« Why AMI technology is so perfectly suited to Competitive Intelligence solutions ? » An interview with Eric Fourboul, AMI’s Chief Technology Officer.

What are the principal challenges of Competitive Intelligence ?

Eric Fourboul : The real challenge is the unknown !

Detection of unknown information: In terms of just day-to-day searching for information, the user  pretty much knows what they want; In Competitive Intelligence, a user will tend to express, “more or less”, descriptions of the subjects of most interest. In Competitive Intelligence the monitoring and collection of information is carried out across sources that are less visible or obvious using tools that are installed specifically to detect new and previously unidentified information. AMI technology makes it possible to build sophisticated Competitive Intelligence solutions combining very high levels of relevance with the use of “expressions” and “related terms”. Our automated data collection technology detects both new and previous un-identified information vital to a users specific areas of interests.

Unknown sources: The significant development of data-processing networks and the availability of the Internet induced new corporate communication strategies. In fact, the volume of information “on line” has now become more important to an Organisation than reliance on the documents and reports which it published a few years ago. The term “layers of information” in connection with Internet has become an accurate one!    

One of the main consequences of the increase in available data is the increasing technological sophistication of the systems containing and disseminating the information. At the beginning of 2000 the contents of most online presences were easily “accessible” allowing the rapid rise of internet search technologies like Google which were able to index the content simply because it was readily accessible. However as an example of the most significant changes since then, in a recently deployed Competitive Intelligence project we were able to prove that standard Internet search engines covered only 25% of the information and sources the customer required and that the information of precise relevance, i.e. that which gave the project and on-going processes real value, was within the other 75%, often referred to as the Invisible Web.

Why the need for dynamic analysis ?

Eric Fourboul: Products which use a linguistic model do allow some degree of competitive syntax-semantic analysis and consequently a terminological extraction with a semantic typology which is used in the subsequent process of analysis. Existing in various forms, “conceptual graphs”, “semantic networks”, “linguistic cartridges”, etc these linguistic models are created in advance, at the start of the project, and thus, are based on known and pre-determined semantic fields. These tools are thus adapted to present analysis based on historical events. However, they loose their value completely as soon as there is a requirement for the system to highlight new phenomena not envisaged in advance but nevertheless of absolutely crucial importance to the Organisation. Competitive Intelligence programs now require more intelligent ways of analysing far greater volumes of information dynamically. The advantage of our algorithms, which do not require this preliminary modeling, is to offer high levels of analytical performance with the additional ability to collect and analyse new linguistic information such as “expressions” and “closely related terms” as they occur.

And tomorrow ?    

Eric Fourboul: Over the last few years we have invested heavily in textual analysis on the basis that unstructured information now represents more than 80% of the information both within and available to an Organisation. Our application AMI Analyze offers advanced but easy to use Charting and Analysis functionality giving the ability to quickly highlight significant events, detect tendencies, identify relationships that may not be immediately obvious, and find even the most obscure items of information. Nowadays it is almost impossible to implement an effective Competitive Intelligence strategy without these functions. More generally, they have become essential tools for the decision maker simply because they allow a far better understanding of customer behaviour and requirements, market evolution, and positioning of competitors in a way that traditional Data Mining tools simply don’t provide.