Deep Learning

A Super Computer Analysis of New LDS Apostle Personalities

IBM’s Watson super computer can win Jeopardy against its greatest champions.  What does it think of the three newly called apostles of The Church of Jesus Christ of Latter-day Saints, Elders Ronald A. Rasband, Gary E. Stevenson and Dale G. Renlund?  According to an analysis utilizing IBM Watson’s Personality Insights service, the one attribute they all exude 100% based on their first General Conference remarks since their new leadership call is sympathy.

This analysis can be done by anyone on any text by visiting the IBM Watson Personality Insights service available here.  The site also has a detailed explanation of the science behind the analysis including its limitations; a direct link to that is here.  Note that the October 4, 2015 remarks were all far under the recommended number of words for an appropriate analysis.  At least 3,000 words are recommended with highest accuracy at 6,000 words or more.  An analysis including the October 4, 2015 remarks and multiple previous General Conference talks by each speaker was performed and is included below.  First, a look at the numbers based on the remarks on October 4, 2015.

Of the personality attributes analyzed by Watson, these are the attributes Watson found Elders Rasband, Stevenson and Renlund to be most similar.

Table 1: Rasband, Stevenson, Renlund October 4, 2015 Most Similar Personality Attributes

Attribute Rasband Stevenson Renlund
Sympathy 100% 100% 100%
Self-consciousness 1% 0% 0%
Agreeableness* 97% 96% 96%
Self-transcendence 96% 97% 97%
Outgoing 90% 93% 92%
Extraversion* 97% 92% 97%
Self-enhancement 8% 3% 4%
Assertiveness 83% 84% 88%

Sympathy was equal across all three at 100%.  There was a 1% difference between Self-consciousness, Agreeableness* and Self-transcendence.  There was a 3-5% difference on the Outgoing, Extraversion*, Self-enhancement and Assertiveness attributes.

*Agreeableness and Extraversion are category headers within the Big 5 (see explanation of the science).  Extraversion represents Activity level, Assertiveness, Cheerfulness, Excitement-seeking, Outgoing and Gregariousness; Agreeableness represents Altruism, Cooperation, Modesty and Uncompromising.

The same analysis by IBM Watson was run with the Sunday morning remarks by Church President Thomas S. Monson (see Table 2).  President Monson also exudes 100% Sympathy and has similar results for the attributes in Table 1.

Table 2: Rasband, Stevenson, Renlund October 4, 2015 Most Similar Personality Attributes with President Monson

Attribute Rasband Stevenson Renlund Monson
Sympathy 100% 100% 100% 100%
Self-consciousness 1% 0% 0% 0%
Agreeableness 97% 96% 96% 97%
Self-transcendence 96% 97% 97% 100%
Outgoing 90% 93% 92% 97%
Extraversion 97% 92% 97% 97%
Self-enhancement 8% 3% 4% 0%
Assertiveness 83% 84% 88% 90%

There are differences in the personalities of the three newly called apostles as one would expect for individuals with varying experiences and backgrounds.  These are the attributes where there was the widest range between the new apostles. President Monson is included but only the widest differences between the three new apostles was measured.

Table 3: Rasband, Stevenson, Renlund October 4, 2015 Widest Personality Attributes Range with President Monson

Attribute Rasband Stevenson Renlund Monson
Openness to change 67% 2% 0% 93%
Conservation 34% 92% 98% 1%
Challenge 7% 31% 64% 82%
Self-discipline 57% 83% 33% 33%
Structure 7% 25% 56% 18%
Conscientiousness 69% 87% 38% 58%
Achievement striving 42% 62% 13% 21%

As was mentioned previously, the number of words in the October 4, 2015 remarks was not sufficient.  In order to analyze across more words, two additional General Conference talks were included in a second analysis for each person.  For all but Elder Renlund, the total number of words reached the minimum required 3,500 words.  (See Table 5 for word counts and full results)

These are the most similar attributes of the three new apostles based on multiple remarks:

Table 4: Rasband, Stevenson, Renlund Multiple Remarks Most Similar Personality Attributes with President Monson

Attribute Rasband Stevenson Renlund Monson
Self-consciousness 0% 0% 2% 0%
Sympathy 100% 100% 97% 100%
Hedonism 9% 9% 6% 4%
Orderliness 22% 19% 18% 13%
Self-transcendence 98% 93% 98% 100%
Assertiveness 87% 83% 82% 92%

With a broader data set, Self-consciousness becomes the #1 most similar attribute with only a 2% difference.  Sympathy had only a 3% difference but with both Self-consciousness and Sympathy the difference comes from Elder Renlund who, as previously indicated, does not have enough words to meet the 3,500 word minimum.

