August 23, 2019 0

Writing Successful Queries Pt. I: Introduction to Predictive Search

Writing Successful Queries Pt.  I: Introduction to Predictive Search

and get started. My name is Tasha
Bergson-Michelson, and I am a Search Education Curriculum
Fellow here at Google. Thank you so much for joining
us today for “Writing Successful Queries.” We’re going
to talk about the basics of predictive searching. So of course the question is
what is predictive searching? First, I really want to thank
both Jole Seroff of Castilleja High School and
[? Jeanine Zaccara ?] of [? West-Bat ?] who helped me think through
what to call this. In my mind, predictive searching
is a process where I visualize the source that would
be perfect, that would have my answer. I actually imagine what
it would look like. And then I use that picture in
my head to build a search to find the closest match to that
which I have imagined. Now this isn’t really
a new idea. I first read about it in a blog
by Karen Blakeman, who is a Search Trainer in the UK. Dan Russell, here at Google has
also written about it, as have many other people,
particularly in the library field. Just at the time that I wrote
about it, or when I first read about in Karen Blakeman’s
blog, I thought it was interesting, but I really
couldn’t see myself doing it. I’m not really a visualizing
kind of person. And so I just kind of, “eh,”
thought it was interesting and set it aside. And then after several years of
trying to teach different advanced operators and different
technical sides of search, and I’d found that
people would get really excited, but it wouldn’t
stick. And I thought, what is it that
I do as a searcher that is different from what most of the
people I’m talking to do. What is it that makes
it work for me? And lo and behold, I realized
that what I did was imagine my perfect source in my head. So I imagine that many
of you who are here today do the same thing. You certainly are welcome to
chat with our panelists and let them know if this is a
strategy you also use. We’d be very interested to know
if it’s a strategy you teach, as well. And actually very quickly,
before we continue, as I talk about chat, I wanted to tell
you that I spoke with tech support just before we started
today, and I realized that we don’t have the ability to let
you guys chat with each other. And so I just briefly wanted
to tell you why that’s so. We either have a choice of
exposing the entire attendee list and letting you chat, or
unfortunately, you can only chat with the panelists and
the host. And because we believe very strongly in
protecting user privacy, we keep the attendee
list private. As a result, I’m sorry that you
don’t have the opportunity to chat with each other. And I see the recommendation
that we have a Twitter tag, and in fact, I had thought
of that just as we were coming in here. So if someone wants to check
on Twitter and see if there is, say, predictive search
as a hashtag– it might not be– and you could start it up and
have a chat right there. And you can use Google
Realtime search. Let me show you how to
do that quickly. So what we can do is, if we
did a search for hashtag predictive search– woops, one word– search, predictive search, and
then if I go over– let me pull up my annotation tool. I go over here to the left-hand
panel and click on More to open it up, and then I
can click here on Realtime. And that gives me
Twitter results. And I see there is nothing right
now showing up for the hashtag predictive search in
Twitter, which must mean that you can go ahead, start tweeting
to each other with that hashtag. We can set it up later. OK, moving on. So we’re going to talk
about this concept of predictive search. So let me give you an example
of predictive search. Let’s say that I wanted to know
who the two lead actors are in the movie, Key Largo. When I do the search, I don’t
immediately see any results for the film. I see things clearly enough
about the Florida Keys and about Key Largo, the location. Now there are a number of ways,
if I want to find out who the two lead actors are
in this movie, that I can approach the question. The obvious one that comes to
most experienced Internet users’ minds is IMDb. And in fact, I can go to IMDb
and find out who the two lead actors are quite easily. But there are a number of other
ways that someone who isn’t as familiar with the web,
can go about finding this information. One of the things that I
absolutely love about predictive search is it draws
on, not just computer web literacy, but many kinds
of literacy. So some of the ways I’ve seen
various people approach this problem is they type in Key
Largo, and they go to “Images,” and they look for a
cover of a DVD or they look for a movie poster or
something like that. Sometimes people will go to
“Videos” and will see that there is, in fact, here
a clip from a video. And the snippet, this part
in black, does give the information of who the
two lead actors are. Of course, because I’m an
experienced web user, I also know that I could just type
the word, “movie” after or “film” and get information
about Key Largo. But if I’m not an experienced
web user, but I am a reader, what I can do is I can think,
when I read something that tells me who the stars
in the move are– what does that look like. And what I realize is that
almost every movie review I have seen in my life starts with
the phrase, “the movie Key Largo, starring…” and
then gives the stars. And indeed, we find this is a
very successful phrase that exposes the information
