Interview with Vladimir Petkov, Chief Data Officer of Updata One and CEO of Identrics in Bulgaria.
Hi Vladimir, Updata One is a newly formed community. What companies are included and what are the main purposes of the community?
The four companies that make up the Updata One community (A Data Pro, Identrics, SeeNews and Perceptica), have a variety of expertise and knowledge. Though they’re different companies, they operate in the very same knowledge-intensive industry and this makes the relationship between them beneficial and essential at many levels while allowing each company to build valuable partnerships with the others.
We have many examples of the group’s collaboration in order to answer a specific customer request or perform internal process optimisation. This collaboration leads to an increase of expertise within the community of professionals and provides an opportunity to deliver better products and services to current and potential clients.
What is your background and what is included in your current role as Chief Data Officer at Updata One?
Almost all of my professional path has been related to data and innovation. I have six years of experience in digital media and business development, gained at Economedia, Bulgaria’s top news publisher. After that, I continued my journey as CTO at A Data Pro, a global company specialising in content, data and business intelligence services. Since 2016, I have been CEO at Identrics, and now I am also Chief Data Officer at the newly established business community of Updata One.
In this role, I am responsible for data. I oversee the effectiveness of processes such as data collection, management and transformations and find ways to improve them. In addition, I pay attention to the development of the market we are operating in and find ways to innovate and implement automation.
Outside my professional life, I am a podcast producer and host and deliver lectures on online business and digital media-related topics at various universities.
You are also the CEO of Identrics where you provide automated abstractive summarisation. Can you explain why that is important for media intelligence?
Abstractive summarisation is a solution we launched recently. The media intelligence industry is in constant search for better solutions of content aggregation, transformation and analysis, and we can all agree that we witnessed the rise of a phenomenon during the pandemic - the infodemic.
Nowadays it is even hard for communications experts to know everything about the organisation they are responsible for and respond to the emerging crisis in the fastest possible way.
These problems need the development of new technologies to help media intelligence company deal with the heavy load of information and improve its process, so they can meet the need for accurate, immediate delivery of news resumes.
Abstract summaries are tough and necessitate significant language modelling. It's also usually more difficult than extractive summarisation. The information in a text is interpreted by abstractive summarisation, which generates new sentences for the summary. It's pretty similar to how long text summaries are written. This method differs from prior versions of summarisation, which used input sentences to construct shorter texts. To develop human-like summaries, Identrics' solution employs AI and Machine Learning.
However, there are certain difficulties in creating a new text. One such issue is that abstractive summarisation algorithms can sometimes produce erroneous data. For this reason, Identrics uses a fact-checking method to address this issue.
How will abstractive summarisation affect copyright, and do you see any challenge in relation to copyright owners?
Abstractive summarisation approaches provide outlines that are easily understood and short. They reduce the length of sentences, resulting in reduced recurrence of tasks.
Because abstractive summaries are wholly new texts, they can assist firms to avoid copywriting challenges. This helps save money and time.
Have you recently, or do you plan to, release any new technology-based solutions in 2022 that will add to or improve the services you offer your clients?
We are planning to launch various solutions this year, starting with entity-based abstractive summarisation.
A project we have been working on in the past months is our Enterprise Knowledge Graph, which will allow us to solve complex problems such as entity resolution and entity linking. It will help Identrics create deep reports and include non-obvious knowledge.
Another development is our model for fake news and hate speech detection. These types of solutions are essential in a world where everyone has the freedom to share their own or another person’s opinion, and the information is spreading at a fast speed.
When it comes to the actual data behind media intelligence, what kind of data or media not currently used can be interesting in the future?
There is always room for additional development and improvement.
Social media is a game-changer for sure, but there are also other emerging communication channels. A lot of podcasts are generating content that could be of interest to business decision-makers and PR experts.
Also, we all know that content is king, but video content is God. There are more and more videos spreading across the internet that carry a variety of information with them. Communicational experts need additional instruments to better explore the information in all types of media.
Can the entire process of media intelligence be automated in the future?
Many tasks can be automated with the integration of artificial intelligence and machine learning models. The truth is, though, that technology is still not mature enough to replace people. In my opinion, there will not be full automation in any industry in the following decades.
Automation is not everything when it comes to media monitoring. While automation saves the analyst a great deal of time, it cannot handle everything. Analysts that track media must also consider their opinion, or whether an article's or a social media post's tone is good, neutral, or pessimistic.
What excites you the most when it comes to automation in the media intelligence industry in the future?
The emergence of new communication channels brings the need for fast development in the media intelligence industry. Think about it: 10 years ago not so many people owned a smartphone, social media didn’t control our world and nobody had even thought of podcasts.
Now we have to deal with new copywriting issues, and the boom of fake news and hate speech. Addressing all of these problems and looking for a way to solve them is something that I am really passionate about.
How do you think the media intelligence industry will change in the next 5-10 years, and what are the greatest challenges ahead?
The media intelligence industry needs to change with the pace that media in general is doing. The term “media” has transformed dramatically in the past several years. Today, everybody can be media, and everybody can spread the news around its network.
I think we will have to invest even more in R&D initiatives and find new ways to monitor all the conversations that are happening around the net, around the globe.
By Peter Appleby
