The 500 million tweets per day sent by Twitter users could soon be read by machines and analysed by IBM data scientists, following an agreement between the companies.
Twitter streams will be analysed by IBM’s Watson artificial engine, to enable businesses and institutions to understand their customers, markets and trends.
IBM CEO Ginni Rometty, said the agreement will help clients make better business decisions.
Companies like Salesforce.com offer sentiment analysis tools for analysing social media trends. Digital marketing agencies often build automated engines to provide their clients with more detailed analysis. The difference with the IBM/Twitter partnership is that IBM is tying Watson analytics, a natural language query engine, with its global consulting practice.
More articles on sentiment analysis
Social media analytics mission: Avoiding mixed sentiments
IBM Global Business Services professionals will use Twitter data as part of their consulting services. IBM and Twitter will also develop industry specific services for sectors such as banking, consumer products, retail, and transportation.
IBM will also integrate Twitter with its analytics services on the cloud, as part of IBM Watson Analytics. It will also offer a cloud-based data refinery service to enable application developers to embed data services in applications. IBM said software developers will also be able to integrate Twitter data into their own cloud services built using IBM’s Watson Developer Cloud or IBM Bluemix platform-as-a-service.
Twitter CEO, Dick Costolo, said: “This important partnership with IBM will change the way business decisions are made – from identifying emerging market opportunities to better engaging clients, partners and employees.”
The first joint product will integrate Twitter data with IBM ExperienceOne customer engagement, allowing sales, marketing, and customer service professionals to map sentiment and behaviour.
A recent survey from analyst Forrester from its Listening Platforms Satisfy Users For Now report found that sentiment analysis received one of the lowest satisfaction scores (3.73 on a 5-point scale). Author of the report, Forrester analyst, Allison Smith said: "This is due to the difficulty of characterising sentiment as positive or negative."
It remains to be seen how Watson analytics will cope with the vagaries of analysing Twitter feeds. Its main application areas to date has been in medical research and financial services, where the vocabulary used by expert users is more defined.