Last week I laid out my understanding of five concepts and technologies that are on the up according to the Gartner Hype Cycle for data science and machine learning. Here are the remaining six that are getting the booster treatment.
Cloud-based workflow app Tallyfy said that human-in-the-loop process takes place when a machine or computer system needs human intervention because it is unable to come up with a solution to a problem. This happens when a computer model is not very confident in the accuracy of its judgement. The crowdsourcing element means that many humans are intervening. US-based CrowdFlower has based its business model on human-in-the-loop crowdsourcing. It allows businesses to access a crowdsourced workforce that would, for example, be able to tell a self-driving car whether someone in an Oscar the Grouch costume is a person or a dustbin.
Q Gate, a CRM consultancy, said there are two types of business intelligence systems - guided analytics and self-service. It added that guided analytics packages allow the company to set up prepared business applications featuring dashboards, charts and calculations with the help of a developer. According to Calero, guided analytics returns reports that are a mix of visual and functional detail and is pre-designed to solve a specific business problem.
These are marketplaces for developers to get algorithms, functions and models - essentially, hubs to find and access algorithms, functions and microservices. Algorithmia is one such marketplace where you can buy and sell algorithms. Gartner itself explained that algorithm marketplaces allow algorithms and other software components, which can be used as building blocks within tailored solutions, to be brokered.
Data science notebooks
The people over at Silicon Valley Data Science said using notebooks makes their data scientists happy and is better than just working with data science tools, like R and Python. This is because the code in a notebook can be executed and results displayed as part of the page. They can be saved as files, easily shared and they can run anywhere due to their browser-based user interface. The Apache Zeppelin notebook can work with Spark, as well as Scala, SparkSQL and visualisation tools. Jupyter is said to be the most influential notebook and can support Python, R, Scale and lots of other programming languages.
Artificial general intelligence
Artificial general intelligence has its own society that describes AGI as an emerging field aimed at the building of “thinking machines”. According to Popular Science, there isn’t a set definition of AGI as some say that AI would need real consciousness to be considered AGI, while others say a simulation of it will do. Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence suggests that a computer with the ability to do what a three-year-old human child can do would be in the realms of AGI. Global Futurist stated that, one day, AGI will be the basis for a new breed of machines with intelligence that will rival humans and the ability to carry out intellectual tasks.
This is the analysis of raw, unstructured conversational data from sources such as virtual assistants (like Siri and Alexa), bots, live chat, call transcripts and emails, according to Artificial Solutions. It means that conversational data alongside context, metadata and other criteria can be summarised. As a result, conversational analytics can detect both sentiment and emotion. The CEO of bot analytics platform Dashbot said that conversational data is richer and more actionable than traditional analytics. It seems that it is very important to developers and creators of bots or text or voice-based conversational interfaces.
In theory, becoming familiar with these concepts will help you to be ever so slightly ahead of the curve.