Tag Archives: search engines

Sensitive Segments

Sensitive Segments    —by Jinny Batterson

A long time ago, before I became totally technically obsolete, I worked for a number of years in what was then called data processing—now more often labeled “information technology.”  The media and the structure of the data I worked with changed over time. First there were punched card files, later storage on computer tapes, then multi-plattered disk drives, and, still later, all sorts of increasingly dense mass storage devices.  At first, each separate computer application had its own files, so there was a tremendous amount of duplication among the various files, with a high probability for errors and mismatches. Later, someone figured out that it would be possible to create a data hierarchy, with a top level “executive” or “parent” piece that controlled access to all the others. This reduced duplication and mismatch difficulties, but meant that to access any piece of data required understanding a rigid data structure, with its different levels and dependencies. 

During the final decade of my data processing work, I was introduced to what were then called “relational databases.” As the expense of computer hardware continued to decline and the speed of data retrieval continued to increase, any data hierarchy that might be lurking in the background was masked. Given a properly constituted database, it became possible for programmers to compose relatively simple inquiries into multiple data fields, no matter how they were positioned in the overall data structure. To me, this way of viewing data was much more intuitive than trying to remember whether “segment A” was a parent segment to “segment B,” or whether both were same-level children of “segment C” in some artificially constructed hierarchy. The catch, in the somewhat hierarchical organizations where I worked, was that some data was deemed off-limits or unnecessary for some members of the agency or firm. For example, the Personnel Department might need to access personal information that was considered either too sensitive or irrelevant for the Accounting Department or the Education Department, and vice versa. 

So some smart computer software guru introduced the concept of “sensitive segments” for data base access.  All the data was stored somewhere, but if you were in the Personnel Department, you could only retrieve data from those segments to which you were given access. If you were in Accounting or Education, you would be blind to those segments whose data was reserved exclusively for Personnel. They did not relate to your job description. From your perspective, they did not exist. 

In the glut of our current information environment, it may seem as though the concept of “sensitive segments” is obsolete. In theory, any of us can access most of the data stored online anywhere in the world via the “world wide web.” However, precisely because there is so much information available in electronic form, it becomes totally impossible for any single person or group to retrieve it all, let alone make any sort of sense of it. Therefore, the Googles and other search engines of our age have devised ingenious algorithms to bring us just those “sensitive segments” they believe will most interest and/or please us. Our search-engine-mediated levels of sensitivity have only increased.

Amid the cries of horror at the polarization and dysfunction of our political and social systems, relatively few point to this sensitization as a partial cause. Few stop to remember that any topical query will bring back the “most popular” web pages on that topic first. For example, if I do a Google search on the word “bias,” it brings back about 207 million results, one screenful at a time, with several dictionary definitions as the leading entries. Such a ranking system helps to make sense of lots of relatively simple topics, yet it also opens a way for more and more extreme distortions of the more complex aspects of reality. Unless I type in a specific web address, I are going to be shown just the information deemed by the search engine software as a “sensitive segment” first.

Each of us reaches adulthood having certain segment sensitivities, based on our genetic make-up, our upbringing, and our exposure to various life events.  Some of us, for example, are drawn to emphasize the role of individual initiative in fostering success; others are primed to stress the role of luck. Some feel entitled to a large share of the world’s material goods; others remark on patterns of systemic discrimination and oppression that deprive many of even a small share of such goods.

It is very difficult, a whole life’s work and then some, to unlearn layers of bias and discrimination we learned early in life. It’s crucial that we minimize the distortions fomented by our increasing dependence on Internet-mediated “sensitive segments.” We need the balance of maintaining and strengthening interactions with real people with real lives whose opinions and experiences may be quite different from our own.     

The Myth of Objectivity: Choosing our Facts

The Myth of Objectivity: Choosing Our Facts     —by Jinny Batterson

This fall’s election season left me both physically and emotionally exhausted. In October, as attack advertising on all sides escalated and early voting started, I finalized plans to visit extended family on the west coast shortly after the election was over. It seemed a good way to regain some longer term perspective—the two generations that come after me would help me regain balance.