I want about who the two lead actors are. So it’s a matter of getting a
question in your head and, actually, stopping and thinking
about what the source that answers it would
look like. Ironically, you could just call
this method, in its most basic form, stop and think,
which I know that we all know that many of our students
may or may not do. I, in fact, had a parent come up
to me after a talk I gave, where I was talking
about this process of predictive searching. And she said to me, you know,
everything you talked about, it boils down to a really simple
point that had never occurred to me before. It just never occurred to me
to stop and think before I typed something into
a search engine. And the truth of the matter is
that for most people who use electronic search tools,
it doesn’t occur to them to stop and think. So why is it important
to do this? Well the best way to explain it
is to understand a little bit about how Google works. So actually, what Google does is
a number of Spiders, which are also called Bots, they’re
a program which crawls out across the Web and looks for
websites and web pages that we haven’t found before, or that
in fact, have new content since the last time
we visited them And then we do what’s called
crawling the page, which is, for all intents and purposes,
taking a snapshot of the content, which we send back to
our home computers, and we break the content down. So for example, if you look at
this Wikipedia result that we have right here, it says, “Key
Largo is a 1948 film noir directed by John Hudson,”
and so on. And what Google actually does
is it takes every single one of those words and places
them in our index. So that when you type
in, say, 1948– I don’t expect this to be the
first result– but when you type in 1948, what happens is
Google, actually, goes to 1948 in the index. And just like when you look up
a term in an index in a book, it pulls up all the locations
that we know of that are associated with the concept. So we do this, essentially,
with all the words you type in. So as you can see, when we
take– let’s take this top link again– when we take this
text and break it down, there are some really easily
identifying or words that are highly associated with the
concept, like year and 1948. And then there words like “was,”
and “a” and “on,” which are not quite as on topic. And you might think, well if
you’re a librarian, for example, you will have used stop
words most of your life, or you will have had databases
that just ignore those small words, because they
get in the way. But what we find is if you do
a search for “who,” you get the World Health Organization. But there might be other uses
of the search term you want, like “The Who,” if you’re
looking for the band, or in fact, “a who,” if you’re
looking for Horton Hears a Who. So we find that in many cases,
we have to take into account the stop words as well,
the small words. And so we can really find
yourself in a difficult situation, if you have, say,
a student who is in, 4H and wants to know, my cow has
blisters on its tongue– and I can’t spell, but
luckily this will be spell-checked for me– what’s wrong with it? And I’m sure that most of you
have seen your students typing in questions every now
and again, even as adults we do that. But you can see here that the
idea that’s needed is so diluted by typing in a question,
that we can’t even get anywhere within the
realm of the answer. So that’s one reason why typing
in a question won’t work, because we’re paying
attention, essentially, to all the words you type in. Along the same lines, we have
a student who types in a successful question like, what
year did the American Revolution start? What year did the American
Revolution start? I see I have another typo. And here what we get, as I’m
sure you were able to predict very well, is a lot of answer
sites,, Yahoo!Answers and so forth. Now the reason this happens is
when you type words into the search box here, and Google goes
and matches it with the index, when we go to rank the
answers, we use over 200 different signals to decide what
pages should come first. So for example, pages that have
your terms close to each other in the same order as each
other, and in fact, in the title, will all
be prioritized. And of course, where you could
click through to an answer site, what you find is the
question itself, that is asked by the user is made into
the title of the page. So when students search for a
question, they are naturally going to find those pages that
have the question as the title or a close match. As you can see, the third result
here actually asked what year did the American
Revolution begin. We have synonymized start and
believe that begin is another close match to that. So one of the reasons I really
like visualization is because a key point that you can take
away and that you can teach to your students is you need to
search not for your questions, you need to search
for your answers. First of all, that will help you
get to more academically valid content. Remember that Google is a
product that’s written for everyone to use. It is not written specifically
for an academic audience. And so sometimes we need to
employ some extra tricks and skills to get to that academic
content, but also just because it helps us more generally
filter the level of quality and of relevance by looking
for our answer instead of our question. So an example of that would be,
you know, someone might ask how fast can
a cheetah run? And if I just stop and think for
a moment, I’ll think what would the words on the
