Article en français

Since the 2000s and the rise of the Internet, access to information and data has become very easy.

This massive data has found its concept: big data. It is a source of major innovation and unprecedented opportunities, but also of colossal challenges. At the heart of these challenges lies the intelligence process: a vital activity for companies wishing to remain competitive and at the cutting edge of innovation.

With the explosion in available data, monitoring has become extremely complex, requiring considerable resources. The volume of new information produced every day is such that, despite very precise search strategies (keyword selection), the noise (information that turns out to be uninteresting after analysis) increases considerably and can in some cases represent up to 80% of the flow of a periodic watch…!

The solution which aims to reduce the scope of monitoring, by being more restrictive in the choice of keywords for example, is no longer sufficient. In practice, we are now seeing that by proceeding in this way, we are no longer capturing information that would otherwise have proved relevant!

So… how do we go about it?

Artificial intelligence, and in particular Machine Learning, one of the most promising technologies of our era, promises to be the key to transforming traditional intelligence into an agile, accurate and efficient activity.

In this article, we explore how Machine Learning is redefining business intelligence, thanks in particular to innovative solutions such as IPMetrix, TKM’s business intelligence software specialising in the analysis of heterogeneous data.

The intelligence process: a job made more difficult by the explosion of data

Digital technology and the Internet have revolutionised the way we access information. The initial promise was seductive: abundant information easily accessible to anyone with an internet connection. However, this promise quickly showed its limits.

The illusion of information just a click away

The ease of access to information has created a misleading impression: while we thought that monitoring would be simplified by digitalisation, it has paradoxically become more complex. The challenge is no longer to access information, but to filter it, analyse it and determine its relevance.

Technology and scientific intelligence: a growing challenge

Scientific intelligence is a striking example of this challenge. With 4 million scientific articles published and 3.4 million patents filed every year, it is quite simply inundated with information. To illustrate this, in a field like cosmetics, there are more than 1,500 new patent applications published worldwide every month. How can a researcher, a watchdog or an IP director keep up with such a volume of information?

The quest for relevance

The overabundance of information poses the problem of relevance. It is not uncommon for only 10 to 15% of the documents in a continuous stream of information to prove to be truly relevant after review by the analyst. This person will then have to devote a large and growing proportion of his or her time to an activity that is not only tedious but also non-productive, which consists of sorting the monitoring flow into « relevant information » on the one hand and « irrelevant information » on the other.


The intelligence process: Machine Learning to the rescue
What is Machine Learning?

Machine Learning, or automatic learning, is a branch of artificial intelligence. which focuses on the construction of models capable of learning from existing data in order to subsequently reproduce tasks such as classification. These models are useful when it is too complex, or even impossible, to define the rules to be introduced into an algorithm. They learn from a set of data already processed by humans to reproduce their actions.

The diversity of models that can be used, the heterogeneity of the data on the one hand and the needs of the users on the other, can make this a complex exercise, with very varied results in terms of performance.

That’s why generic AI modules that generate results with a high level of accuracy in just a few clicks, for any type of user or need… are still just a promise. For technology or scientific intelligence, where both the mass of information is colossal and the need to be precise is paramount, AI is a powerful tool provided it is well mastered. Simply detecting 60% of potentially blocking patents as part of a watch will probably not satisfy any researcher or industrial property manager…

This is not the case for those we have been working with and supporting for almost 20 years at TKM.

The impact of Machine Learning on intelligence

Machine Learning represents a major development in the field of intelligence. Tools based on this technology, such as IPMetrix, are capable of analysing and sorting very large volumes of information with very high levels of accuracy, saving the person responsible for monitoring a great deal of time. Firstly, by eliminating the noise (irrelevant data), but also by putting the relevant information in the right places.

10 Minutes instead of 4 days each month…

One of the most tangible benefits of Machine Learning in market intelligence is the time it saves. Whereas traditional, manual monitoring can require several days’ effort each month to filter and qualify information, Machine Learning-based tools can accomplish this task in a fraction of the time. With a Machine Learning-based tool such as IPMetrix, the monitoring filtration process is reduced to less than 10 minutes, freeing up precious resources for other, higher value-added activities, such as analysis and dissemination.

TKM: combining Machine Learning and Data Mining to optimise the intelligence process

The combined power of semantic AI, statistics, Machine Learning and data visualisation.

TKM does not rely on a single technology to facilitate the intelligence process. On the contrary, it combines several disciplines to create a robust solution. But TKM’s major breakthrough in recent years has been the extensive integration of Machine Learning to support its customers’ business intelligence.

Whether it’s technological, competitive or start-up intelligence, TKM addresses a wide range of issues for users in all fields by providing them with Specific AI (and the performance it enables) at the price of generic AI.

Thanks to its ability to learn and continually improve on the basis of user data and usage, this technology enables TKM to offer a cutting-edge monitoring solution that saves precious time, enhances the value of the in-house monitoring profession and enables analyses that were simply unimaginable before…


IPMetrix: the ultimate monitoring tool

IPMetrix reflects this technological expertise. Designed to be both powerful and easy to use, it offers companies the opportunity to conduct effective monitoring without requiring in-depth expertise. For those looking for an even more customised solution, TKM’s team of specialists is always on hand, combining their know-how with the power of the tool.

More than just technology watch

While TKM and IPMetrix shine particularly brightly in technology intelligence, their usefulness does not stop there. The approach adopted by TKM is flexible enough to be relevant in other areas such as state of the art, industrial property strategy and even market research. This makes TKM a must-have solution for all companies seeking to remain competitive in an ever-changing world of information.

Exploiting information to its full potential

With the Internet, we were promised information at the click of a mouse, but what we got were mountains of data and new filtering and analysis challenges. However, thanks to innovative technologies such as Machine Learning, it is now possible to navigate effectively through this deluge of information. Tools such as IPMetrix can simplify and accelerate the intelligence process, making it not only less daunting but also infinitely more accurate.

In the future, the need for effective intelligence will only increase as the volume of data continues to grow. But with tools like IPMetrix, businesses are better equipped to meet this challenge. By combining machine learning, data mining and other advanced technologies, solutions like IPMetrix will usher in a new era of intelligence, where information will not only be accessible, but also exploited to its full potential. Would you like to find out more about our tool? Contact us.



Author y.belrhiti-alaoui

More posts by y.belrhiti-alaoui