Early days of information retrieval

“A new chapter for libraries” (features, Jan. 28) brings to mind one of the most foolhardy actions in my job as manager of information services for the American Management Association approximately 40 years ago – attempting to engage in dialogue with the late H.P Luhn, senior scientist at IBM, during a Q&A session ending a conference devoted to developments in what we used to call information retrieval.

My recollection brings up Luhn as the first to program a computer to play chess. Along the way, in punchcard days, he developed a method for what Professor Dan Cohen ’90 describes as the work of an “active library.”

SDI, or selective dissemination of information, involved an IBM scientist’s key-punching his information needs, in terms of publications, into the standard IBM card. Key title/subject words informed the librarian as to what to send in terms of abstracts and/or full-text hard copy if available.

Later, about 30-plus years ago, Lockheed Dialog and Systems Development Corp. computerized this system, and an organization, utilizing a cumbersome local/long distance phone and modem with its billing ID, received microform of the specialized abstracts/full-test. At the time, Lockheed charged 50 cents per abstract. Although we rarely used this at AMA because our service was gratis to members, we did obtain a billing number and trained departmental staff at periodic Lockheed seminars.

Google discovered advertising online, not state-of-the-art digitalizing/information retrieval. It pays for database licensing that we were indirectly charged for. Of course, any Internet user knows not all readouts are free. If memory serves, newspaper advertising is off 10 to 12 percent of late and online advertising up 26 percent. But do not forget really old times. Dewey decimal has always used terminal digits for specificity, and the accounting profession has used such for aging receivables.