Slashdot Mirror
Putting Google to the Test
Big Nothing writes "Google has built its reputation on being the fastest and most accurate way to find information. But is the internet really the quickest way to access facts - and get them right? The Guardian puts Google to the test against more old-fashioned methods."
To win radio trivia call in contests a few years ago. Using Google to answer trivia just like in Ghost World.
Great comparison but they don't take into account how long it takes to get to the library, phone charges, etc. For me, 15 minutes on Google is faster than 30 seconds at the library.
Instead of comparing them against each other, its more important to use both internet based as well as "old fashioned" resources together. Its important to realize that hard backed enyyclopedias are better than google and wikipedia for some things, and not for others. The younger generation needs to learn how to recognize what source to use, instead of automatically going to google. The internet should not replace old fashioned resources but merely embrace them.
the byproduct of years of oppression by the white man
What is this li-brar-ee thing you speak of? That must have been where people accessed the internet before computers...
That none of the questions included something likey what is the maximum sustainable speed in Mb/s of the alcatel 8100 series router
Thats the stuff where Google with kick everyones trash, not complete list of authorships
He didn't count the time it took for him to leave his office and drive to the library. So add another 20 minutes to all of the library times.
I'd have to say that google deliberately alters certain rankings. The proof is when you do a google search for search engine. By almost any measure of page rank google would have to be listed first, but it is not. This means that google deliberately lowered their rankings. It makes sense that if you searched for a search engine on google, they've already got you and it doesn't really matter at what position they rank themselves. But by doing so they can potentially shield themselves from antitrust issues, and help hide the likely fact that for other categories they do alter the rankings for their advantage. Up and coming competitors in other arenas than search can be quietly disposed of if no-one can find them.
This article assumes that the person looking for information already knows other means of finding information. I didn't know about the Who's Who book and even if I did, I might not have one handy. I think these results are a little less than accurate for most people. Also, the author directly phones some people. What if I don't know exactly who to phone? I think Google will win in that case.
... is YES. Books on, say, Napoleon, you can find in a library. But ever tried to find, say, ten pages, or even a paragraph, about Napoleon? Especially in a 1200+ page book, or several? Ever tried looking something up in a card catalog? Or finding one sentence in a huge book? The answer to the question in the article is a definite YES.
How does the "Library" provide sub two minute results? The article doesn't say.
The comparison dosent seem to be so much as google / other means - its more of an Online V/s Offline means to search for specific stuff.
I think its wrong to brand Google as the only means to look for information online.
Secondly, the issues that the reviewer raises are also adhoc - they cant be used to generalise the entire deal / spectrum of infomation that people need / want / desire.
Try looking for a code sample that shows you how the GTK# can be used from Mono to display a Multi level Outline filelist. What are the options that you have for this in the Non - Online world ?
The guy already knew who to ask / who to talk to - what if you dont know that - what then ? how do you go about finding the best non-online resource to speak with / enquire from ? My guess is that you are going to be heading right back online.
What about the fact that the online resources / google are avilable to you when you want it - how you want it and where you want it. Ever looked up what a word from the bible meant in the middle of sunday mass at the local church using a Wap phone over gprs at wml.google.com ? Me neither....
I bet those who he/she called, immediatly fired up google to find the answer ;)
After reading the article, I feel there is a slight bias in favour of the libraries when looking at the questions. Of course a library has a master index of books of one author. Or - to find out about some very specific question about an event you immediately know what kind of journals to look in.
The only question really geared for search engines was the Thatcher quote (as that would be a full text search).
Would this be the time to create a true categorisation of questions to be used in comparisons? (Note - not the ACTUAL questions, so that search engines could optimise for them, but only specify the general direction of questions).
I admit, it would be pretty hard to do, but I guess it could be worth the effort...
This article, while interesting, failed to mention several other aspects that, at least for me, make google the best source.
:).
1 When using the phone, there are really two searches. The one you care about, and the one before that where you try to find the correct phon number. This can take quite a while in some instances.
2 I have to leave my house (which could entail getting dressed, which adds more time) and drive 4 minutes to the library. Once I get my online library account through the county, however, this will no longer be a factor
3 I actually have to have a conversation with someone on the phone. Google can be a more private experience, which depending on what I'm searching for, can allow me to better focus on finding the information I need.
