Heartbeat Ai 2020 Year In Review
With one Quarter left in 2020,
This year's impact on global and national norms has already solidified change for years to come. Every year has defining moments, but 2020 has been packed with paradigm-shifting events that have changed all of our lives forever.
It’s hard to believe that on January 31st, the global media cycle was being dominated by Britain becoming the first country ever to leave the European Union and COVID-19 had less than 10,000 cases worldwide (7 in the US).
Fast forward to March, COVID-19 cases sky-rocketed to 858,340 cases worldwide and 188,724 cases in the US - leading the World Health Organization to declare a global pandemic. Markets around the globe faltered, spiraled and crashed, causing months of economic suffering and unemployment levels to reach record high.
Going into April, the term “Social Distancing” recorded its highest ever search volume around the world, and the shift to a “new normal” started to formalize. Face masks became a usual sight, transportation ground to a halt, and city-centres became dystopian-like ghost towns.
Entering October, COVID-19 has eclipsed 35 million cases and 1 million deaths worldwide, with the United States accounting for 7.5 million cases and 211,000 deaths. With the potential for a major second wave, and US election in a matter of weeks, social and economic forecasts are uncertain at best.
Why Is Emotional Intelligence More Important Now Than Ever?
While some jobs have always come with a heavy emotional toll, like healthcare workers on the front line, or a police officer in a gang-heavy district, the shift to the “new normal” has forced almost everyone into areas of discomfort and uncertainty. Whether you are now working from home until an undefined date or lost your job altogether - individuals' stress levels are extremely high. Concurrently, those of us on the frontline of service industries are being subjected to more “emotional labour” than ever before. COVID-19 has shifted the goalposts, requiring service-focused employees to not only have to manage the stress of the duties of their positions, but also the stress of the public they are required to interact with.
However, studies suggest that emotional intelligence (EI) can aid our well-being by helping with both burnout and wellness. That is why it’s more important now than ever for organizations to understand their users, customers, employees, patients etc.
With Endless Data - Emotional AI Is The New Frontier
Every day we produce over 2.5 quintillion bytes of data, and 90% of that data is unstructured data, which is mostly made up of different text inputs.
Technology has quickly become an avenue for companies to gather an even higher volume of data on users, and concentrating on the emotional aspect of customers can ensure better loyalty, lead generation, and retention. Brands can leverage emotional analytics to identify what are the unique expectations of the customers and create a customized and personalized experience for them.
AI comes in many aspects, but leveraging cutting edge technology is essential for the continued success of your business. Identifying new trends in your consumer base, increasing the quality of your employee engagement and understanding the feelings towards your brands are just a few of the many ways Heartbeat AI can empower your organization to the next level. The adaption of new technologies, specifically AI, is vital for sustained growth in uncertain times. A recent study found that 84 % of C-suite executives believe that you must leverage AI to achieve high growth objectives.
Where Is Heartbeat Ai Amongst This?
Although virtually no industry has gone unscathed, the first two quarters of 2020 saw effectively no organizations taking on new software. At Heartbeat, we utilized this time to strategize and reposition. We saw this as an opportunity to improve our service offerings, explore new avenues of analyses, and improve our team’s effectiveness.
We focused on expanding our accessibility for different industries, improving our algorithms to uncover deeper emotions in context, and developing our dashboard for greater theme identification and actionable insight extraction. But we realized...
We Needed More Human Power
Recognizing that our ambitious goals would need more human power, we hired Tobin Sydneysmith, to the position of Director of Human Experience Analytics. Tobin has a background in economics, marketing, statistics, user interface, and building data-driven strategies. He has worked with multiple tech startups across many industries and is an avid fan of the outdoors.
Tobin will be leading our growth/sales efforts, helping improve the analytic capabilities of Heartbeat’s HX platform, and improving our dashboard’s navigation and insight extraction capabilities.
Technical Improvements And New Tools
At Heartbeat, we are always striving to take our software to the next level for every client, and it is through this constant working process with our clients, those essential improvements to our tool are uncovered.
As of the beginning of 2020, we have launched our fully operational API, allowing us to connect Heartbeat’s HX analytics to virtually any source. We have already integrated the Heartbeat API with organizations like Tableau, Qualtrics and Telegram.
To Improve Analytical Ability Of Our Software…
We added a statistical significance calculator for primary emotions. Statistical significance is the likelihood that a relationship between two or more variables (in this case, emotions) is caused by something other than chance.
At Heartbeat, we utilize statistical significance tests to understand whether the differences in primary emotions and sentiments among different groups are significant. These tests help formulate better analyses and understandings of the emotional data.