page look like. Now another thing that makes
searching Google hard is that we have this phenomenal
creativity as humans. Each and every one of us will
express the same idea in a slightly different way. And now our job as searchers is
to get the index to give us what we want. Our job as searchers is to
figure out how would someone else answer this question. What words would they use? And so I’d think about it,
and I think about terms like miles per hour. And so I know that I want
the cheetah, right? And so I think, maybe someone
will say, “cheetahs can reach speeds of up to…” See I can
imagine a page where it would say cheetahs can reach speeds
of up to x number of miles per hour. So now I can run this search,
and I can see in my results, although I clearly have
to check the results, because they vary. But I see that again and again
and again this phrase appears. And in fact, though I said I’d
talk about operators in the next talk, you can– you might be familiar with
putting quotation marks around a search in order to glue
the words together into an exact phrase. But I have this problem, because
I have one element of my phrase that I don’t know. I can, actually, use the
asterisk, shift 8, . I can use an asterisk to
stand in as a kind of fill in the blank. And now I’ll see that all my
results have exactly the phrase I want and that they will
give me the number that will change from site to site. So here I haven’t looked for a
question, I’ve looked for the words that I anticipate
to see on the page that will be the answer. Now I once showed this to a
phenomenal librarian that I know, and she immediately said,
oh yeah, so that would be like “what kind of speed can
a Ford Escalade reach.” “How fast can it get to its top
speed.” And I said, OK, so let’s stop and think. I want you to take a second,
because every time you’ve seen a commercial break in your life,
there’s been a phrase that describes how fast a Ford
Escalade speeds up or accelerates. And if you stop and think about
this, you’ll realize that it is– I don’t mean a Ford Escalade,
I’m sorry, it probably works, too– I mean a Ford Escape, because
I like their hybrids. So “0 to 60” in the number of
seconds is a phrase that I hear over and over again. And again, I like this example,
because it’s not drawing on any wild knowledge
of the web, though we have learned the quotation marks
and the asterisk, but a concept of what you might
want to search for. It doesn’t take a knowledge
of the web. It just takes having exposure to
a television or a newspaper or a magazine. And you’ve heard the phrase “0
to 60” in a number of seconds. So this is a fairly
common topic, the acceleration for a car. But I know that we often want
our students to be finding more academic information. You can apply the
same principle. So if we had someone
who is working on– excuse me– on tooth decay in children,
“tooth decay in children.” You find a bunch of different
results. But let’s say I’m looking for
statistics, and a lot of times, I’ll see people using
search terms like facts. You’ll see students doing a
search for “tooth decay children facts” or “tooth decay
children information.” And these search terms are,
I find, a little fuzzy. Because generally speaking,
there will be pages that people have put together– in this case you can see the
suggested search suggests that it’s information for children to
use, not about children and tooth decay. But also these words don’t tend
to be used as much by more academic sources. So I try to think I’m looking
for general information, and I want some statistics. What kinds of things
am I likely to find in an academic source? And one thing that I think I
might find is “table” or “figures.” Now you’ll notice
that I’m using a capital OR. That’s because I want to either
find “tooth decay children table,” those words
in a document, or I want to find a document that has
tooth decay, children, and the word, figures. And by putting in OR in all
capital letters, it makes Google search for either
table or figures. So now when I run the search,
I can see that a lot of the sites that I’m getting here, for
example, “Tables 1 through 8” are providing latest data
on tooth decay in children. I am suspecting that when I
click through on this site, it is, actually, going to
have a lot of good information for me. And as I look down my page, I
see that there are a number of sources which can work
very well for me. This came out of running into
educators who use real life math statistics– sorry, figures from real life
situations in their math classes, who are having a lot of
trouble, actually, getting to the numbers they wanted. And they hadn’t, kind of,
stopped and thought about what is this information going
to look like. It started with someone
who wanted world population growth data. And he spent an entire 45-minute
class period having his students look for
information on world population growth, and in the
same way, he hadn’t thought about the fact that it would
be in tabular form. And in fact, today if you ask
a student to sit and think about what world population
growth data would look like, they’ll say, oh, it
would be a graph. And they will immediately go
to Images to look for it. And while we may have a couple
of off topic, we also can find that there are a lot of tables
that appear in our results. So again, it’s a matter of
stopping and thinking, what is this information, what is it
going to look like, and what words can I draw off of what
I would anticipate seeing on the page? So, again, if you have someone