4 With google and the library, I can have multiple searches running at once. With the phone, I'd have to pay extra per search.
5 With the library and phone, I can only use them during business hours. I can use Google 24/7.
"The best laid plans of mice and men gang oft agley..." - ROBERT BURNS
Had you even bothered to read the article (Duh, this is slashdot, what am I thinking).
You'd notice this was the whole test.
We asked various "pub quiz" type questions and then comapres the speed of response of various methods of finding the answers, such as telephone, library and of course google.
Then again I wouldn't have needed to write this post if you'd bothered to read the article.
The article report compare between Google, Phone and Library.
I think this is a biase comparison.
In the phone and library search, it is assumed upon a narrowed subject or particular topic. Where the searcher knows where to look for the _authorative_ answer, for example the title of the particular book to get the answer.
Overall, I think the winner is pretty inconclusive, but it still does shows that Google is a really good search engine - where you can actually find a reasonable result.
Hey, that's my password you are typing
The methods all have their uses. I don't use google find old classical music scores, but I don't use the library to find recent news stories, software, and such. There's a use for all these things, and it's important to preserve it all.
:-)
There's nothing that beats human interaction and direct knowledge in many cases, but people are not there all the time. If I had them right at me, I wouldn't need google. Google (and the library) is a compilation of what a bunch of people once knew, worked on, built further on, et cetera. Now, since it's impossible to reach these people, we wrote books. Books that we can read, to learn what people found out. That has it's value. Now, we can find the book, read about it, even read it, using google, or we could find other information rapidly that the library won't have for a long time - at least not before the next day's newspaper.
After all, the library might even have their search engine against a GoogleServer in the back room
Final point: Cherish all sources of knowledge, and use them appropriately. That will give you the best results.
I'm a researcher with Google Answers and it's never a must that we use Google to answer questions that come in. Actually, for one-off's and questions which I know can be answered with a phone call or two, I won't even bother with the internet. Of course, then again, there was that one time I called over a hundred different restaurants to see who was serving on Christmas Day. Hah.
-Christopher Wu
http://www.christopherwu.net/
Question 3: Who is the vice chairman of the all-party parliamentary group on back care?
Google - 6min 27sec (3rd)
Quote: "Unfortunately, "back" is rather a common word, and is turning up in all sorts of irrelevant documents..."
Entering "back care" in quotation marks got me the answer in 25 seconds, much less than either of the "offline" sources. If they're going to have an accurate test, at least make sure the person performing it knows how to use a search engine.
Or maybe I'm wrong; maybe most people don't have these basic searching skills, in which case the test is accurate after all?
-- If no truths are spoken then no lies can hide --
I would think it would take 5 mins just to find the phone number to the library, get some one to answer the phone and then explain the question, PLUS have them search for answer. This also would all depend on the library you call. Google is google no matter where you live, but not every library is staffed when the same people.
Dear God,
Google can search anything that can be searched.
Unfortunately, Google is unable to search my soul and desire.
You need to give me something better.
Henry, 19 years old
I've often said that I'd have to quit programming if Google ever disappeared. I lean on it for information in the same way that excessively using a calulator will lead you to punching in 1 + 1. In fact, I'm so good at it that people sometimes think I'm a genius problem solver, when really it's just a matter of creative googling on an error message.
The meek shall inherit the earth, in 3 by 6 plots. - Lazerus Long
it's interesting that your 'library searches' seemed to take only a minute or two to perform. You didn't bother counting the time it took to actually go to the library and find the relevant book, then?
I'm a Google Researcher. Generally most Researchers use the Internet or their own knowledge as their primary resource when answering questions. This is because most questions have a low fee and can be given answers which are merely pointers to further resources. Researchers are also encouraged to give the methods they used in their answer, to help the person who asked the question.
However, there have been quite a number of extraordinary cases where people have been so interested in answering the question that they've made phone calls, chased people, and dug out answers to incredibly complex questions. In one case, a Researcher managed to track down someone's obscure pre-20th C. German heritage.
Do remember that Google Answers is primarily for people who don't have excellent research skills of their own. While easy to use, finding certain things with Google (and other engines) requires skill, time, intelligence, and abstract thinking faculties that many people lack. Also bear in mind that most Google Researchers don't do it for the money. You will inevitably get a far higher quality (and longer) answer than you could possibly expect for the money. This is why tips are given to Researchers so often on the system.
