Wednesday, February 22, 2017

Artificial Intelligence Public Relations



In wandering through the Suffolk countryside, it struck me that there is very little by way of a Public Relations perspective of  Artificial Intelligence (AI). We need a simple introduction to the subject. As one does, I took out my phone and dictated a ten-minute introductory lecture. I have embedded it into this blog post.  It is only ten minutes long to be easy to listen to rather than reading a long boring blogpost. There are other explainations, you may like to explore as well.

Listen to "Artificial Intelligence - a ten minute lecture for PR folk" on Spreaker.

Basically, I took as a simple example the application of AI in the analysis of Twitter in its reportage of Brexit. Initially, the issue is one of the collection of data. This can be done in a variety of ways. Twitter has an API  (IBM has a good resource for API beginners here) and this gives access to Twitter  content (there are some services you can use too).
You may have to make decision at the start  would data be collected by the minute, hour day, would one collect the names of the people posting content and what words do they use is the first part of this exercise. Of course, this quickly becomes a lot of information. It is "Big Data". This can be presented in list form. For example how many citations per hour, who are the top contributors, what are the subjects being discussed etc.
We already have services that can do this for us. The media analysis agencies do this. NodeXL  is a useful tool for all of us. Examples show Twitter profiles with tens, even hundreds of thousands of followers.
But we may want to look for hidden perspectives such as a perspective of the expression 'Hard Brexit' or Remain centric content. We might want to work on predicting content and who will drive it (using this open source programme from Google). All such considerations are hard enough for us to assemble and reconfigure the data to provide the answers. But, what if the computer programme was to find similar perspectives for you to consider? These findings are already available using (open source) software like Watson and gain valuable insights from social trends in real time with the solutions integrating natural language classifiers.
What if it was also able to additionally predict the coverage of such perspectives and tell you the level of confidence it had in its predictions.
It will be important to be able to represent such findings in an understandable way. There are examples such as this one.
That would be pretty awesome. But what if we were to collect the data from other media such as Facebook, LinkedIn (with permission, of course), newspapers and other media. Additionally one can add sentiment analysis for even deeper understanding. That too would be very interesting and helpful.
Now, what if one were to combine all these findings. It could, for example, tell us which media leads opinion and excites social media or journalistic activity. Building such capabilities are available to everyone and produce amazing results.
This then describes a scenario for the application of Artificial Intelligence. It is a capability that will be available in months, not years. In the meantime, have a look at Big Data analysis using NodeXL.
All this information is very interesting but so what! It's just lists. How does it become useful?Using such findings, you will be able to inform the Board about where the debate and interest in the media, both traditional and social, has come from and is going to. It can identify the key influencers who can be the target for your views and perspectives. It could suggest what the key words and ideas are for sections of society in various media.
So much for Brexit, but in using Artificial Intelligence in this way, we can examine other subjects and interest in organisations, products, services, employees and other key publics (what really drives politicians?).

The PR industry should be prompted by Karen Bradley, Secretary of State for Culture, Media and Sport, to make AI in the PR industry more relevant. She said: "I want the UK to lead the way. We are already pioneers in this exciting technology. We have some of the best minds in the world, working in some of the world’s best universities. We have earned a reputation for brilliance in AI.'

Perhaps we can now look forward to AI integrated into the work of public relations. Can it, for example, propose content for traditional and social media (yes, the actual words and pictures). Can it be used to seek out the relationships between content and PR activities? For example online content and sales? Is it a tool for developing the relationship management tasks in public relations? Such activities will be useful too, but there is much more by way of AI opportunities.

We have to note that there is a lot of copyright and patent application being implemented now.

However, analysis of various factors such as media coverage might identify the nature of the differences between one organisation and its competitors (stability, profitability, progress, people, products, services, and vendor relations, etc.).
There are other areas for the application of AI from defining the nature of organisations (for better mission statement development) to methodologies for enhancing relationship's between organisations and 'publics', but that is for another day.
Suffice to say, that the practitioner armed with what is outlined here, is the person with both the intelligence data and capability to be one of the most powerful people in the organisation and beyond.


Here is a short history of AI. Thank you Albert Puig.




For more about AI this is a slideshow to have a look at.



There are slides for the technicians - like marketing.





Picture: Wintery sunshine over the river Deben.