@fspaggiari

We’re currently exhibiting on the ad:tech London show floor in booth #209. If you have planned to attend today, we hope to see you there.

We’re featuring demonstration of our semantic advertising Web service that allows publishers and advertisers to develop more effective online advertising campaigns.

In particular, we could talk about a new perspective on the relationship between contextual advertising and Big Data, starting from the state of the art of online advertising in the Big Data era to the benefits of semantic contextualization for cookie-less ad targeting.

There’s still time to request an in-person meeting at the show; simply fill out this form!

@fspaggiari

The world’s most innovative technology companies will demonstrate their newest offerings and engage with marketing leaders throughout the Advertising Week Experience (AWE) which will be housed at The Times Center and staged concurrently with Advertising Week IX in New York City October 1-5 2012.

Advertising Week is the world’s premier annual gathering of marketing and communications leaders each year in New York City and ADmantX will be the only contextual data provider showcased during the show. We are proud to have been selected and would be delighted to meet with you there.

In the meantime, you can learn more about us reading this blog entry that came out recently on the AWE social website or listening to this podcast, a really enjoyable conversation between Doug Zanger from the AWE and my CMO at ADmantX, J. Brooke Aker.

 @BrookeAker

Many current contextual ad targeting approaches claim to use semantics as their underlying processing approach.  But there are many varieties in applied semantics.  One such approach most commonly used is a keyword & statistics approach used by Autonomy.  The second is a Natural Language Processing (NLP) approach used by ADmantX. This document does a comparison between two vendor’s approaches that typifies the differences and argues that the incremental benefits of a NLP approach out way any additional costs.

Keyword based systems combined with statistics go from stemming (eliminating the last part of a word in order to work with the base form) to simple linguistic operations (finding the terms with standard patterns of inflection). The use of a thesaurus is also possible – archives of keyword relations. The statistical methods then aim to catch the keyword relations using mathematical algorithms of probability (Bayes theorem and Hidden Markov Chains being the most prominent); because of their objectives, sometimes these are inappropriately labeled as semantic technology or natural language technology.

Classification is done through Bayes’ functions, therefore creating values during the training phase and when in use, calculating the probability of a group of keywords which are “the closest” to a training group with respect to others.  Entity Extraction is done through the use of regular expressions: logical operations on characters sometimes assisted by Markov Chains.  Limitations exist in the approach in constructing the fullest expression of language logic often referred to as a semantic triple (subject-predicate-object) used in establishing sentiment, direction of action, motivation, intention etc.  The limitation is due to the variety in human expression that current probabilistic models cannot capture.

Probability based methods suffer from a zero sum problem in terms of performance.  That is to say when an increase in precision is needed it always comes at the expense of recall, and vice versa.  Probabilistic systems draw “boundaries” between documents relevant and retrieved documents.  The boundaries are ridged in the sense that that cannot be adjusted for single or small groups of documents or other special conditions.  They change probabilities for all documents.

Diagram 1: The diagram above represents both precision and recall.  The relevant documents are to the left of vertical line.  The documents inside the oval are the retrieved documents – the documents the system thinks are correct.  The proportion of documents not retrieved but should have been relative to those correctly retrieved is the precision measurement and marked as arrow 1.  The proportion of documents not identified as relevant that should have been relative to those correctly identified as relevant is the recall measurement and marked as arrow 2.

In a probabilistic system when work is done to increase precision by increasing the number of relevant documents by definition the recall worsens as seen in the Diagram 2 below.

Diagram 2: Changes that improve precision worsen recall e.g. ratio 1 improves but ratio 2 worsens.

Conversely in a probabilistic system if work is done to improve recall by increasing the number of retrieved documents the precision suffers as is shown in Diagram 3 below.

Diagram 3: Changes that improve recall worsen precision e.g. ratio 2 improves but ratio 1 worsens.

@fspaggiari

In a few days, we will attend the Technology For Marketing & Advertising conference in London where we will launch a new feature that allows our users to directly try our contextual advertising web service semantic analysis.

We’ve been testing our new “copy, paste and test it” functionality (that’s how we call it internally), and we’ve invited users like you to take part in our tests (you may have already received our invitation or read this other post.)

To test the new functionality, visit www.admantx.com, and enter a URL in the blue search box:

Once you have clicked “Go” after copying and pasting your URL…

… ADmantX will display the results of its semantic page-level analysis and contextualization, showing categories, topics and emotions present in the page content.

What do you think? Contact us or request a demo now if you would like to discuss your experience with us or to understand more.

If you are planning to visit the TFM&A show, we hope you will join us on the exhibit floor in booth #213!

