Frequently Asked Questions
Who is Receptiviti?
Receptiviti was co-founded by Dr. James W. Pennebaker, the former Chair of the Department of Psychology at the University of Texas at Austin, and the inventor of LIWC -- the gold-standard algorithm in the field of language psychology. Receptiviti was spun out of Dr. Pennebaker's lab to develop software that detects and understands the human emotions, psychology, personality, mindsets, and motivations hidden in virtually any source of language data. Our platform makes it easy for both novices and experts in data science, machine learning and natural language processing to generate hundreds of novel psychology-based features from language data and create highly-sophisticated models that inform decisions and predict behaviour. These same features describe the specific psychological processes that contribute to the behaviours you are observing, predicting or eliciting. Receptiviti also owns the global commercial rights to LIWC.
Is it validated?
Yes. Our core science, LIWC, was developed by one of our co-founders, Dr. James W. Pennebaker, and has over 19,000 academic research citations. LIWC has been applied to virtually every business challenge you can imagine – including mental health, organizational psychology and culture, medicine, recruitment optimization, leadership and succession planning, market research, customer intelligence, insider threat detection, finance, fraud detection and more.
Are there benefits over ML and AI models like BERT?
Yes there are, and they are significant:
1. Black-box models, such as BERT, can be highly predictive but they require large training data sets. Receptiviti generates highly actionable insights about your data with zero training data required. Sounds impossible? It's not, because we already have your training data: Our models pull from over 20,000 validated research studies into the linguistic fingerprints of emotions and personality traits, the linguistic signatures of psychological states, and language-based predictors of human behaviour. And if you need to, it's easy to re-norm the insights to fit the distribution of your dataset, or you can start with Receptiviti's scores while concurrently building your training dataset. Receptiviti provides out-of-the-box insights about the emotions, psychology, personality, mindsets, and perceptions of the people behind your language data.
2. Black-box ML and AI models lack the transparency required for trustworthy and explainable decision-support and predictive models. Transparency and explainability isn't just desirable, it's also required in many countries when applied to personal data. The transparency inherent to Receptiviti’s approach allows for greater insight into the specific variables driving the model output.
Does it scale?
It certainly does. Both the Receptiviti Insights Platform and Insights API scale horizontally, respond instantly to high volume requests, and are perfectly suited for the majority of businesses including startups and large enterprises. Our on-premises API is our most flexible and powerful option. It offers unlimited volume and bandwidth, and a single instance can handle sustained rates of 100,000 words per second per CPU.
What about data security?
Receptiviti was designed with privacy and security at the forefront. We offer multiple options depending on your needs: Our web-based API is highly secure - it doesn't store data and all requests are encrypted. For users with very high bandwidth demands or elevated security requirements, we also offer a containerized version of the API that can be installed within any technology environment.
What does it cost?
We offer a Web-based API a Container API depending on your needs, and we offer discounted rates to academics. Check out our pricing here.