who hasn’t spent a lot of time on the web but, for example,
on the older end of our students who has had exposure
to academic writing, if you ask them to think about the
caption that would go with data they want, they can pull
terms from one of those captions and fold them
into their search. Now interestingly, I used to
teach students to do something like that if they wanted to find
a more academic source. That they would put in their
search terms, and then add search terms like bibliography,
or works cited or sources or suggested
reading– that that would all make for
very compelling searches that would bring up pages that had
bibliographies or lists of work cited on them that would
help them check the credibility of a page that
they were looking at. These days, unfortunately,
I find that the results can be messier. Although you can still do a
search like “cardiovascular system,” and if you know that
you want a bibliography, you can, actually, ask for it. These all make really great
search terms. Bibliography is a great search term. Album is a great search
term, table, figure– words that describe what you
want that are likely to actually appear on the page. Some words don’t make such great
search terms, like quit smoking documentation,
or death certificate documentation. Documentation is not such a
strong word, because I have trouble imagining it actually
appearing on the page. Some of these others
are great. Now if you’re a teacher in, say,
an elementary school or a middle school, you’re saying
it’s wonderful to find, say, Roman Army– if I do a search for “roman army
bibliography.” But a lot of these look like they might be
for a higher reading level. I should try this out
with cardiovascular. So let’s try this. You can, actually, add a site
search. site: is another operator, which we will talk
about later, but I’m going to cheat a bit and talk
about it today. And I, actually, know that
teachers are going to have compiled bibliographies about
the cardiovascular system, or classes will have compiled
bibliographies. And I know that a whole lot
of education, school sites, are at k12. and then the state abbreviation,
and then “.us.” And in fact, I can put an
asterisk here and do a search for site: no space k12.*.us
and run the search. And then I’ll find that I have
a number of sites from different schools and that
they’ll, perhaps, offer me bibliographies on the
cardiovascular system, and they’ll be across a number of
different grade levels. So it’s an interesting question
finding the right source for the right kid,
because we often talk to our students about selecting
scientific words. For example, one of our favorite
ones here is we point to that little indentation above
our lips, and we ask, what’s that thing called? It’s really funny, because,
actually, if you do a search– if you think about it,
what’s it going to look like when someone– the web page that has that
question, it turns out that “thing above lip” is a pretty
good search, just to get the proper terminology. Or another one that works is
just “indentation above lip.” These are all ways to just get,
if you need a piece of vocabulary, you can ask that
that casual language question. On the other hand, once we start
getting vocabulary, we find, if you look at a bunch
of sites, you’ll find that it’s called a philtrum. And in fact, you can look it up
in the dictionary to make sure you’re right. But once you search for
philtrum, you start getting much more technical
literature. For example, you get information
about what medical personnel should look for in a
philtrum as a diagnostic tool. You will, in fact,
see there that there are Medline sources. And you’ll find that there
are a whole bunch of edu and gov sources. You’ll find information about
normal and abnormal philtrum. So once you’ve hit on that
technical term, you’ll get a lot of technical information
with it. You will often teach students,
find the terms of art, find the technical terminology and
look for it In order to find more academically valid
information, and that’s absolutely fabulous. On the other hand, what we see
a lot of here on our search education team is that once we
teach people to look for technical information, they
start writing ridiculously over-technical searches. So you’ll see people starting
to, if they just want to know what this philtrum is, they’ll
start searching for things like “infranasal depression.”
And there are people who will write really long searches and
start piling on the search terms until they either are
only finding things so complex. that they have no hope
of reading them or they just end up with no results at
all or just really messy, bizarre results that are off
in some corner of the web. So kind of the next step beyond
visualizing your source and thinking about what the
words are that are going to appear on the page, the next
step is thinking about what kind of page is it actually
going to appear on? So we’ve talked about avoiding
answer sites by not typing in the question, but along the
same lines, sometimes an answer site’s exactly what you
need, as we, actually, saw with this “indentation
above lip” search. We don’t need anything beyond
what the answer site– if you just want to know what
something’s called, all you need is a answer site and
then check your answer. So an example of this, is my
supervisor– he and his son like to go to basketball
games. And there is a song at the
Stanford basketball games that everyone sings in the audience
to cheer on the team, and the lyrics are, “oh oh oh oh oh.”