I found the answer to the question 3 in about 30 seconds, well under the 6m 27s quoted by their researcher. It's clear from their comments about irrelevant pages that they hadn't enclosed 'back' in quotes to form "back pain", as '"back care" parliamentary group' puts the result on the third link (from google.co.uk). Also, it doesn't seem very fair to compare a researcher who doesn't think to use quotes round that expression with a librarian who knows to look "on page 242 of the excellent Vacher's Quarterly", a publication with which I (and most of the public, I would imagine) have no familiarity whatsoever.
Google knows nothing, except where words are placed.
But really, that test does not consider the fact that it takes a while to go to the library and that you actually need to get out of your house. Plus, library isn't available at night, neither is most people you can try to call.
Google sure wins any convenience test.
Qui ne va pas à la chasse n'a pas de gibier
PHP Queb
This isn't a comparison of anything...
The library searches don't include travel time. They also appear to only count the time it takes you to read the text in the book... not to:
a) Find WHAT book you want (Card catalog?)
b) Locate the book on the shelf
c) Find the correct page
All those things take the MOST amount of time, not reading the actual text. This is assuming that you KNOW what book you're looking for to begin with. I had no idea Who's Who would be a good place to look for the answer to the author's books. Google would have given me the answer pretty quickly without the need to know that information. How much more time would it have taken to find out Who's Who is the book you wanted?
Add on top of the fact that I'd have had to drive ot the library, and the time increases dramatically.
Calling a friend? Maybe faster, but I don't have many friends that would know answers like that... nor do I have the number to railway stations on speed dial... especially those in other countries.
Google is simply the fastest AND most convenient method to find the information. Or at least, if not Google, SOME search engine. If you're already at the library and KNOW what book you want, it might be a better choice, but seriously, how often does that happen? How often do you sit at the library and think of things you want to know?
I don't... I'm usually sitting at home reading, or surfing the web and come across something I want to know more about. Driving to the library to find that information would be ludicrous... and calling my friends regularly with mundane questions would cause me to lose what little outside life I already have.
Bleh... this isn't even an aritcle worth reading... jeez.
Yes they were able to get the required information from the library, but the only way to confirm that the data was accurate was to compare it to a known "up-to-date" source.
If I used the encyclopedia that was available in my High School library (in 1983) I would have learned that because of the recent Sputnik launch that man would someday walk on the moon.
Try calling Piers Morgan's press office at 4:00am. I bet your friend James won't appreciate you calling him when he's just sat down to dinner. The Library is a very poor information source when it's closed.
Google would beat any of those methods 'out of hours'.
MacBook Pro. Worst name since the Bicycle
So you're telling me that the first library test which took 20 seconds, involved looking in a card index, fetching the book and looking it up? Or did he have the relevant book in front of him already? - That strikes me a bit as cheating, otherwise I'd say it would take longer than 20 seconds just getting the book and opening it.
My <1000 UID is with a hot chick
Well, this isn't a very good test. Really what we learned is how effective one individual can be using google compared to other methods. Even then the tests didn't try to see how this applied to different kinds of information or how these results may have different from a group perspective. Worse still, the results for the author's first question makes me question if the author knew the answers ahead of time and had no way to call a result 'correct' or otherwise.
Also, more comprehensive searches at a library could involve actually having to visit the library... with it's associated drive time.
A good test would have had more questions, more participants and questions selected for a vareity of information types. The premise of the article I think is interesting: what kinds of research is the net really good for? Other than porn, of course, which is a given (try not finding it).
The problem with Google (et al) isn't finding information: it's finding reliable information (for most subjects). There's a hell of a lot of noise out there.
Cheers!
SCB
Google has some other advantages that phoning and the Library don't have:
1. Google is pretty much 'always on'. I can do a Google search any time of day where as I can't use the phone or the library at 3 am.
2. The ability to Find a keyword. Usually when I use a google search I use the google cache. This highlights the terms I am looking for so I can find them easily on the page. This is an inherent advantage of the computer over people or your eyes - scanning through text looking for what you really want.
humble and proud of it.
Libraries are expensive dinosaurs. All information in book or journal form should be digitized and put on the internet.
How about a meaningful comparison of electronic data retrieval services?
Compare Google to Lexis-Nexis.
