One Toronto firm is using cognitive artificial intelligence (AI) processes to mine big data sets in social media to help asset management firms make better decisions for investors.
Based in Toronto, Buzz Indexes claims that machine learning has evolved to the point where developing models to monitor and understand the context within the millions of posts and comments around stocks and investments made on online social media platform such as Twitter is a reality.
The company initially launched its Buzz Social Media Insights Index this past spring and offers regular methodology and stock rebalancing updates, most recently this past month.
According to Buzz Indexes founder Jamie Wise, the offering takes a big data approach to social media such as Twitter, aggregating chatter on investment opportunities to track potential actionable insights: “Each month we look at an index of the 100 most talked about stocks in the media landscape and the tone and depth of the conversation,” he said.
Specifically, the firm reviews social media platforms, online news sites and web forums to identify “influencers” whose tweets, comments and posts are most likely to impact collective opinion. This involves a process of tracking which members of the online community have historically been the most successful in their forecasting accuracy, according to Wise.
The company claims that its AI and natural language processing software approach has outperformed the Nasdaq and S&P 500 for the past three years; using a proprietary algorithm model, the Buzz Index technology is able to determine whether a comment or conversation on social media — including hashtags and forum posts — is bullish, bearish or neutral, with an eye on longer term trends.
Particularly within the financial services and investment space, big data is certainly where everything is moving, notes Wise. Buzz Index recognizes that there is a huge amount of customer feedback within various social channels and focuses on helping users make the best use of it, he said.
Big Data: Going Social
Whether it’s hedge funds or exchange-traded fund (ETF) investment offerings, using big data to gauge the depth of investor community conversation on social media can be beneficial in determining the best way to extract value from all of the information that’s out there to make better investment decisions.
“The nice thing about the stock market is that you can measure success in a black or white format — you can identify success over time in terms of performance,” Wise said: It’s all about using social media to help organizations better understand customer behaviour.
The ultimate aim for the index is to identify the stocks with the most social momentum, which helps firms define and identify market trends and develop products that meet the needs customers are looking for, he added.
“Everyone in the financial space is figuring out how big data can complement their research and product offerings,” Wise said, adding that the social factor is a potential key. Using machine learning models can help firms better use big data to potentially forecast economic events, engaging in stronger portfolio diversification, and tailor retail strategies, for example.
“The advantage of machine learning is around training machines to understand context behind longer form sentences or blogs… (and) trying to train the model to read what is being posted from the social media landscape,” he said.
It all speaks to how cognitive AI is developing, he added. There are now a lot of developing machine learning frameworks in place for better understanding the sentiment of text, video and photo content.
“It’s evolved significantly. Three or four years ago, this would have been something cutting edge or a side project for a computer science lab,” he said.