And one day our boss’ son said to him, hey, dad, I’d really
like to download that song onto my mp3 player. I really like it, what’s
it called, will you help me find it? And he thought about it, and
he said, you know, I don’t think you can search for that. There’s nothing searchable that
we know about the song. So his son said, oh,
let me just try. And he actually did the search,
“oh oh oh oh song.” And in fact, if we clicked
through, we’d find this is the proper one. If you listen to this
YouTube video, it’s also the proper one. And in the comments at the
bottom, the song is actually identified as being
by Zombie Nation. And it’s called, I think,
Kolkraft 4000. But again you can then find
this song and play it and confirm that it’s
the right one. The interesting thing to me
about this story is not just that sometimes the obvious,
simple answer will work, but that if you stop for a minute
and think, OK, I want to identify a song, where is that
going to happen, what kind of source is going to
do that for me? The clear answer is, well,
an answer site. And there are probably
some discussion lists that would be helpful. I think there’s a “what’s that
one called?” in LiveJournal. You know, but there are these
informal meeting places where people discuss these
questions. And the key word in my mind
there is that it’s informal. An informal source calls for
an informal search, just as much as a formal source calls
for a formal search. So at this point, we’re
really moving– we start to move from predictive
searching as an opportunity to draw on any
kind of literacy you have available to you and start
moving into an area where a greater familiarity with what
the web offers provides more opportunities to run really
great searches. And this is something we’ll be
looking at a lot when we talk about advanced operators
next time. But actually to make a quick
note, I should point out that predictive searching
is not just an online search technique. Because it’s fairly common, I
think, that if you stop and think what is this source going
to look like, that what you actually find is it might
not be an online search. For example, when I wanted to
know what kind of cherry tree could I grow in my backyard
in my neighborhood and be successful growing cherries,
there are places I can think of online that would answer
that question. But the fastest way was really
for me to just pick up the phone and call my local nursery
when I considered all my sources. And in fact, I know, too, that
my librarian could have given me books at the public library
that talk about what grows in my area. But the most specific
information is from the nursery down the road
from my house. Another example of that is I
heard that Napoleon kept up a big, long term correspondence
with his wife, Josephine. And when I think about that
correspondence, I think, you know, I’d really like
to see that. I wonder if it would show me
a totally different side of Napoleon than the one I learned
about in history. If you’ve heard my webinars
before, you know that I really like primary sources and
especially reading correspondence. So I wanted to know
about this. And when I think about it, I
think probably that is going to appear most reliably
in a book. And it’s interesting because
when I do a search and get it spell checked, thank goodness,
I find that, actually, the. top results are,
indeed, books. So if I’m in the library while
I’m thinking about this, I can see if this book is
on the shelf. If I want to see what I can find
out while I’m still at home in front of my computer, I
can, actually, go over here to Google Books, I can click on
Google Books, and indeed, the book is right there,
and I can view it. And if I click through to the
book, I have the opportunity on the left-hand panel,
here, to find the book in the library. And so, if this is my zip code–
or I can change it to match my own– then I– you may be familiar with
WorldCat– but here I can see where the book is available
near my house. I can, if it’s available at a
public library, I can click through and put it on hold in
their catalog, or I can figure out which university library
I have access to. And I can go see the book. Myself, I can also read
portions of it here. This one is, actually, a free
book, which means I can read the full text– as much of it as
I need online. So I have a number of options. To me, this is the most
reliable source. To me, it seems natural that it
would be in a book, and it was just a matter of thinking
that through in order to find the right place. Now of course, the flipside
of this is I need to know what’s possible. If I’m not aware that Google
Books exists, I would never know to come over here
and click on it. In the same way, one of the
things that we find often in our Search Education Department
is that people’s ability, not just to find the
information they want, but to ask questions that would help
them find the information they want or even to desire specific
information is based on their knowledge that
information even exists in a certain way. Let me give you an example. I might want to know, for
example, just to draw on recent news, [TYPING] “…osama bin laden.” I can
learn a lot from reading online news or reading my
newspaper or listening to the radio, watching the TV about
Osama bin Laden. But I would tend to think, if
I wanted to know what Arabic sources, sources written in
Arabic, were saying about this event, I would be dependent on
