Google's search results are not relevant, example: PDF

So he wants info about 'pdf reader software' but only searches for 'pdf' and doesn't get what he wants? Duh. Guess what happens when you search for what he really wants?

Google's search isn't psychic. It also isn't personalized. It will give you the best guess it can make based on your search terms. But just like almost everything else in the world, if you don't supply enough information, you won't get what you want.

If I want a car repair shop, I don't search for car. If I want a pet crematorium, I don't search for pet.

They are relevant to him.* He is currently logged in, and his location is visible on the left side of the image. Google uses his previous search queries as well as information about him and his location to provide a relevant result. He has probably searched for Java and R before. If you try searching pdf in your own browser, it is likely there will be different results.

*I meant that it is what Google thinks is relevant to him based on the data that they have.

I think Google can be proud that a broad query with only 8 out of 10 results relevant is a newsworthy occurrence.

This is the definition of a corner case. Google is smart, but it can't read your mind. Give it a hand, and think like a computer.

The most desirable results for PDF (and, by extension, most file extensions) are

1) Information about a program that can read or write PDF files

2) Information about the PDF format

The fact that both are shown as the first two entries is desirable. And the fact that wikipedia rounds out the top 3 is great.

So in response to a vague and ambiguous query, you got highly relevant results including the one you wanted on the first page. And you're complaining about that?

Incidentally, the R Programming Language manual contains stuff like this:

  pdf() Produces a PDF file, which can also be included into PDF files.

so it's quite relevant too. Admittedly, there are probably a few bogus results here due to incoming links like "Download the PDF file here". But given how vague your query was, and how common PDF documents are, it's remarkable that Google did so well.

Before your next post you might like to read up on search engines, tf/idf, pagerank, and what 'relevancy' actually means.

The only irrelevant thing here is your complaint.

Guys, I can understand on when you say query is lacks context but does my Point is those two results are fully irrelevant to my query. its maybe its I have held bar for search engines to high but I don't expect them to read my mind. Because most people when searching with "PDF" keyword will look for:

* pdf reader * pdf editor * pdf converter * pdf search engine etc etc.

But Java or R programming language results, that's not what I am NOT looking for. these two results are not even related.further, for other queries too -- I see ton of such irrelevant results stuffed into google which have to NO relation to original search term. This wasn't always like that -- earlier I used to get better results.

That's why I used this query to express my opinion that search result quality has gone down in past few years & you are now more likely to find spam results for your queries.

--Cross posted from Google Plus

The cause of this is probably that those documents have been linked to with anchor test like "this PDF". Do a search for "click here" to see a similar phenomenon.

So if you want to order pizza, you ask for "Food" ?

Anyway, searching for PDF might well mean you are looking for any pdf file (to test your shiny pdf reader maybe?) so its reasonable. I m surprised that there's no wikipedia result for probability distribution function though.

I don't think this is a valid argument. By the very definition of search engines and the web itself, tech-related stuff will be more highly-ranked than less-techier alternatives (to a certain level, obviously).

It's a self-reinforcing cycle. My favorite example is Googling for "wine" where the emulator comes in first place. There are many other similar examples as well.