Lexis-Nexis has boolean logic driven search (not natural language), and lacks "PageRank", but it includes all sorts of major periodicals not offered and certainly not archived on the web.
Lexis-Nexis would win hands down in all sorts of categories of questions.
It's an object lesson in the impact of intellectual property laws on access to information in our societies.
Ricochet was around $70 amonth, but at 20-60 bucks a week it more than paid for itself. Best thing, there were no rules that said you couldn't access the internet. People were amazed at my trivia knowledge.
Question 3: Who is the vice chairman of the all-party parliamentary group on back care?
A google search for:
"vice chairman" all-party parliamentary group "back care"
resulted in *exactly one* hit, a pdf document listing all parliamentary groups. A click on View As HTML, a find on "back care" and Voila, the answer took about 30 seconds to get.
An experienced googler can find things faster than they did. This particular case was just a matter of knowing the difference between words and phrases and putting quotes in the right place. But there are many other tricks (such as negation and using 'site:') that their google searches could have benefited from.
bp
- the value of the information presented is inversely proportional to the effort made presenting it
- the accuracy of the data is inversely proportional to the effort required to collect it
I've noticed that the most useful pages to me are the ones that contain plain text, like mailing list archives, for example. Maybe that's just an artifact of the type of information I'm typically trying to collect (technical questions, etc).The second point is just a general observation. When I was in school, the web was a wet-behind-the-ears DARPA project that nobody had heard of. To write a paper, I had to go to the library and look stuff up in books and periodicals. It took friggin' forever, but the results were pretty accurate. Now, I can type something in google and get a bazillion hits pretty much instantly, but I have to carefully search through the results to weed out lunatic fringe webpages (unless that's what I'm looking for), out-of-date webpages with no date on them, etc. I wonder how that affects kids today doing research papers? Imagine never having to go to the library, but, instead, having to hone your skills of scepticism.
If you love something, let it go. If it comes back to you, hump its leg.
Who cares if you don't know where to look for a piece of information? The reference librarian does. In larger libraries, there are usually librarians who specialize in particular fields of research. My university's library, for instance, has at least one research librarian assigned to each college or school within the university-- all degreed, and many dual-degreed in library science and their respective specialty fields. And they don't care in the least who is asking them for help-- it's not like the CS librarian will only talk to CS students.
Google is convenient, and fast for most searches, but there's a lot of information that just isn't available to it. Libraries buy access to that information, both in print and in databases, and they hire people to help you find the stuff you need.
The most important library skill, and the one that is most often overlooked, is recognizing the reference desk and asking for help.
Both the tests and the replies miss the most obvious problem: Google, libraries and friends answer different information needs.
Google is a fantastic way to find web sites. That's both the massive scope and the cramped limitation of it. It's up to you to sift through the web site result for the specific bit of information you want and then determine its accuracy. Google itself makes no claims on providing informationally accurate results, it claims to provide contextually accurate results.
If you want a significantly higher chance of information accuracy, a library is your ideal choice. For comprehensive information on the topic, a library is a better choice. You have experts on hand to steer you towards the most useful/reliable sources, and information pre-catalogued and cross-referenced for you.
If you want a an answer to a question that's particularly obscure, highly specialized, or couched in necessarily vague (or, worse, common) terms, a human expert is your best bet. If you want to find the last time the Milwaukee Brewers were over .500 in June, you talk to your baseball-enthusiast friend (substitute in appropriate football clubs and stats if you happen to be in the 90% of the world that prefers football). If you want to know the name of that one blonde girl your ex-roommate dated sophomore year, you call your ex-roommate.
Somewhat tangentially, the other glaring problem with most of the responses I've seen is they ignore the skill required to use any of these sources. Plenty of people have complained how they wouldn't know what books to reference or what people to call...often the same people who mock the author for not knowing what search terms to use. It's all learned skills. Google-fu is learned, not natural. Just like library research (anyone who's played Call Of Cthulhu should know that), and knowing who to call. Knowing how to differentiate a web site that's probably authoritative from one that's at best shaky is a skill that's really no easier or harder than being able to recognize a publication as reliable or a rag.
Anyhow: my point is that the article is neither right nor wrong. Google vs. libraries vs. phone-a-friend is a pretty meaningless question. They're different resources for different jobs.
Reality has a conservative bias: it conserves mass, energy, momentum...