the interpretations that I’m seeing on the news, but maybe I
want to find out for myself. If I know that I can come over
to this left-hand panel– let me show you– I opened more search tools,
and I’m looking here. And right here, there is
translated foreign pages. Now unless I know this exists,
it will never occur to me to look at results in Arabic. But here I have a source that
will, actually, take my search and will translate it
into Arabic for me. It will bring me back
the results, translated back into English. Well, maybe I want to know
what people in France are saying about it. Or maybe, if I click on this
triangle here, I can find out what people in Japan are saying
about this subject. And unless I know that this tool
exists, I’m never going to be able to even really ask
that question thoroughly. The same way, if I wanted to
know, OK, so Osama bin Laden– I’m very curious about this
spike here that I see. If I want to know, OK, Osama bin
Laden is in the news now, and I know about 2001, but I’ve
heard that he was around before, and what
else did he do. I’ve seen surveys that say that
teenagers are a big group that’s been asking about the
history of Osama bin Laden over the past week. If I wanted to know that, if
I didn’t know that I could, again, come over to this Google
left-hand panel and a little further down– pardon me, just a moment. If I didn’t know that I could
come into the left-hand panel here and look at Timeline,
then I wouldn’t know that there was necessarily a source
on the web that looked for the search term, Osama bin Laden and
looked for dates related to it, and mapped it out for me
in this graph, that I can see the spikes of the events
that are talked about online in his life. So it’s an interesting problem
we have. As we work on predictive search skills, we
need to help students think about what’s out there. And we need to help them think
about how to draw on the literacies they have available
to them, the different types of information they know or they
could know or they should know in order to visualize
this source. On the other hand, I have seen
it work very successfully with kids as young as, say,
third grade. I had a group of third graders
who encountered the idea of official websites, which
they thought was hysterically funny. I have an opportunity to
misspell something else that I should be able to spell, here. Roald Dahl– I’m sorry, I always spell
his name wrong. Oh, thank goodness. So I had a third grader, who
looked up Roald Dahl and ended up on the official website,
which, of course, played music and all kinds of
awesome things. The other kids in the class
wanted to know where she was. And when she ran the
search, one of the kids noticed the title. And he said, official,
that’s hysterical, what does that mean? So we had a really long
discussion about what “official” means in the
title of a website. Well, interestingly enough, the
next week it came up that Rick Riordan, the author of
the Percy Jackson books, himself, read the first
chapter of each of his books online. Unfortunately, it’s since
been taken down. But at the time, you could hear
audio recordings of him reading the first chapter
of each of his books. And the kids were,
like, awesome, where can I find that? And I said, well, where do you
think you can find it? And you could see them, they
stopped and they thought, and they said, official, official,
it would be on his official website. And so from as young as
third grade, I saw kids be able to learn. For them, it was stop
and think– what do you know, what kinds of
literacies do have that you can draw on that will help you
figure out where to find information? So these kids were huge, huge
readers and this particular group was very articulate. But I think from fairly young,
we can start framing questions in a way that will help students
draw on what they know in order to write really
great searches. The world population growth
question, anything having to do with graphs, things that are
related to visuals work really well with younger
students to get them thinking, what would this information
look like, would it be a paragraph of text, would it
be best in a pie chart? And it’s, actually, interesting,
because it’s a flipside of the standards, the
AASL and the other information literacy standards we have out
there, where we want students to learn the best and most
efficient and effective ways to present different kinds
of information. This is just turning
that on its ear. How do you think other people
will have presented it, and what makes that searchable? So we’re going to look at
this more in a second . I split this into two, because
there are a lot of operators, and they take some time to talk
about, and I want to make sure we have a chance to really
look at them in depth. So we’re going to be looking
even more at new ways that you can draw on how you envision
sources in your head and how that ties into using operators,
but for now to step back to just using visualization
and predictive search to choose keywords. I had another interesting
conversation with a librarian who was going to be working with
a lot of students on a paper, and she thought a lot
of them were going to do childhood obesity for
their subject. And we were talking about how
to find statistics, and how to, kind of, compare statistics across different sources. And it was interesting. We were thinking about when
you look for statistics, you’ll get a page that’s put
