Friday, September 10, 2010

The dangers of obscenity tagging ...

I was pulled up sharp by this report about an XBox user whose account was suspended because he used "offensive language" in his details, the offence being that he lived in a city in the USA called Fort Gay.

I'm continually astonished by the stilted sensibilities of some Anglo-Saxon communities (I have never come across policies banning "offensive language" in systems for any other language) - like most of my fellow Europeans I can't imagine why the word "gay" should be deemed offensive. But this incident highlights the problematic use of obscenity lists which cause problems all the time for large numbers of innocent Internet users.

Gay, for example, is a common given and family name in the English-speaking world, and there are many streets and places named after those people. There are many people with the surnames Duck or Mouse, and some are called Donald or Michael; and they have a constant battle to achieve anything online. The good burgers of Scunthorpe in England have a very trying time with web sites and spam filters (think about it); and imagine the issues that the inhabitants of Dildo in Newfoundland, Condom in France and Fucking in Austria have.

If you do feel the need to check for obscenities, it needs to be done in a more knowledgeable, culturally-aware way than is currently the case.


Karen Lopez - said...

I have also seen terrible implementations of obscenity screening.

However, most of the time I've seen it done is not to block customers from submitting aledgely bad data, but to block company staff from adding data to accounts that could, if exposed, have negative impacts on the company.

For instance, many CRM systems have screening (not usually blocking) in place to stop CSRs from adding notes and records that could be perceived as insulting.

I've also worked at companies that have firewalls that block mail and access to websites with what are perceived as offensive. These "nanny" walls can get in the way of real business, especially as you note, the international boundaries of the internet.

Kathy Hunter said...

Obscenity screening should be just that - screening. Before a person is blocked, someone should review the suspect blockee. That's usually all it takes to separate the real obscenity from the machine-perceived obscenity.

However, obscenity screening does have a legitimate place. I have had several experiences where employees of a firm embedded obscenities into customers' addresses to get revenge or just to tease.

When dealing with international addresses, the person who reviews needs to be able to impart cultural wisdom to the process as should the person or persons controlling the list of potential obscenities.