The following table presents the results of both times the IBM Watson Personality Insights service was run.  The October 4, 2015 remarks came from the Sunday Morning General Conference Session.  The Multiple Remarks included the October 4, 2015 remarks for each speaker plus remarks from April 2014 and April 2012 for Elder Rasband, April 2014 and October 2012 for Elder Stevenson, April 2015 and October 2009 for Elder Renlund, and Priesthood October 2015 and Sunday Morning April 2015 for President Monson.

Table 5: Full Personality Data Results for October 4, 2015 Remarks and Multiple Remarks

October 4, 2015 Remarks Multiple Remarks
Attribute Rasband Stevenson Renlund Monson Rasband Stevenson Renlund Monson
Words Analyzed 902 1054 982 1765 3792 4173 3073 4770
Big 5    
Openness 40% 72% 48% 70% 65% 83% 63% 64%
Adventurousness 37% 72% 43% 66% 62% 84% 57% 67%
Artistic interests 97% 94% 85% 87% 91% 69% 53% 81%
Emotionality 91% 67% 76% 54% 66% 30% 53% 56%
Imagination 16% 25% 27% 40% 39% 62% 44% 38%
Intellect 28% 64% 45% 75% 61% 83% 66% 66%
Authority-challenging 12% 30% 13% 16% 23% 40% 30% 11%
Conscientiousness 69% 87% 38% 58% 65% 85% 36% 54%
Achievement striving 42% 62% 13% 21% 29% 56% 13% 12%
Cautiousness 49% 75% 62% 77% 65% 90% 57% 68%
Dutifulness 24% 41% 12% 30% 27% 45% 23% 31%
Orderliness 55% 60% 30% 17% 22% 19% 18% 13%
Self-discipline 57% 83% 33% 33% 50% 72% 21% 35%
Self-efficacy 73% 92% 59% 93% 84% 89% 60% 87%
Extraversion 97% 92% 97% 97% 96% 78% 91% 97%
Activity level 2% 16% 2% 1% 2% 32% 7% 1%
Assertiveness 83% 84% 88% 90% 87% 83% 82% 92%
Cheerfulness 88% 54% 73% 76% 70% 34% 52% 78%
Excitement-seeking 45% 9% 23% 16% 21% 6% 21% 25%
Outgoing 90% 93% 92% 97% 93% 86% 87% 96%
Gregariousness 89% 75% 82% 82% 79% 44% 68% 87%
Agreeableness 97% 96% 96% 97% 96% 85% 90% 97%
Altruism 82% 91% 73% 93% 85% 85% 70% 92%
Cooperation 65% 82% 36% 59% 70% 83% 42% 55%
Modesty 33% 60% 69% 7% 28% 48% 53% 21%
Uncompromising 73% 90% 71% 66% 72% 74% 56% 67%
Sympathy 100% 100% 100% 100% 100% 100% 97% 100%
Trust 50% 81% 51% 60% 66% 73% 50% 59%
Emotional range 53% 32% 56% 34% 31% 11% 58% 31%
Fiery 56% 35% 65% 51% 38% 26% 64% 48%
Prone to worry 76% 45% 61% 45% 42% 11% 56% 42%
Melancholy 28% 17% 55% 16% 19% 13% 56% 22%
Immoderation 36% 13% 22% 1% 14% 7% 17% 3%
Self-consciousness 1% 0% 0% 0% 0% 0% 2% 0%
Susceptible to stress 71% 45% 46% 38% 35% 7% 49% 31%
Needs
Challenge 7% 31% 64% 82% 13% 11% 21% 71%
Closeness 42% 68% 84% 64% 58% 47% 73% 74%
Curiosity 7% 11% 48% 19% 14% 14% 33% 38%
Excitement 37% 38% 60% 33% 33% 22% 44% 46%
Harmony 79% 84% 100% 88% 87% 77% 97% 95%
Ideal 6% 19% 18% 8% 6% 6% 12% 15%
Liberty 54% 50% 81% 49% 52% 35% 67% 64%
Love 81% 85% 98% 92% 87% 76% 96% 95%
Practicality 76% 73% 100% 81% 85% 74% 98% 92%
Self-expression 66% 35% 60% 14% 46% 21% 46% 21%
Stability 22% 36% 64% 42% 33% 25% 50% 60%
Structure 7% 25% 56% 18% 17% 19% 44% 35%
Values
Conservation 34% 92% 98% 1% 11% 41% 42% 2%
Openness to change 67% 2% 0% 93% 80% 27% 51% 92%
Hedonism 13% 1% 0% 1% 9% 9% 6% 4%
Self-enhancement 8% 3% 4% 0% 6% 20% 7% 1%
Self-transcendence 96% 97% 97% 100% 98% 93% 98% 100%