together by someone, and you can, maybe, find a very
good quality website. Perhaps the CDC is a highly
authoritative source, something like that. But if you want to look just
across different sources or if you want more commentary
on the statistics, how do you do that? Like if you’re looking for a
forecast, the problem is that “forecast” is a difficult word
to search for, because it has a lot of synonyms. So it could
be forecast, it could be predictions, it could
be estimates. There are all these different
words that are associated with this idea. But when we thought about it,
what we discovered is that if you imagine the sentence in
your head, it would be forecast that something to do
with childhood obesity by the year, and then there
would be a year. And something in 2011, that year
would probably be, like, somewhere between
2015 and 2050. So we thought, what would happen
if we just drop the obvious key word that we’re
trying to catch, “forecast,” and instead search for the terms
that we expect to find right next to that– “2015,” and I’m going to sneak
another operator in here. If I put in two dots surrounded
by numbers, Google will look for the first number,
the last number and everything in between. It will look for a
range of numbers. So if I run a search, “childhood
obesity,” and then the phrase “by 2015..2050,”
what we should find is a number of phrases where any year
where there’s a statistic associated with the phrase
“by” and then “that year” shows up. So you can try this– I see a lot of 2015 here, but
you can try this with other trends as well. You can try immigration, and
you’ll find statistics. You can try diabetes, which I am
thinking of because of our first subject. And you could do the same
thing with the idea– with a phrase like “from..” some
statistic in somewhere between 1940 and 1970 to another
statistic in, let’s say, 2005 to 2011. Here, what I’m doing is, I want
to look for statistics to do with diabetes and a phrase
that tells me there was some old statistic relating
to it and a new, more modern statistic. And so then I will see I
get results like this. So these searches are a little
funky and complicated looking. I absolutely grant you that. But I think it’s a really
interesting– this is an idea I’ve been
wanting to play around with. Sometimes we want keywords. We want search terms. But
sometimes when the search terms are messy, when the
keywords are messy, there are associated search terms that
appear near them, maybe like even a boilerplate language
that we can use. And we can search for the thing
that we know will come next to it. I guess that’s what you’d say
with the old search we did at the beginning for who the
stars of the movie were, although that’s more
a keyword. But I think if we start thinking
really creatively about the language that we
experience as readers, we will find that we have a
lot more options available to us as searchers. Now the funny thing is on
the other end of this. So I had a student, a fourth
grader, and she had to write a paper. She was allowed to pick any
topic, and it was supposed to be a mini-research paper, and
she was supposed to spend just not too long doing
research on it. It was practice with the
elementary research process. And she wanted to know– she wanted to just
write about cats. So she decided she wanted to
write about how many different kinds of cats there were. So she thought about
what is it going to look like on the page? And she picked the search,
different breeds of cats, because she anticipated
seeing– “in the world there are ‘blah’
different breeds of cats.” So she ran the search, and in fact,
she seemed to get really great results. So you’ve got all these pages
that talk about the different breeds of cats that
are out there. The only problem she had was– here’s one that says
there are 43. There’s a list of 43 different
breeds of cats. And then she clicking on
different ones and finding that each one had a different
number of breeds of cats. And in fact, actually,
they were all listed. Most of them didn’t say how many
breeds there were, and she was counting them. So she was going, “one, two,
three, four…” on this tiny screen and losing count,
starting again, went through several sites, got different
answers on each site, counted again to make sure they were
really different, and called me in tears. And so I think that not only
do we need to teach our students to use prediction in
order to think what is a good source going to look like? But also when you’re looking at
your search results, spend some time talking about how do
you predict what a given result is going to look like
when you click on it. And there a lot of ways
we can do that. We have the title
of each page. We have the snippet, which is
this text in black, which is the keywords and the context
as they appear on the page, and then there’s the URL. And I have, actually, found,
again, that same class of third graders was very
good at parsing URLs. They found the vocabulary
available in URLs to be easier for them to understand than a
lot of the other vocabulary they were seeing on the
search results page. And so if they were looking for
scientific information, I’d have students say, well,
this page has a web address that says it’s a trivia page,
and I’m not looking for trivia, I’m looking for real
scientific information. So I found that it was very
productive to talk to them about how to read URLs. So this girl saw that these
