the patent landscape in order to produce competitor and industry reports. Risk management and business intelligence companies such as Digimind, i2, and Datops have also embraced visualisation solutions.

Such companies have discovered that the technology can be used to discover a variety of trends and relationships. In the area of STM information, applications include tracking research trends through an analysis of patents or articles written in specific fields. This helps users to see if a research area is on the rise, steady, or declining. The technology can also help people check on competitors by analysing the subject and location of patents filed. Comparing geographical patterns of research by analysing patents filed or origin of article authors is another application. Visualisation also provides a way to examine relationships between authors and subjects covered, as well as frequency of authorship, and peaks and troughs in publishing output.

Despite the range of potential applications, the inclusion of visualisation in STM information products and use by information specialists and researchers is still limited. The likes of Thomson and STN, however, seem to think that this will soon change. They have invested considerable marketing resources in recent months to encourage use of their new visualisation products and to try and stay ahead of the competition.

Thomson Data Analyzer was launched in November 2005 as the successor to Derwent Analytics. Its visualisation technology can be used to analyse information from a variety of Thomson Scientific databases including its patent databases, Web of Science/Web of Knowledge and the various databases in Web of Science. Thomson Data Analyzer can also be adapted to analyse information from in-house databases and Excel files.

Visualisation tools such as Thomson Data Analyzer can open up information to a wider audience such as those in sales and marketing and business development.

In summer 2005, Chemical Abstracts Service (CAS) and FIZ Karlsruhe announced the release of STN AnaVist. Its key features include analysis of information from three leading scientific resources: CAS's CAplus database of scientific literature and patent information; and the USPATFULL and PCTFULL patent databases. It also offers an interactive workspace that can display a range of data visualisations that are dynamically integrated, including cluster and contour maps, histograms, and co-occurrence matrices. Harmonisation and standardisation of data prior to visualisation using algorithms based on intellectual data analysis is also included. Also, a data grouping feature helps to minimise scattering of results by permitting data elements to be edited and customised across the databases.

The challenges before acceptance

Visualisation tools can display large data sets in an easily digestible way and provide a concise picture of trends and issues. The use of colour and shading also makes most of these products pleasing to the eye. So can they become a key tool for information specialists and researchers, or will they just be an interesting extra feature?

The first concern is that visualisation tools are only as good as the data mining or information retrieval systems on which they are based. Many users would identify improvements to search technologies as the key priority before considering visualisation solutions. It is also important that analysis and visualisation tools should be integrated with the search and retrieval tools. Recent visualisation applications, for example, use a variety of tools to seamlessly analyse and visualise search results.

Another challenge is to develop easy-to-learn systems. These would encourage use by those people who prefer information to be presented in a traditional textual context. Some system providers appreciate this latter point and still use text options to accompany the visuals. The interface should be easy to use and the resulting analysis or visualisation should be easy to interpret. There can be a temptation to try to use the technology to provide over-complicated solutions when the tools are probably best suited to providing a general overview or a general trend analysis. If these tools are to reach a broad audience, then simplicity and ease of use are essential. Tools that are easy to use will gain acceptance more quickly, and have lower training costs.

In addition, products should be able to analyse and visualise information from a variety of sources, both in-house and external, and from structured and unstructured data.

There will also be hardware and software issues to consider. More complicated solutions may be needed when data is from multiple sources, especially if those sources are a combination of internal and external data. As well as the costs of purchasing and running visualisation software and programmes, which can in themselves be expensive, there may be costs for new hardware and other software.

Attracting new users

So, will tools for analysis and visualisation solve the problem of information overload? Content and cost issues, plus a reluctance of some traditional information users to turn to graphics rather than text, suggest that the growth of these tools will be gradual rather than dramatic. However, they may have a role in extending the use of information across organisations.

The eye-catching appeal of the technology has already allowed vendors with these tools to show the potential of visualisation to a broader audience. For example, patent analytics services have used visualisation tools to encourage the use of patent information beyond just R&D and specialists. The tools can take the information into new areas such as business development, competitor intelligence, and strategy.

This aspect of visualisation may suit end-users that have not been heavy users of text-based information sources in the past, and may be confused by the range of information options now available. Such users could be persuaded to use graphs and maps if they are relatively easy to produce, can present a quick overview of a trend or issue, and can reduce the time involved in the information retrieval and analysis process.

David Mort is director of IRN Research, a UK-based market research and information company specialising in the analysis of European information and content markets. Contact him at: