IBM’s Watson wants to assist you with your tax returns


If anyone can make sense of the over 74,000 pages of the US tax code, IBM’s Watson can. Or at least that’s the plan as Big Blue has teamed up its Watson cognitive supercomputer with the tax return specialists at H&R Block to help customers with tax filing options.

As part of the first phase of the collaboration, H&R Block and IBM development teams trained tax language Watson, first applying the technology to the myriad questions and topics discussed during the return filing process.

The service uses cloud-based Watson services to understand context, interpret intent and draw connections between clients’ statements and relevant areas of their return, the companies said.

IBM said that Watson’s initial training was validated by H&R Block tax experts – who have filed some 720 million returns since 1955 — and the initial corpus will expand over time through each subsequent tax season. During the next phase, H&R Block tax professionals will work with IBM to continue teaching Watson all about the tax code and apply the technology to innovate in other areas of their business.

Watson uses machine learning and natural language and image recognition to develop all manner of intelligent answers to tough challenges. The system has been successfully deployed in the healthcare industry where the system has become a trusted adviser to hospitals and research centers working for people fighting cancer.

With the ability to read more than 800 million pages per second, it can analyze vast volumes of data — including the unstructured kind — processing it by understanding natural language, generating hypotheses based on evidence, and learning as it goes, according to a recent IDGNS article.

Watson is also making inroads into cybersecurity. Most recently IBM bolstered the Watson for Cybersecurity project by adding 40 new enterprise IT companies to help develop the system to automate IT security duties and analyze the tons of alerts generated daily by security operations.

Read the source article at Network World.