The above results were obtained October 7th and 8th, 2015.

Personality Summary

The IBM Watson Personality Insights service provides a short personality summary compared to most people who participated in the surveys.  These are the summaries from the Multiple Remarks analysis for each of the new apostles and President Monson:

Elder Ronald A. Rasband

You are social and confident.

You are empathetic: you feel what others feel and are compassionate towards them. You are confident: you are hard to embarrass and are self-confident most of the time. And you are laid-back: you appreciate a relaxed pace in life.

Your choices are driven by a desire for well-being.

You consider helping others to guide a large part of what you do: you think it is important to take care of the people around you. You are relatively unconcerned with achieving success: you make decisions with little regard for how they show off your talents.

Elder Gary E. Stevenson

You are rational, helpful and genial.

You are empathetic: you feel what others feel and are compassionate towards them. You are confident: you are hard to embarrass and are self-confident most of the time. And you are calm-seeking: you prefer activities that are quiet, calm, and safe.

Your choices are driven by a desire for well-being.

You consider helping others to guide a large part of what you do: you think it is important to take care of the people around you. You are relatively unconcerned with taking pleasure in life: you prefer activities with a purpose greater than just personal enjoyment.

Elder Dale G. Renlund

You are social.

You are confident: you are hard to embarrass and are self-confident most of the time. You are empathetic: you feel what others feel and are compassionate towards them. And you are laid-back: you appreciate a relaxed pace in life.

Your choices are driven by a desire for efficiency.

You consider helping others to guide a large part of what you do: you think it is important to take care of the people around you. You are relatively unconcerned with taking pleasure in life: you prefer activities with a purpose greater than just personal enjoyment.

President Thomas S. Monson

You are social and generous.

You are empathetic: you feel what others feel and are compassionate towards them. You are confident: you are hard to embarrass and are self-confident most of the time. And you are laid-back: you appreciate a relaxed pace in life.

Your choices are driven by a desire for connectedness.

You consider helping others to guide a large part of what you do: you think it is important to take care of the people around you. You are relatively unconcerned with achieving success: you make decisions with little regard for how they show off your talents.

Personality Visualization

The IBM Watson Personality Insights service also provides a graphical visualization of the personality data.  Here is President Monson’s visualization from the Multiple Remarks analysis:

MonsonVisualization_MultipleRemarks

Advertisements

Affective Computing

My latest comes to you via LinkedIn. Catch me 5/21/2015 at Stanford GSB where I’m co-chairing an ’emotional’ night.  Check out: Computers can know your feelings. How does that make you feel?

 affective computingnew banner

Deep Learning

“A machine learning approach inspired by the human brain, Deep Learning is taking many industries by storm.”  Visit this link to learn more about the VLAB Deep Learning event.  The event video will be uploaded within a few weeks.  In the mean time, please check out this article by Forbes contributor Robert Hof: AI For Everyone: Startups Democratize Deep Learning So Google And Facebook Don’t Own It All

It would be an understatement to describe the VLAB event team journey as rigorous and rewarding.  The team? VLAB volunteers.  The topic? Deep.  The result? Learning…by all attendees from a superb panel. #VLABdl

