Search Engine Aims to Make IRS Site More Usable

A new search engine developed by a California CPA tries to make it easier to find relevant documents on the IRS’s Web site.

Brian Dooley of Newport Beach, Calif., introduced his search engine, IRS.gov Wizard, at http://irswizard.net, earlier this week after testing the technology for two months on a private server. “It’s something I had wanted to do for about four or five years,” he said. “I’m a CPA and what’s always bugged me about the IRS Web site is that they have all this information, but like Google or Yahoo, some of the documents may be 20 or 30 pages long, and the [words in the] search terms may not really be related to each other. This way, all the words are in the same sentence or paragraph. If you were to type in ‘corporate elections’ on the IRS site, you would get about 1,000 documents, but to refine what you want, it would take forever.”

His search engine aims to narrow down the searches so that the words in the search term entered by the user are close together, rather than just being individually found somewhere in the document. Dooley believes his search engine can help taxpayers and tax preparers avoid costly mistakes by finding the relevant information from the IRS.

“When you deal with the IRS, if you make a mistake, it’s really serious and you can get a big penalty ranging up to half the value of a bank account,” he pointed out.

The search engine on IRS.gov Wizard relies on Dooley’s knowledge of tax law and tax jargon. He holds a graduate degree in taxation from the University of Southern California and was chairman for 16 years of the California Society of CPAs’ International Tax Committee in Orange County. He has also taught at the University of Southern California and testified as an expert witness before the Treasury Department and the Senate Finance Committee.

His search engine narrows down searches by asking users if they are a business, individual, nonprofit or fiduciary. Searches can be input based on a first search term such as “rental property,” a second term such as “passive activity,” and can exclude a term the user does not want.  The search engine then ranks the best-matched 25 documents, with the most relevant document listed first.