Towards smarter data - accuracy and precision

sasha uritsky

Nov 01, 2014

There is a huge amount of information out there. And it is growing. To make it efficient and increase our competitive advantage we need to evolve and start using information in a smart way, by concentrating on data that drives business value because it is accurate, actionable, and agile. Accuracy is an important measure that determines the quality of data processing solutions.

How accuracy is calculated?

It is easy to do with structured data, because the requirements are formalizable. It is less obvious with unstructured data, e.g. a stream of social feeds, or any data set that involves natural language. Indeed, the sentences of natural language are subject to multiple interpretations, and therefore allow a degree of subjectivity. For example, should a sentence ‘I haven’t been on a sea cruise for a long time’ be qualified for a data set of people interested in going on a cruise? Both answers, yes and no, seem valid.

In these cases an argument was put forward endorsing a consensus approach which polls data providers is the best way to judge data accuracy. This approach essentially claims that attributes with the highest consensus across data providers is the most accurate.

At nmodes we deal with unstructured data all the time because we process natural language messages, primarily from social networks. We do not favor this simplistic approach, as it is considered biased, inviting people to make assumptions based on what they already believe to be true, and making no distinction between precision and accuracy. Obviously the difference is that precision measures what you got right, and accuracy measures both what you got right and what you got wrong. Accuracy is a more inclusive and therefore more valuable characteristic.

Our approach is

a) to validate data against third party independent sources (typically of academic origin) that contain trusted sets and reliable demography. Validating nmodes data against third party sources allows us to verify that our data achieves the greatest possible balance of scale and accuracy.

b) to enrich upon the existing test sets by purposefully including examples ambiguous in meaning and intent, and providing additional levels of categorization to cover these examples.

Accuracy is becoming important when businesses move from rudimentary data use, typical of the first Big Data years, to a more measured and careful approach of today. Understanding how it is calculated and the value it brings helps in achieving long-term sustainability and success.

Interested in reading more? Check out our other blogs:

How AI is changing the work landscape

"For better or worse, robots are going to replace many humans in their jobs,” says analysts from BBC, and the coronavirus is speeding up the process. Consumer preferences are evolving and recently consumer behavior demonstrates that we as a society have become more tolerant accepting of using automation in our daily routines.

In the professional workspace, most if not all companies have moved towards working from home. Given the unprecedented times, recruitment, the employees management, and the corporate governance processes and communication have moved online. As a result of pandemics many companies are experiencing hiring freezes, but many others have moved their recruitment efforts online. A few companies have begun piloting recruitment with the help of artificial intelligence. They are now leveraging AI to conduct online interviews and assessments and deliver data back to the employer. Now more than ever, companies are realizing the importance of moving towards a remote-friendly workforce. Being able to scale human capital on a larger scale online has definitely been accelerated recently.

I know for myself, as a current student who recently had their internship offers rescinded due to COVID-19, I’ve put myself back into the market. I’ve seen both small businesses and corporations utilize screening questions, video pitches, and unique riddles to test students’ critical thinking and how they fit into the company culture. This experience in itself has been revealing – after so many years of in-person interviews to suddenly having to emulate the same energy online or via video. Given the adjustment, at times it definitely felt unnatural to sit in front of my computer camera and pitch myself or answer video questions. However, going forward, I can see how automation and online platforms will become more explored given the time it saves and the bias it could remove during the recruitment process.

Yet it is not just a change in the recruitment process that we are seeing. The customer service environment, as I have seen first-hand, is under large stress. One of the first calls I had made was to an online retailer, to try and put in a return order. What seemed to be an idea that everyone else had as well, I was put into a queue that lasted more than 30 minutes. After hitting that 30-minute mark, I gave up and put off the task for a later date. Now, a month later, more and more companies are adopting chatbots and artificial intelligence into their customer service processes. These companies are beginning to provide information in a more efficient manner, and with less human capital.

Moving forward, in the next few months and post-COVID-19, it would be interesting to see which companies are focusing more on their digital transformation efforts. I believe that a larger number of universities and educational institutions will partner with tech companies to help digitize their working environments. And private businesses will continue to implement some of the already existing practices and produce products that cater to the remote working lifestyle and online interactions.

Microsoft AI products

Microsoft product strategy has always been and still remains that of ‘zero alternative’. Their ultimate policy is for their customers to have no choice but to embrace only Microsoft products. Consequently they created and are offering products and solutions in (almost) every segment of IT enterprise and consumer market, including, but certainly not limited to, their own data base, their own cloud services, operating system, office tools, programming language, and many more.

Not only do Microsoft offer wide variety of products, they tie them up together in a unified ecosystem that makes it easy for components to connect and interact. At the same time, this ecosystem is hostile to non-Microsoft products.

Microsoft strategy for the burgeoning, fast growing AI segment is similar:

Create products to address all parts of the AI market, add them to the ecosystem to ensure easy compatibility from within and difficulty of use from outside.

Currently the products on offer are:

- Microsoft AI engine, called LUIS. It is supposed to compete with other major industrial AI systems such as IBM Watson, and has similar training methodology. It offers webhook interfacing via endpoints.

- Microsoft chatbot building platform, called, surprisingly, Microsoft Bot Platform. It addresses the popular demand for easy chatbot design and provides seamless connectivity with main user interfaces, such as web interface, SMS, mobile, and messaging platforms.

- In addition Microsoft offers their own messaging platform in Skype.

The main advantage of using Microsoft AI products is the built-in connectivity with user interfaces.

The main disadvantage is in their ‘zero alternative’ policy - once you’ve chosen a Microsoft product you are likely will be forced to choose only Microsoft products for the duration of your project.