Adding Multiple Ways To View Emotions
To allow for more granular viewing of our primary emotions breakdown, we added Binary Mode to our main report section. Normal Mode allows users to understand the distribution breakdown of primary emotions in the data; summing to a total of 100% between all of the primary emotions.
However, Normal Mode takes into account multiple of the same primary emotions triggered in a single verbatim*. For example, if two emotional words were triggering “Fear” in a single verbatim, both of those emotional words would be utilized in the calculation of the “Fear” Distribution in Normal Mode.
Binary Mode does the opposite. It extracts all emotional words in a single verbatim as one, allowing you to get an alternative view on the emotional breakdown of your data. Let's take our previous example, if two emotional words were triggering “Fear” in a single verbatim, the emotionality attributed to those emotional words would be aggregated and count as one in the Per Response distribution in Binary Mode. This allows for users who may have some respondents who respond at length, and others who are very brief, to not be skewed by the lengthy responses and have a more holistic view of their data.
*A verbatim is unit of text occupying one cell in one row of a data file, and has a max character count of 1000 characters
The Heartbeat Aggregator
At Heartbeat, we understand the importance of good data. We are continually creating opportunities for our clients by analyzing their unstructured text data and unveiling insights. Our familiarity with different types of unstructured data has equipped us with knowledge of how to extract data from third-party sources properly.
We created the HB Aggregator to assist our clients with extracting data from third-party sources as seamlessly, and in the best form possible. Unlike some scrappers, that extract data in bulk from third party sources, and often result in large datasets with inconsistent, unclear, and sometimes unusable data. The HB Aggregator allows you to create specific queries on third party sites, and then selectively tag posts, comments, and feedback from that specific query. Allowing you to efficiently collect well-sourced, well structured and easily-accessible data and storing them all within one secure location.
Data extracted using the HB Aggregator is tagged with the appropriate meta-data. When uploaded to the Heartbeat HX dashboard, it allows the user to sort by the appropriate metadata filters, such as source, sub source, query, time, etc. making analysis super easy and efficient!
Current Aggregator Sources
Currently, the HB aggregator has 3 possible active sources, Reddit, Instagram and Transcription. When utilizing Reddit or Instagram, you’re able create a query for either general search or post-specific, then allowing you to tag the desired comments or posts. The transcriber allows you to upload desired transcription data to be appropriately formatted. More sources will be available over time.
Our New Website
With all of the improvements to our platform and algorithms in 2020, we decided to also rebrand and launch a new website. We have expanded our service offering and the industries we cater to - learn more in our “Solutions” tab. Curious about the science behind our classification? Want to know more about how our platform helps extract actionable insights? Check out our “Platform” Section. Want to know more about the people behind Heartbeat Ai - check out the “About us” page!
Going Forward - Our Plans for 2021
At Heartbeat, we pioneer with purpose. We are innovators, leading the evolution of artificial emotional intelligence (AEI) and equipping our partners with the tools necessary for them to get the full value of their data.
We are never satisfied, continuously pushing ourselves to expand the possibilities of Heartbeat Ai. So what’s coming in 2021?
Topic Identification
While topic and theme extraction is already made easy with the Heartbeat software, it’s not automatic. Our features allow users’ to easily identify actionable insights, themes and pain points in their users journey. In 2021, we are taking that to the next level, with Heartbeat Topics. Automatically analyze the language structure of your data; and populate your dashboard with “Topics” found in your data set. The automizing of topics within Heartbeat, combined with our already-automatic emotions identifier, will create an even more powerful tool to perform robust analysis on unstructured data.
New Emotional Classification
We are always looking to improve, and we do so best by working with clients in different industries. Over the last 5 years, working with thousands of different sets of data, leading us to deepen our understanding of how clients utilize our emotional classification. This continued learning allows us to continue to develop the world's most sophisticated algorithms at understanding human language and behaviour. In 2021, we are releasing Heartbeat’s Emotional Classification 2.0. Our newest classification will see us reduce the number of secondary emotions, from 97 to 70. We are doing this to improve our accuracy and consistency when analyzing your data.
Predictive Modeling and Analytics
Predictive modelling is a process that uses data and statistics to predict outcomes with data models. These models use known results to develop (or train) a model that can be used to predict values for different or new data. The modelling results in predictions that represent a probability of the target variable based on estimated significance from a set of input variables.
At Heartbeat, we have utilized predicting modelling with our software to predict results in the 2016 US “Swing states” election with incredible accuracy. Our data is aptly suited for random forest models, which, to put it simply, builds multiple decision trees and merges them to get a more accurate and stable prediction.
In 2021, we are partnering with Function Group Analytics, to develop predictive modelling, specifically random forest models, into the Heartbeat Ai dashboard. Allowing users to not have to have a background in statistics or math to derive concrete insights into potential outcomes of their user base from their data.