Marcia Davis Kerchner (PhD’71) attended Cornell at a time when the field of computer science was newly forming. At the time of earning her degree, it had been less than ten years since Cornell University and the State of New York officially recognized the Computer Science faculty and the department. Although the department was around since 1965, the first bachelors’ degrees weren’t even granted until 1980. Computer Science at Cornell just celebrated its 40th anniversary, reminding us that there was a time when computers were not at all main stream.

Originally a math major, then Marcia Davis decided to enroll in her junior year in one of the first programming courses Queens College had to offer, a FORTRAN course that sparked her interest in computer science. Turned on to the field by the broad opportunities available to computer scientists, she chose to apply for the PhD program at Cornell. She had only three credits in computer science under her belt but was accepted alongside students graduating with complete undergraduate computer science degrees from other universities.

Kerchner explains that she gravitated toward a topic that is today relevant to both Computer Science and Information Science – Information Retrieval. She studied under the late “Father of Information Retrieval” Gerald Salton. The Computer Science Department still honors Professor Salton with an annual Salton Lecture Series. Described by Kerchner as a man with a gruff exterior and soft heart, Salton guided Kerchner’s dissertation study on “Dynamic Document Processing in Clustered Collections.”

Since 1984, Kerchner has been at the MITRE Corporation, a not-for-profit organization chartered to work in the public interest. She has worked with government agencies in many areas, including information architecture and system design, information retrieval, and web usability. MITRE, organized to apply “systems engineering and advanced technology to critical national problems,” is currently contracted by the IRS to improve IRS.gov. Kerchner has developed methodologies that have led to greatly improved user experiences.

A lot of what she’s doing to modernize government services and interfaces, Kerchner calls “common sense.” She asked, for example , the simple question, how do people search for the 1040 EZ tax form? By analyzing users’ information-seeking behaviors through site usage and help desk reports, she found that there were more than a dozen ways that people asked for it but, unless they typed in ‘1040 EZ’, the search engine would not find the right link. Based on this data analysis, Kerchner adapted search results to the user behavior rather than expecting users to get the form numbers just right and was able to increase user satisfaction scores and significantly reduce the number of calls and e-mails to the IRS.gov Help Desk.

As “common sense” as this may seem, elements of information retrieval Kerchner studied under Salton over 30 years ago have only gotten into commercial systems over the last 10-15 years. Although much of her work is unpublishable due to security restrictions, she was able to publish her innovative work improving the search experience on IRS.gov, “A Dynamic Methodology for Improving the Search Experience,” in a refereed journal this year.

You may wonder why Kerchner ended up in the commercial world instead of academia, as she had planned. In the 1970s, female computer science professors were as in demand as they are today. But Kerchner's anecdote is telling about how times have changed for women in computing. At the time of her search, she was receiving a lot of interest. When she asked a CS faculty member for yet another recommendation, the professor advised Kerchner to wait and delay her job search until her husband found a position first. When the university that hired her husband had only a part-time opening left for Kerchner, she weighed the larger salary offer outside of academe – then the hefty sum of $9,000 – as temptation enough to forgo the continued search for a teaching position.

But for this small piece of advice, millions of Americans would be spending many additional hours driving to the library or post office to get an IRS form.