All Posts By

David Amerland

A quick introduction to Machine Learning

By | Artificial Intelligence, Google Plus | No Comments

Understanding Machine Learning

Machine learning (also popularly called “Deep Learning”), cognitive computing and AI are words that now feature in our daily vocabulary. Yet understanding what they actually mean is not quite as common as you might think.

The closest misapprehension we get is that Deep Learning is some form of artificial intelligence and computers are networked in ways that are like a human brain. Deep Learning is a very weak form of artificial intelligence where a complex algorithm is trained to recognize patterns during a supervised stage which then leads to greater confidence of accuracy in its unsupervised stage.

Considering that I just introduced two new terms trying to explain the first one it will probably be best if we take things from the beginning. The real breakthrough in machine learning has been in the use of convolutional neural networks (CNNs) which are networks of processors (the neural nets in question) structured in an overlapping way where processing redundancies are introduced in an attempt to correct pattern recognition mistakes on the fly and save computing cycle time.

Convolutional neural networks were initially used, with great success, in visual image recognition and have been inspired by the way the human brain processes visual information, but they have since found uses elsewhere, anywhere as a matter of fact where patterns emerge that can lead to a greater understanding of what is being indexed.

When we talk about machine learning we basically imply the use of a convolutional neural network of some kind and an algorithm that goes through two stages: a supervised stage where a human operator helps the algorithm go through masses of data and adjust its understanding of it and then an unsupervised stage where the algorithm is left to work on its own with occasional quality control intervention by a human, as warranted.

If this sounds familiar consider semantic search and Google’s human rater guidelines used to train its algorithms. Similar logic applies to Google’s Voice Search, YouTube recommendation engine and Google’s Image Search.

Deep Learning is not really ‘Deep’

With all these terms out of the way the question is do machines exhibit learning behavior and are they truly intelligent? These are two very different sets of questions. The term “Deep” as applied to “Deep Learning” has often been misinterpreted to mean learning of the same type that the human brain exhibits and that is not what is happening there.

Deep Learning refers to the convolutional neural net architecture where more than two layers of neural networks are involved. Neural nets these days can go pretty deep and what sets them apart from an ordinary computer network is that each stage of the neural net can actually be trained. The intelligence aspect of the behavior is also the result of the network’s architecture where a memory is attached that remembers the results obtained which means the neural network apportions resources to ‘learning’ new stuff by recognizing new patterns. Where the new patterns are derivatives of the patterns it already knows. An example of this is teaching a neural network what a dog is by looking at a few thousand pictures of Alsatians, Labradors and Poodles and then allowing it to understand that “dog” refers also to Pit Bulls, Dobermen and Chihuahuas.

This is not true intelligence in the way humans exhibit it, though within very narrow contexts it’s close enough, especially if a recursive neural network is created where the computational outcomes are fed back into the network (which is why Google Voice Search can hold a conversation with us on Barrack Obama’s marital status and children, without our having to specify the context every time).

Machine Learning Everywhere

Moving forward we will see the use of machine learning everywhere. The uptake is driven by the same reasons that made semantic search a no-brainer: lower cost implementation, greater savings and greater reliability in complex results.

And Google have just opened sourced too:

This will mean that our devices and apps will become ‘smarter’ in the sense that they will now be more responsive to the environment and more focused in the context of our needs when we interact with them. Smarter devices means we can now make better use of information and, as neural networks spread everywhere, better informed decisions and choices.

It is our ability to use information in such a smart way that actually makes us, humans, intelligent.

How “Keeping it Real” Became the Next Thing In Marketing

By | Google Plus | No Comments

Marketing has never been about lying. But that does not mean that it is there to tell the truth, either. Because we make purchasing decisions based on emotions rather than logic, marketing has always been there to help push the right ‘buttons’ to make us purchase.

In the offline world with its centralized authority figures, marketing pushed boundaries that sometimes required advertising standard authorities, the government or the police to intervene. Thus we were protected. If things went wrong because of a marketer’s successful manipulation of our emotional triggers, we could be certain to find some redress the moment we complained.

You see in the offline world, marketers, just like centralized authority figures, are real. They have reputations that can be burnt (always costly) and corporeal forms that could be punished. The moment marketing reached the web, however, things changed. In the decentralized, authority-figure free world of 1s and 0s there was a sense that anyone could be anyone and do anything. The freedom was exhilarating, the ability to reinvent oneself electrifying but it also led to a distinct lack of trust.

Without trust there can be no relational transaction of any kind. Clearly, on the web, relational transactions take place all the time. We consume information, news, articles, videos, images and we all make purchases of one kind or another. Analysis of each time we do so shows a relationship of trust has, inevitably, been formed.