Now that I’m in sunny California, we’ve mostly avoided talking politics. I’m sure we don’t agree on everything. Avoiding hasty words is an antidote for having to take them back later. The grandkids are cute, and generally uninterested in grown-up concerns.

However, I haven’t escaped political repartee entirely.  My first Sunday here, I was sitting under a shade tree, watching the kids play as the awards ceremony for our son’s final cross country race of the season droned on. A man also watching from nearby piped up with unsolicited advice: several of the kids were playing in a way that might result in injury, he said. A friend of his had had a dangerous fall and lost an eye. I thanked him, mentioned his concern to the kids, asking them to be a little more careful, and thought that would be the end of our interaction. No such luck. 

“What do you think about the results of the election?”  he inquired.

“I’m not too happy,” I responded. “I would have preferred that the other presidential candidate had won.” 

“How could anyone support someone who violated the first tenet of public service?” he jibed. “I was in the Coast Guard. If I had sent classified information over a private email server, I’d forever have been disqualified from further service, let alone from becoming Commander in Chief.” 

Not wanting to get further enmeshed in a discussion that didn’t seem likely to have any positive result, I tried for more neutral ground.

  “Neither candidate was all that appealing,” I ventured. “I didn’t like the way Mr. Trump repeatedly insulted all sorts of people.” 

“Most of that was just bluster,” he responded. “After all, Trump is a brash New Yorker. Still, he has valid points. For example, there are over 750,000 illegal immigrants in New York City alone. That makes more than 10% of the population. They’re taking up housing that should be available to those of us who were born here or came here legally.”

Succumbing to continuing a conversation that I was pretty sure would end badly, I responded, “In the part of the country where I live, immigrant labor harvests nearly all of the crops. We need those workers.”

“Ha!” he said. “There was a time, during World War II, when there was a shortage of local workers for harvesting because so many American men were serving overseas, so immigrants were allowed in. However, once the war was over, Eisenhower sent them all back where they came from.”    

‘Look,” I said, “it seems that you and I view very different parts of reality. I respect your perspective, but I can’t agree with it.”

“Lady,” he said, “you’re entitled to choose your opinions, but you cannot choose your facts.”  With that, he strode off, triumphant. 

Not one gifted in coming up with snappy responses, I later did a bit of further research. Estimates of illegal immigrants/undocumented workers in New York vary widely, with the high-end 750,000 figure most likely coming from an online 2015  post by conservative-leaning publisher Newsmax. Every article about immigration has an editorial slant, explicit or implied—emphasizing either the costs or the contributions of this section of the American population. Some sources stress the need to keep families together, the need for skilled workers, or the aging of the U.S.’s native-born population. Others emphasize the need to reserve jobs/housing/advancement opportunities for the native-born. During the presidential transition, immigration remains a hot-button issue. 

I don’t have a pat answer. I’m not hopeful that our national immigration policies will become either more sensible or more humane for the near-term future. However, I believe that there is a way toward consensus in our nation of immigrants.  As recently as 2013, the U.S. Senate managed to pass a comprehensive immigration reform bill,  S.744, with a substantial bi-partisan majority.  No comparable bill has so far passed the House of Representatives.

To further chances for reform, I believe we’ll need to listen to each other’s stories more deeply. We’ll need to broaden our array of “facts” beyond the mostly biased reporting we’re being subjected to on all sides.

To get a more general view, I turned to a different Internet site and searched for TED talks about “search engines.” I found the following brief talk by Swedish journalist Andreas Ekstrom—“The moral bias behind your search results.”  (http://www.ted.com/talks/andreas_ekstrom_the_moral_bias_behind_your_search_results/transcript?language=en)  

Ekstrom uses two examples of image search results that got temporarily distorted by massive bias: the first, a 2009 derogatory representation of Michelle Obama, was quickly removed by Google; the second, an equally derogatory 2011 representation of mass murderer Anders Breivik, was allowed to continue until it died out on its own. Ekstrom concludes: “And I (emphasize) this because I believe we’ve reached a point in time when it’s absolutely imperative … to remind (ourselves) that that wonderfully seductive idea of the unbiased, clean search result is, and is likely to remain, a myth.” 

Let’s choose our “facts” carefully.