By @SimonCocking

A preview piece by Parsa Ghaffari @parsaghaffari from Aylien about the upcoming Predict Conference. Where Business Decisions Meet Data & Predictive Analytics | 15-17th September | RDS, Dublin, Predict Conference @predictconf | Registration now open.

Who does Predict Conference appeal to?

In today’s World, the importance of data, and the idea of looking at data to make better/smarter decisions is no longer a secret. More than 60% of top 100k websites on the Web today use Google Analytics.

Businesses large and small can benefit from a wide variety of data analytics tools and solutions to understand their markets, their customers and their own business better, and ultimately make better decisions.

Prescriptive Analytics is fundamental and necessary for getting an idea of where things have been in the past and where they are right now, but it’s through Predictive Analytics that you can get a truly macroscopic view of things and identify trends, patterns, and ‘predict’ where things might go in the future.

Predict Conference is for anybody who’s interested in discovering, utilizing and building data-driven solutions, with a special focus on Predictive Analytics.

Why were you interested to speak at it?

Although we’re seeing a high availability and adoption of data and content analysis solutions, and democratization of tools and instruments in this space, most of the efforts are still limited to structured data (think databases, spreadsheets, etc). But there’s a wealth of information trapped in unstructured data such as text and images, which must be extracted and taken into account.

At AYLIEN we utilize AI techniques and applications such as Machine Learning, Natural Language Processing and Image Recognition to make sense of unstructured data (text and images) and we think it’s crucial to look at both types of data to get a holistic view of reality.

What can people expect from your talk?

The topics aren’t finalized yet, but most probably I will be talking about how NLP can be used in predictive models, e.g. can we predict which headlines will receive most attention from people on Social Media?

Last year we collected about 80 million tweets related to the World Cup 2014, and we discovered some really interesting patterns between the sentiment of these tweets and various events happening during the games. For instance, when the Uruguayan Suarez bit another player in a critical match, we could clearly see a negative trend forming in tweets that contained a mention of him or the incident (although he quickly recovered from it thanks to his PR team!):

Screen Shot 2015-08-03 at 5

This is the type of thing I’ll be talking about.

What are you excited about?

What I’m really excited about in this field is the combination of cheap/highly available computation and storage infrastructure, state-of-the-art techniques and algorithms and large amounts of data, which I’m very hopeful will give us an unprecedented understanding of how things work in different verticals, and in new formats (intelligent business applications, analytics solutions, etc.)

Which future trends could make the biggest change to people’s lives in the next 1 – 3 years?

Again I think the three major trends that will have a key impact here are:

– Algorithmic improvements: there’s a lot going on at the moment with Deep Learning and related ideas.

– Big Data: it’s easier than ever to access, collect and process large amounts of data.

– Cheap/highly available computation and storage infrastructure: cloud economics has continuously moved in favor of consumers – storage is almost free now, and I suspect computation will follow the same route soon.

These combined with innovation at startups/bigger companies will result in cheap, accessible analytics technologies and solutions that will ultimately benefit the end-user. Think less spammy ads, better personalization, better search results, etc.

What are the potential challenges to these not being achieved?

Well I guess the biggest unknown is how far and deep we can go with the algorithms and how quickly. That’s the hard part. Same goes for innovation.

There are challenges associated with Big Data but a lot of those are getting resolved, as solutions and ideas mature and become more streamlined (e.g. Apache Spark).

We should expect the least risk/challenge from infrastructure.

Talk about the sweetspot between predictive analytics and Aylien’s expertise, what sort of opportunities might this offer for you / businesses and how could it improve the lives of us / end users?

There’s a strong demand for smarter solutions (incl. Predictive Analytics) in the content space i.e. News & Media at the moment, and I think AYLIEN is in a very good position to move fast and serve this market with its novel technology. That said, many of the use cases are still unknown and that’s what makes attending a conference like Predict exciting.

For the end-user this ultimately means more engaging content, better discovery, better personalization, better ads, etc.

Pin It on Pinterest

Share This

Share this post with your friends!