But that is not really what it’s all about. Left to its own devices to work this way, the entire web ecosystem could easily fragment into enclaves that would mark perhaps ‘safe zones’ areas where the proverbial wagons have been circled and those inside them know and trust each other, and the badlands of the ‘wild web’ where trolls and tricksters roam and in which no decent netizen should venture after dark, alone.

Semantic Search and the Question of Identity

Of course, the web has not quite been left to its own devices. It is being mined, indexed, categorized and archived by search engines designed to make it easier for us to navigate its vastness. This is where things get interesting.

Semantic search does not only index information, it actually understands it in relative importance to its context. It then goes on to understand the relative importance and context of those who share it. For me, for instance, to share a link leading to my book, Google Semantic Search, creates a moment of transparent intent that deprecates its value: it is my book, I am bound to push it, in order to help its sales. But an independent review, a mention, a share by others, creates moments of different context.

Taken all together, all these moments, create their own sense of gravity that can, depending on their sentiment, greatly enhance the sales of my book (and my reputation) or damage both. And that is really what’s so different. When marketing is no longer controlled by marketers no amount of advertising, social media posts, blogging and paid ads is going to change the fate of a product or service that fails to be valuable to the end user.

Trusting strangers with your fate, however, requires something more than just having a great product to sell or a needed service to offer. People, primarily, do business with people. Whether you are a brand or a small business or a single individual who has created a brand value around their name, the challenge is the same for everyone: how do you make real contact in the first instance? The kind of contact that leads to a mutual sense of understanding and trust developing?

The answer is also the same for all: through making contact at a very human level.

The web never sleeps. And it never forgets. It becomes therefore, next to impossible to try and act like there are airtight compartments in which you can keep all the different ‘messages’ of your brand. It is only when you come out from behind the façade that you begin to realize the power of the human connection.

By establishing who you are through:

  • Picture sharing
  • Posts
  • Blog posts
  • Videos
  • Podcasts
  • Everything else you can think of here

You begin to join all the dots of your digital existence into one coherent whole, and as that whole gets more and more fleshed out, so do trust levels begin to go up. So does the kind of marketing that helps the picture to be fleshed out and trust levels to go up, increase.

It’s a relatively simple cause-and-effect chain that has a tremendous impact on the way we do business.

Structured Data and Semantic Search Visibility

By | Google Plus | No Comments

When it comes to semantic search (and we should really be thinking about it as just search) the question of structured data introduces an oddity. Really semantic search is all about structured data. Google’s bot indexes the web, collecting everything it finds, most of which is in the form of unstructured data, and Google then creates a structured data image of that inside its index.

Search then is almost like a brain, applying a thin but critical layer of order upon the chaotic mass of information that gets dumped upon the web every second of every day. Within that context everything that we can do to help make the indexing process easier is a plus.

Structured data is a W3C validated way of annotating the content you produce on your website so that search engines can better understand what it is and how it fits in with everything else they have indexed across the web. Because the application of structured data requires some knowledge of programming in order to implement correctly it has always been left to those who actually can code, to work with.

This raises an interesting question. If you can’t code or if you create content using a platform like Google’s blogger or Word Press or if your website is ran by a dynamic content management system like Joomla or Drupal is there anything you can do to help Google understand your website’s content better?

The answer, obviously is “yes” (otherwise this article wouldn’t make much sense). It does however require some planning and forethought and it will also require some discipline as you move forward. So if you really want your website’s content to feature in Google’s search here are the steps required:

Step 1.

Create content that answers specific search queries. Rather than simply having content that is there to feature a keyword or two, think of what your online visitors really need to know about what your website does and focus on answering the questions they really have about its services or products.

Step 2.

Have depth in every article you write. Link to additional, external, authoritative sources that broaden the scope of your articles and further enrich the experience for your online visitors. Point to other pages on your website only when the content found there is relevant and adds further value for the online visitor who is reading the current page.

Step 3.

Use hashtags. Whenever you share your content in social media networks like Twitter and Google+ use hashtags that help quantify it. Be consistent in your use of those hashtags and make sure that they adequately reflect the content being shared. An article like this, for instance might appear under #marketing but really that’s a very generalized category. It would be much better to present it under #semanticsearch.

Step 4.

Create ontologies in your content. Ontologies are groupings of content that share a common identity of purpose. They help create a sense of order, depth and expertise in your website content.

Provided you take the care and attention required to meet all these points in your content creation plan, then your website will stand a good chance of being indexed properly and rank highly even if you have zero structured data markup in your implementation of its code. More than that it will appear exactly when it’s supposed to meaning that it will be found by those whom its content can benefit the most and that is the best win you can wish for on the semantic web.