A special shout out goes to the event co-chairs Jeff Stevens and Shane Gu as well as my fellow team members Adwoa Boakye, Natalia Erokhina, Priyan Guneratne, Fred Stein, Alexander van Dijk and Wolfram Willuhn.  Countless hours were spent researching, screening panelists and coordinating what seemed like a never-ending list of tasks to ultimately construct the panel that so eloquently addressed the topic for the sell-out VLAB audience.  As lead for marketing and demo exhibits, I must remind everyone it takes a team to sell out an event.  Thank you to all who answered the clarion call to action to support event marketing.

 

IMG_4710

IMG_4711

IMG_4712

IMG_4714

IMG_4717

IMG_4716

IMG_4721

Smiles in a packed crowd.

Wearables Predictions: Who to Watch for Prediction #4

This is the fourth post in a Wearable Industry Watch Series for each of the 10 Wearables Predictions.  Follow this blog or Twitter handle @WorkTechWork to be notified of each part of the series. To view all predictions and links to the other parts of the series, visit the Wearable Industry Watch Series.

Prediction #4:  Wearables will become more intelligent because of developments in sensor technology and the ability to translate data from these sensors into insight via analytics.

Developments in Sensor Technology:

One of the challenges of wearable electronics is the rigidity of the components and the resulting space requirements to manage ridged components.  Consumers want small, comfortable devices but the inherent rigidity of metal and silicon require that components be kept from bending or flexing or else they can becoming brittle and break, a particularly challenging issue wearable textiles.

Dr. Wenlong Cheng and a team of researchers at Monash University in Melbourne, Australia have come up with a sensor that can bend or be twisted without cracking (read more here).  I had the opportunity to talk with Dr. Cheng about this technology last week and the conversation has me excited for potential applications in the wearable industry, as well as the wider Internet of Things because of this material’s fast response times, high sensitivity and stability under lots of different situations.  This technology could disrupt how all wearable devices are designed in the Biosensing Wearables Landscape described by Rock Health. (image source Rock Health)

RockReportWearables4 from Rock Health

Ability to Translate Sensor Data to Insight via Analytics:

Knowing how many steps taken in a day is one thing, being able to detect anomalies in behavior that resulted in more or less steps, or understanding how the number of steps taken in a day affects behavior, particularly spending behavior, is another and data scientist are working hard to translate sensor data into insight.

It is no easy task.  The Cityzen Sciences team produces D-Shirt, a smart shirt that generates 200,000 data points in one hour.   Big data suddenly became humongous data with such finite data for the quantified self.  Fortunately, the team also offers the Cityzen Data platform which enables data collection and storage so that value, created through analytics, can be created from sensor data.

Cityzen is not alone in analyzing wearable data.  Empath Analytics acquires data from multiple wearable devices and is positioned as a Backend-as-a-Service API to help developers create apps leveraging data from wearable devices.  Empath Analytics can collect, parse, and clean data as well as apply machine learning techniques enabling developers the luxury to focus on generating value for users and improving user experience.

Moreover, big players in the space including Google, Facebook, Microsoft, and others are acquiring sophisticated analytics producing companies with skilled data scientists almost as quick as they can incorporate and build websites.  All that glitters is not gold, though, because it is easy for a bus dev team member to say, “We can turn data into actionable insight” and difficult for a team to actually deliver on that promise.

On a lighter note, what is good for man may also be good for man’s best friend.  Pet wearable players like Whistle  and  Voyce are enabling the quantified pet and using data analytics to enrich the lives of both pets and pet owners.

Lastly, I’ll mention a point on the immense quantities of raw data (recall 200,000 data points and hour from one smart shirt!).  With large data sets, it becomes challenging to identify features in raw data that are meaningful.  Deep learning  may find attributes in quantified self data that humans simply cannot detect.

Are you aware of developments in sensor technology or are you working to translate sensor data into insight?  If so, I want to hear about it so express your thoughts in the comments below or reach out on the contact page.

Next Prediction: Who to Watch For Prediction #5

Previous Prediction: Who to Watch for Prediction #3

Back to Wearable Technology