What is web intelligence? To the uninitiated, it probably sounds like something used exclusively by groups like INTERPOL and CSIS. For researchers in the areas of artificial intelligence (AI) and information technology (IT), it represents an exciting field of study. To professionals accustomed to tech and business buzzwords, it may seem to have no meaning at all. That’s what happens when you’re inundated with overused terminology. It all starts to sound the same.
The concept of web intelligence has many different interpretations, but this post is about how we see it at the Gnowit office (at least, it’s about how I see it, based on my ongoing conversations with experts in the field). Let’s start with a bit of background.
The roots of web intelligence
As a field of scientific research, web intelligence (WI) is relatively new. According to Wikiepedia (not always reliable, but the site echos authoritative sources in this case) WI “explores the roles and makes use of artificial intelligence and information technology for new products, services and frameworks that are empowered by the World Wide Web.”
What does this look like in practice? Developments in the WI field of information retrieval have lead to technologies that automatically “crawl” the web for useful, high-quality pages (this can greatly improve the accuracy of search engines). The area of web mining (finding patterns in large quantities of web data) can be used in targeted online marketing, government security, and a whole host of other applications. And one of WI’s biggest focuses – personalization – can steer users towards web content they’ll actually relevant.
The key takeaway here is that, far from an abstract concept, web intelligence is extraordinarily useful in the real world. Plus, thanks to recent buzz surrounding artificial intelligence (AI), it’s in demand. Just as some buzzwords represent truly valuable ideas, some frequently-discussed technologies are deserving of the hype.
The media-monitoring connection
The previous section touched on some of the ways WI can help businesses and internet users exploit relevant online information. These types of solutions have become crucial in recent years, as the democratization of the internet has lead to unmanageable volumes of published content. This content shapes public perception, which can lead to threats and opportunities for companies, associations, and nonprofits. Automated media-monitoring systems have emerged as a way of keeping track of this information, but they’ve suffered from some major limitations.
Fortunately, research is leading to the development of technologies that allow for more comprehensive monitoring and increasingly-accurate results. For example, sophisticated internet crawlers make it possible for monitoring systems to improve themselves automatically. While traditional systems generally monitor a set number of sources acquired from third-parties, advanced WI systems seek out relevant, high-quality pages to add to the sources they monitor. This means a much larger pool of sources and, consequently, a much lesser chance of missing vital information.
But these benefits go further. Because systems backed by WI research are more flexible, they’re able to accommodate not only more sources, but sources of many different types. Trade journals, blogs, transcripts published on government websites – these are just a few examples. In addition, AI research has made the relevant information from these sources easier than ever to digest (by providing analytics, context, accurate summarization, and more).
These developments have expanded media monitoring to the point that, often, the descriptor is no longer accurate. A new category is now required. At Gnowit, we call it “web intelligence” – not only because of the science behind it, but because it’s an accurate descriptor. Intelligence is solid information that can guide strategy and big picture decisions. It’s also privileged. As a still-young and developing field, WI has been most attractive to organizations with innovators at their helms, which has resulted in its use being considered a “trade secret” by many.
The web-intelligence advantage
When it comes to gathering and analyzing intelligence, a comprehensive approach is best. Web intelligence fosters a more complete awareness of the threats and opportunities facing an organization, as well as an understanding of the larger landscape in which it operates. Below are some of the major advantages of developing this type of strategy.
A lowered risk of missing relevant content
This point is addressed above, but it bears repeating. Systems capable of automatically seeking out relevant web pages acquire source pools many times greater than those that were used in the past. Careful source selection is crucial, but having a comprehensive set of choices to start with can make all the difference.
More accurate information
Quick (better yet, instantaneous) methods of narrowing down massive volumes of information can result in leaps in analyst productivity. As previously mentioned, finding the most potentially-relevant sources on the net to gather information from is a big (and complicated) part of the equation. But filters play an equally important role.
Filters powered by artificial intelligence – such as those capable of, for example, removing results that contain the word “apple” when it’s used to describe a fruit and keeping those that contain references to the brand – are the gold standard of accuracy.
The usefulness of intelligence is greatly increased when it can be acted upon immediately. Usually, a good deal of analysis and investigation is required before an organization can take action based on a piece of information. But consider the potential drawbacks of inaction (public relations crises, missed leads, etc). The need for quick turnaround becomes obvious.
Receiving “smarter” communication can go a long way in this respect. Topic detection, sentiment analysis, trend visualization – these are just a few features that can make the information you receive easier to understand. In some cases, users can comprehend an entire set of results at a glance, greatly reducing the time between information acquisition and action.
As we see it, the degree of sophistication determines whether the communication you receive through a monitoring and alerting system constitutes web intelligence.
Last but certainly not least: web intelligence systems are flexible. They’re designed with the realities of the digital world in mind. Obscenely large quantities of data, low-quality published content, web pages with poor or complex navigation – a system that provides truly useful intelligence must be capable of dealing with such issues. One advantage of these systems: they can be adapted quickly.
Users may need to monitor sources of a new type, filter content by a specific criteria, or have results automatically translated into another language. A resilient system can be tweaked to meet the unique monitoring needs of most organizations.
It’s all about the user
So, what does web intelligence mean at Gnowit? From my point of view, the science is just one aspect. Ultimately, it’s all about the user and her experience. It’s about source selection. It’s about access to precisely the information you need – no more, no less. It’s the ability to understand large quantities of information at a glance. It’s superior engineering for better results. That’s worth getting excited about.
Feature Image: Surlan Soosay