weren’t really that authoritative looking. And what she actually found– the results have changed a bit
since she did this search– but the particular results– it looked like different breeds
of cats, which had made sense as a predictive search,
actually was not formal enough for what she needed. And in fact, one of the top
results that she came across at the time was from a website
she didn’t recognize. It was a “.org” called CFA. And it said that she
recognized– it said that the CFA recognized different
breeds of cats. And she was really curious
about that. And so she thought, well that
sounded much more scientific. And so she tried the search,
“recognized breeds of cats,” and found out that the CFA was
the Cat Fanciers Association. She then determined that, in
fact, there are groups out there that talk about how to
define a cat and who are the gatekeepers for cat shows and
for cat breeders and what’s a breed and what’s not. Now in point of fact, that is
hinted at here in this snippet from, of all sources,
Wikipedia. It says “animals may be
considered different breeds by different registries.” The
question is what is a registry, and it turns out
that the Cat Fanciers Association is one of those. The reason I bring this up is
that not only should we be teaching students in the
predictive process to think about what they’re looking
for, look at their search results and think about what
do I think I’ll find when I click on this result. If there’s something that says,
“cat breed profiles, personality and breed
descriptions with photos, different cat breeds, pictures
and information,” does that sound like it’s going to be
authoritative for me? How do I make that decision? Also, I think it’s really
important to teach students when they are looking at their
search results and predicting what they’re going to find, that
they shouldn’t look at just one result at a time, but
they should look at the whole page, because I pull a whole
story out of this page. I see that “different breeds
of cats” is a search that brings back very informal
sources. What I’m looking for
is more formal. And I see the suggestion here
that, perhaps, I’m not asking even the right question to begin
with and that I need to be asking, how is a
cat breed defined? And that may cause me to
reconsider what I want my search to look like and to
go back and try again. And you can see that in lots
of places, for example, my coworker, Trent, introduced me
to this wonderful story about Franz Ferdinand, who was not
only the Archduke that we know about from World War I History,
but also now a popular band. And she recounted to me a story
of an interview they had where the band Franz Ferdinand
said, “we’ll know we’ve made it when we’re number one in the
Google search results.” So now it is incumbent upon our
students to be able to look at this page of search results and
not go one at a time, but to look and predict. The second one, here, is fairly
clear that this is going to be– it tells
me it’s a band. This one, here, actually tells
me some search terms that it would be very smart if I go and
fold into my search for better results. Some of the rest are not quite
as clear, and it takes a moment to look and decide which
ones are on my topic and which ones are on another. And that comes up again
and again [TYPING] as I do searches– Boxer Rebellion, bands,
underwear company, and important part of
Chinese history. So it takes a predictive eye
when looking at search results to find exactly what you need
and, ideally, to pull tips on building a better search. I see this result. It looks good. It has some extra information
that I’m going to fold back in to write a better search. So I think, at this point,
I am going to wrap up. Next time we’re going to be
looking, kind of, a lot more minutely at some of the
operators we’ve seen today, like the number range search. We’ve looked at using
an asterisk to fill in the blanks. We’ve talked about site
searching, searching within the site with a site
colon operator. There are a number more that I’d
like to introduce to you. So I hope that you will
join us next time. I hope, too, that you will take
a moment to fill out our survey and give us good,
constructive feedback on how to make these better
in the future. Among other things, our survey,
this time, is asking whether you are around this
summer and want more webinars over the summer, and in fact,
what subjects you’d like to hear about. So please do give us feedback,
and we will happily integrate it into our next webinar. Also, I have a couple of links I
would like to share with you of videos that can help explain
to your students how search engines work. And I will send that out to
the group when we have an archived version of this
up and ready to go. Thank you. Oh, sorry, what I wanted to
ask is if you have any questions, please do use your
chat box, and let us know if you have any questions. So it looks like there might
not be any questions coming in, and so you may, as always,
email me at [email protected], if you have questions
in the future. It’s a good way to get in touch
with us and ask anything that comes up later
about the session. So thank you so much
for your time. Have a lovely day. And I see that you all got
going on Twitter, which I appreciate so much. Have a lovely day. Bye bye.

Leave a Reply

Your email address will not be published. Required fields are marked *

© Copyright 2019. Amrab Angladeshi. Designed by