Site navigation

Artificial Intelligence in Retail: An Emotional Chat Bot Example

Pelin Ozbozkurt


Artificial Intelligence in Retail

Pelin Ozbozkurt PhD, Data Science Leader of Oracle EMEA, offers up some insights into artificial intelligence in retail ahead of Oracle’s ‘Future Ready Data Management Forum’ later this month.

When it comes to artificial intelligence (AI)/machine learning (ML), retail is one of the most innovative industries. Growing customer expectations and the competition within the market force them to be so. Retailers are looking for new ways of using AI-enabled technologies to enhance customer experience and operational efficiency.

Some of the AI-related examples in the retail domain are self-driving trolleys following customers at shops, virtual assistants that can analyse facial expressions, robotics that optimise warehousing and picking/packaging operations, drones that decide the best route by handling unexpected cases (like traffic, weather etc.) and intelligent chatbots to improve customer interactions.

In this blog, I would like to mention an innovative chatbot example that combines various technologies (image processing, sentiment analysis and machine learning) to create a more personalised product recommendation engine.

An emotional chatbot!

Wouldn’t it be great to have a free offer that makes you smile when you feel a little bit down? or happy suggestions when you feel excited and ready to discover? This is what an emotional chatbot aims to do. The objective is to improve the efficiency of the recommendation by considering not only behaviour, profile and historical transactions but also the emotions.

The bot can predict real-time customer moods either from a selfie or a text (depends on the customer choice) and connect that information to the emotional score, calculated by ML algorithms, which shows a customer’s emotional connection to the company.

This example falls under the concept of Emotional AI – in other words, affective computing. Emotional AI is a well-known concept that combines data science with psychology. The reason it gets more attention today is the technological improvements in the AI era, as well as the increased availability of the new data sources.

It has been shown that emotionally-connected customers, if identified correctly, create more revenue compared to loyal only customers. With the technology today, it is possible to detect emotions by using face recognition, voice analysis and text mining and convert them into numerical information.

Today, most of the retailers have product recommendation engines. However, it is up to them whether to use AI technology and enhance the efficiency of the recommendations or not. Understanding the real-time mood of the customers and converting emotions into numerical data can help to generate more efficient and timely offers.

Moreover, it is possible to combine emotional AI concept with other analytical approaches. For example, if an emotionally connected score calculated by ML is merged with social influence score, calculated by graph theory, that indicates how influential a person is in delivering a message within his/her social network. From there, it may be possible to create a much more efficient product recommendation engine compared to the traditional approaches.

Machine learning, conversational bots and graph analytics are all pieces of a big toolset. Outputs of one piece can be used as the input for the other piece. This is the flexibility of data science comes with the available technology. Retailers, as wells as other industries, have been looking for an integrated data management platform with AI/ ML capabilities to get the best value out of an embedded approach.

However, technology itself is not enough for that. An innovative team with appropriate skill set is a must to have as well! That is one of the reasons of why we see more investments in AI teams today. If you are a retailer that aims to have a more innovative and competitive business strategy in the AI era, there is almost no limit on what you can do with the data. All you need is an integrated and powerful data management & data science platform, a creative team equipped with the right skills set and an idea to start with.

You can hear more insights from Oracle’s data athletes for free at the Oracle Future Ready Data Management Forum in Linlithgow on 25th September. You can hear from a packed agenda with topics like the UK data economy, bringing Data Science to the masses and the Oracle and Microsoft Accelerate Enterprise Cloud adoption.

Register now to confirm your place!

Pelin Ozbozkurt

Data Science Lead, Oracle EMEA

Latest News

%d bloggers like this: