As in the prior method, each class is given with some number of example sentences. Once again each sentence is broken down by word (stemmed) and each word becomes an input for the neural network. The synaptic weights are then calculated by iterating through the training data thousands of times, each time adjusting the weights slightly to greater accuracy. By recalculating back across multiple layers (“back-propagation”) the weights of all synapses are calibrated while the results are compared to the training data output. These weights are like a ‘strength’ measure, in a neuron the synaptic weight is what causes something to be more memorable than not. You remember a thing more because you’ve seen it more times: each time the ‘weight’ increases slightly.
However, if you’re trying to develop a sophisticated bot that can understand more than a couple of basic commands, you’re heading down a potentially complicated path. More elaborately coded bots respond to various forms of user questions and responses. The bots have typically been “trained” on databases of thousands of words, queries, or sentences so that they can learn to detect lexical similarity. A good e-commerce bot “knows” that trousers are a kind of pants (if you are in the US), though this is beyond the comprehension of a simple, untrained bot.
If you are looking for another paid platform, Beep Boop may be your next stop. It is a hosting platform that is designed for developers looking to make apps for Facebook Messenger and Slack specifically. First, set up your code using Github, the popular version control repository and Internet hosting service, then input it into the Beep Boop platform to link it with your Facebook Messenger or Slack application. The bots will then be able to interact with your customers with real-time chat and messaging.
Conversational bots can help a business’s customers with difficult transactions, plus collect data and give recommendations. For example, a conversational bot integrated to an airline’s website can answer questions regarding flight availability, rebook tickets, fees and suggest add-ons like hotels. Though a conversational bot may not be able to finish the exchanges, it could still be able to gather preliminary data and pass it on to the next available customer care agent. In both cases, the airline will save considerable time in its call center.
Before you even write a single line of code, it's important to write a functional specification so the development team has a clear idea of what the bot is expected to do. The specification should include a reasonably comprehensive list of user inputs and expected bot responses in various knowledge domains. This living document will be an invaluable guide for developing and testing your bot.
While messaging and voice interfaces are central components, they fit into a larger picture of increasing infusion of technology into our daily lives, which in turn is unlocking new potential for brand-to-consumer interaction. The fact is, technology overall is becoming more deeply woven into our lives, and the entire ecosystem is enjoying tighter cohesion through the increasing availability and sophistication of APIs. Smart companies are finding new and innovative touch points with consumers that are contextual, relevant, highly personal, and yes, conversational. Commerce is becoming not only more conversational but more ubiquitous and seamlessly integrated into our lives, and the way we interact with brands will be forever changed as a result.
Dan uses an example of a text to speech bot that a user might operate within a car to turn windscreen wipers on and off, and lights on and off. The users’ natural language query is processed by the conversation service to work out the intent and the entity, and then using the context, replies through the dialog in a way that the user can understand.
The sentiment analysis in machine learning uses language analytics to determine the attitude or emotional state of whom they are speaking to in any given situation. This has proven to be difficult for even the most advanced chatbot due to an inability to detect certain questions and comments from context. Developers are creating these bots to automate a wider range of processes in an increasingly human-like way and to continue to develop and learn over time.
It won’t be an easy march though once we get to the nitty-gritty details. For example, I heard through the grapevine that when Starbucks looked at the voice data they collected from customer orders, they found that there are a few millions unique ways to order. (For those in the field, I’m talking about unique user utterances.) This is to be expected given the wild combinations of latte vs mocha, dairy vs soy, grande vs trenta, extra-hot vs iced, room vs no-room, for here vs to-go, snack variety, spoken accent diversity, etc. The AI practitioner will soon curse all these dimensions before taking a deep learning breath and getting to work. I feel though that given practically unlimited data, deep learning is now good enough to overcome this problem, and it is only a matter of couple of years until we see these TODA solutions deployed. One technique to watch is Generative Adversarial Nets (GAN). Roughly speaking, GAN engages itself in an iterative game of counterfeiting real stuffs, getting caught by the police neural network, improving counterfeiting skill, and rinse-and-repeating until it can pass as your Starbucks’ order-taking person, given enough data and iterations.
The classification score produced identifies the class with the highest term matches (accounting for commonality of words) but this has limitations. A score is not the same as a probability, a score tells us which intent is most like the sentence but not the likelihood of it being a match. Thus it is difficult to apply a threshold for which classification scores to accept or not. Having the highest score from this type of algorithm only provides a relative basis, it may still be an inherently weak classification. Also the algorithm doesn’t account for what a sentence is not, it only counts what it is like. You might say this approach doesn’t consider what makes a sentence not a given class.
But, as any human knows, no question or statement in a conversation really has a limited number of potential responses. There is an infinite number of ways to combine the finite number of words in a human language to say something. Real conversation requires creativity, spontaneity, and inference. Right now, those traits are still the realm of humans alone. There is still a gamut of work to finish in order to make bots as person-centric as Rogerian therapists, but bots and their creators are getting closer every day.
[In] artificial intelligence ... machines are made to behave in wondrous ways, often sufficient to dazzle even the most experienced observer. But once a particular program is unmasked, once its inner workings are explained ... its magic crumbles away; it stands revealed as a mere collection of procedures ... The observer says to himself "I could have written that". With that thought he moves the program in question from the shelf marked "intelligent", to that reserved for curios ... The object of this paper is to cause just such a re-evaluation of the program about to be "explained". Few programs ever needed it more.