 @BrookeAker

BrandTrust CEO Daryl Travis coined this phrase- The Un-Known Thought - in a short talk I heard him give here.  He talks about the high percentage of our everyday activity that we do “automatically” without thinking.  We are, he says, creatures of patterns – looking only for anomalies to react to.  We need to do this since the amount of information flooding into our brains everyday would otherwise overwhelm us.  So we know things without thinking about them, hence the Un-Known Thought. 

Daryl also makes the case that emotions are a human adaptation to processing information.  He suggests the following;

We are actually driven by feelings not facts.  The non-conscience part of the brain actually operates more on the basis of feelings, patterns, mental models, rules of thumb, and heuristics.  Ultimately we are not thinking about everything that we think we are thinking about. 

Advertising’s best weapon is a tug at the emotions of the buyer.  This is well known.  Yet in the last 5 years or so the mathematical, rational, measurable pattern of behavior has driven the growth of online advertising.  Emotions got left in the dust. 

Not so at ADmantX.  We look at the content of a page where ads will be placed and ask some simple yet profound questions.  Like; “What does the reader feel when they read this?”  Or; “After reading this what would the reader do next?”

Using the science of semantics we answer these questions.  The result is a series of improved, expanded and corrected tags about a page that can be used to match far more impactful ads with.

 @scagliarini

I am fascinated by the changes that are happening in the publishing industry. I am fascinated because this once very stable industry had to go through one of the more disruptive changes since the industrial revolution, but I am also inspired by them because this abrupt change has opened many opportunities for entrepreneurs (technology and non) to implement or drive even more change.

These changes will derive from a combination of technology (hardware and software) and a new way of doing business. As a semantic technology and online advertising professional I, obviously, have opinions on what should be done to turn this into a win-win proposition for all the players in the market place.

The debate in the publishing sector today is particularly confusing. To simplify it, you can say that on one side, you have the traditional players that are under a great deal of pressure to try to retake a leadership position, and financial profitability, after having allowed Google to almost destroy their business model and significantly weaken their competitive position. On the other side, you have the new online-only players that are forced by their somehow unexpected success to continue to innovate to ensure that they will be able to offer a unique and different experience to their readers.

Both sides see semantic technologies as strategic because the applications deriving from them can help publishers:

a)      increase revenue by improving the user experience and providing a more effective way to serve advertising, and

b)      reduce costs by automating the work of content creators to let them focus on the most valuable part of their job (creating content) instead of wasting time in low-value activities like manually tagging content to make it easier to search and access by users.

In any case, the change this industry is experiencing cannot go unnoticed, even if you don’t deal directly, like I do, with semantics or online advertising, as we all are online newspaper readers and Web users.

Let’s consider a couple of examples on what semantics can do for publishers and brands that are wondering how they can launch more effective or less invasive online advertising campaigns.

Semantic technology can enable the creation of advertising campaigns linking, for example, pieces of content automatically to other content, either to another article or to links to real-time feedback or opinions submitted by users. Advanced semantic technologies enable the brand to use many criteria (topics, motivations, feelings, emotions) to create this link. In this way, brands can be more creative and increase the possibility to develop unique messages without compromising the reader’s experience.

For example, an advertiser for a new trendy luxury car could link the online ads not only to content about cars but also to topics of interest for the target audience (i.e. finance, golf, holiday resorts) and to other emotional or behavioural concepts like “success,” “modern” and “wealth,” attributes that are independent from the actual topic covered in the article (“success,” in fact, can be linked to sport, politics, economy, etc.)

Semantic technology also allows  content creators to make their content automatically available in the standard formats required by the semantic web. This makes it possible to immediately activate all the described features and to reduce costs so that journalists and bloggers can spend more time on other, more important activities (like creating content instead of normalizing the content already created).

I’m convinced that this revolution will continue to bring about historical change, and that conditions exist that will make it possible for big publishers and small innovators to live and compete together in a transparent, fully functioning market. To make this possible, big publishers should not focus only on big legal battles to reverse what they made possible (the availability of virtually unlimited amount of free information) or on neverending discussions on how to reinstate a pay-per-view model. It is necessary and cheaper for them to understand and take advantage of the opportunities provided by new technologies (hardware and software). The profit that could be made by applications that already exist or for now, only exist in the minds of some innovators…are potentially much larger than we could imagine.

 J. Brooke Aker

The Semantic Web Summit East wrapped up just a couple of days ago. It was a great mix of business oriented Semantic Web believers. We explored use cases and heard how companies are now moving from pilot to production in their use of semantic technology. On Wednesday, ADmantX announced itself to the world and made the case that semantics and advertising are a huge opportunity. In fact, last Tuesday at Web 2.0 in San Francisco Mary Meeker, Morgan Stanley analyst, predicted online ads to hit $50 billion shortly. Funny to think the digital ad world could go so high on such brittle and imprecise technology that matches the ad to the content. We’ll be changing that – starting Wednesday !!