Perri Nelson's Website 

Babelfish, blogging, and a new ranking scheme

There are times I wish I was multi-lingual. As it is, I only really learned to speak one language, English. I took a semester of French in junior high school, but only learned how to say I am a pencil, or something like that. I did enjoy looking at the Asterix the Gaul comic books though.

When I was a very young child I could carry on conversations with some of my Turkish friends who didn't speak English, but I don't think I ever learned enough Turkish to have made much of a difference. I certainly can't speak the language now.

Over the years I've had friends who spoke other languages. In high school there was a girl who spoke fluent Spanish. When I moved to Washington, a coworker spoke Russian. A neighbor across the street speaks Japanese.

Some of the blogs on the blogrolls I participate in are in Italian. Today in the comments julia974 said "Hi". Her blog is in Italian.

I don't speak Italian, and I certainly can't read it. That's a bit of a bummer too, because it looks like an interesting blog. Fortunately there's Babelfish.

Or maybe that should be unfortunately there's Babelfish. Babelfish is an automated translation site that can translate several different languages to other languages. It's pretty cool, but it's still a bit awkward.

For one thing, it only translates words, and it obviously can't translate every word. It doesn't translate idioms, so some things are really hard to understand even after translation. Finally, it doesn't adjust the syntax of what it translates. This makes it a chore to read the translated text.

Still, for a person who can't speak Italian it's a wonderful thing.

Julia's blog is called "La Toga Strappata", or as Babelfish renders it "The Torn Toga". There's a post on her blog I found fairly interesting, if hard to read, even with Babelfish... Here's a sample of the translated output.

From some year to this part, the WEB, the virtual ?rricchito or impoverito net, to second of the linen, points of view, of migliaia of weblog per diem on.

Lately, also the politicians, between which the min. of infrastructures, Of Peter ( you read its blog.).e the min. of the communications, Gentiloni ( you read its blog.), they are using this new way to make communication, putting to knot their political ideologies on blog personal, quite acquiring of the screw interactive, like in the case of Of Peter, on Second Life, the new phenomenon, you allow to add, to say me little maniacale, of Internet. But, the argument that we will face in post, concerns the affirmation of the American writer, books of fantascienza, Bruce Sterling, ( reads its blog ) which it has supported, to the South by Southwest Festival di Austin, than from here to ten years, stragande the majority of blog on linens sar?estinata to extinguish itself.

Like the dinosauri of the preistoria.

If I understand her post correctly, and I'm sure there's a lot I don't understand about it, there's the prediction that in ten years blogs as we know them will be extinct. Partly because in the long run they become boring, even to their authors, but mostly because of the schemes that are used to position blogs in the blogosphere.

I think she's talking about a practice known as link whoring. Blogs are ranked by the number of incoming links, and there are a lot of schemes to get lots of incoming links. One is to participate in a bunch of blogrolls. Another is to participate in linkfests.

These are great ways to get a blog noticed, and to some degree they drive blog traffic, but if they're all a blog has to offer, then the blog isn't going to last. What truly distinguishes a great blog, and what ought to be used to rank blogs is the quality of the content.

Julia says (Babelfish translated) "E' rare to find a blog rich of news, contents, value." In that, I think she's right.

We need a new blog ranking system. That is, we need one if we're really interested in ranking blogs. Maybe we shouldn't be.

I'm interested in writing about stuff that interests me. Sometimes it's politics. Sometimes it's news. Sometimes it's software. Sometimes it's just about something that happened.

I'm also interested in having people read what I write. If other people don't read it, then why write it?

I suppose I could just write whatever I felt like writing, and not participate in blogrolls to get links. I tried that for a while. The only traffic I saw was from my own browser.

It was only after I had passed some magic threshold that I started to see traffic on my site. At first it was one or two hits here and there, from other members of the same blogrolls. There weren't a lot of return visitors.

Maybe that said something about what I was writing. It either wasn't very interesting, or it was pretty much the same as what other people were saying.

After a while, I started to see an increase in traffic coming from search engines. So people were doing searches and my site was popping up in the search results, and people were finding my site by accident.

Now I receive between 40 and 60 hits per day on the site. That's not a lot of traffic, but it is enough to keep me at it. Somewhere between 5 and 10 of those daily hits are return visitors, regular readers of the blog.

Those regular readers would have never found my blog if it weren't for the massive proliferation of links provided by participation if blogrolls and linkfests. The search engines wouldn't have turned up my site in results if there weren't enough incoming links to drive my site up in the query results.

All of this is because incoming links are used to rank websites. The fact that about 90% of the people that visit my site don't come back says something else to me.

It says that either 90% of the people that visit a web site didn't find what they were looking for, or that they don't find it interesting enough to come back.

We really do need a new way to rank blogs after all. I think Julia's got it right. We need a system to rank blogs based on the richness and quality of their content.

I don't think that's an easy thing for computers to manage. After all, a computer is a horrible judge of content quality.

Well, maybe not entirely, SPAM filters are getting pretty good at weeding out junk mail and junk comments, but they're generally looking for a few keywords and phrases. They still trap legitimate mail and comments though.

People are still the best judges of content quality. To be really effective, a new blog ranking system based on content quality will need to be based on human input.

The pieces all exist in one form or another. Some sites allow people to rate the content. What is needed is a system that allows people to rate the content of a posting, one vote per I.P. address per posting, and aggregate the ratings in a central location, and calculate relative rankings on a post-by-post and blog-by-blog basis.

I think I may know how to implement a system that can do that, after having observed other systems that bloggers use. If you think this is a good idea, let me know in the comments. If you have suggestions to make it a better idea let me know too.

Let me know if you think it's a bad idea too. I'm interested in knowing why it's either a good idea or a bad one.

Personally though, I find the concept interesting enough that I'm going to try out a prototype of the idea. After all, it's something new to code up.