Searching Canberra Bird Notes
The search box labelled “Search CBN” near the top of the CBN web page is specifically set up to search Canberra Bird Notes (CBN). If you wish to search within the CBN archive you should definitely use this box rather than the general site search box that is located in the header above every page.
How to use the CBN search facility
- Go to the Canberra Bird Notes page http://canberrabirds.org.au/publications/canberra-bird-notes/
- Enter your search words including any special operators (see below) into the “Search CBN” box near the top of the page.
- Click the search icon or press Return.
- In the resulting list of possibilities returned by Google, click on a desired link to open that edition of CBN. This will open the relevant issue of CBN as a pdf file in your pdf viewer or browser (depending on how your system is set up).
- Search again within the document for the same words (or similar or simpler words). In Windows and most browsers and pdf viewers, Control-F will open a search box at the top or bottom of the window. Put words to look for in that box and click to search. Use the small up and down arrows if necessary to move to next or previous occurrences until you find the occurrence that Google identified.
Note 1: You may be asked to prove you’re not a robot at various times, be prepared to do what it asks and then “submit”.
Note 2: The results may include standard Google advertisements disguised to look like results. They can be identified by [Ad] in green beside the result link.
Note 3: Google change their operating methods regularly, so some things may change from what is written here.
Understanding Google search
To use this facility efficiently it is necessary to understand a little of how Google search works, and be able to use one or two of their ‘advanced’ search operations.
When you search as in steps 1 and 2 above, Google will search through every edition of CBN in the archive, that is, every issue ever published.
Google will return only one result for each issue of CBN in which your search is successful. So for example if you searched on [ honeyeater ] (without the square brackets), it will likely find several occurrences in each issue, but only one result will be presented — the first mentioned occurrence of ‘honeyeater’. (This likely means you’ll have one return for every issue of CBN, since every issue will probably contain the word ‘honeyeater’ somewhere).
If you search on [ scarlet honeyeater ] it will again return only one result per issue, but this time it will offer the most relevant (not the first) occurrence of “scarlet” and “honeyeater”. i.e. it will offer the first occurrence where the words appear exactly and in the same order, or if not, the first occurrence where they occur beside each other in reverse order, if not, the first occurrence where they occur close to each other, etc.
Using ‘advanced’ search operators.
Narrowing down the number of results at the first stage is important for two reasons:
- Obviously, this will reduce the number of results you have to check through.
- Not so obviously, remember that we only get one returned result for each CBN issue. If there are a large number of returns within that issue, the chances are that the single result that Google has chosen to show will not be the one we are looking for. In that case we wouldn’t know by reading the result that this issue actually does contain the desired information, because the context info given with the single result would be irrelevant. This is covered further below.
To reduce the number of results and improve the chances of getting Google to highlight and return the correct occurrence in the results, we need to: a) choose good search words and b) almost always, use one or more search ‘operators’. The most useful of the operators for our purposes are described below.
The three most useful operators to refine searches…
SEARCH FOR AN EXACT PHRASE
Using quotation marks “like this” around your words helps by restricting to that exact phrase, though be aware that then it will not find plurals or other parts of speech. You’d need to search for [ “scarlet honeyeater” ] and again for [ “scarlet honeyeaters” ] to get them all. On the other hand if you search for [ scarlet honeyeater ] without quotes, it will find honeyeater or honeyeaters but will also return documents that contain the two words anywhere within the document. For example there may be an article on Crescent Honeyeaters on page 2 and another on Scarlet Robins on page 20, which would cause a valid return for [ scarlet honeyeater ]
AND/OR
Google allows basic logical operators including AND and OR in searches. These can be used between search words, making sure that AND or OR are in capital letters. By default (in recent times) Google assumes an AND between all the words you specify, so [ word1 word2 word3 ] is the same as [ word1 AND word2 AND word3 ]. Unless you use OR, searches will only return pages that contain all of your words. For this reason [ scarlet honeyeater myzomela ] will not find any mention of the Scarlet Honeyeater unless the word ‘myzomela’ also happens to occur somewhere in the document.
You can make more complex queries using brackets together with the logical operators eg:
scarlet AND (honeyeater OR myzomela)
will find both ‘scarlet honeyeater’ and ‘scarlet myzomela’.
AROUND(X)
‘AROUND’ is the single most useful (but practically unknown) Google operator. It returns a result where the two words or phrases are within (X) words of each other. This is especially powerful in situations like ours, where the ‘pages’ that Google identifies are actually quite long documents containing many unrelated articles. For example, say you are looking for an article by Michael Lenz on Avocets.
If you searched for [ avocet “michael lenz” ] you would get many returns (46 out of 179 issues when I looked). This is because Michael is mentioned in many CBN issues in one capacity or another, and avocets are also mentioned in many issues. Furthermore, the context given by Google on those results would likely not usefully point out whether the issue contained an article by Michael, or just unrelated mentions of his name plus, elsewhere in the document, the word ‘avocet’. In fact in my test none of the 46 results contained context that indicated an article written by Michael. Even if I’d scanned through every result I’d not have been able to identify which of them contained what I was looking for.
Instead, we can look for {any mention of ‘avocet’ occurring within 10 words of “Michael lenz”} by searching on:
avocet AROUND(10) “Michael lenz”
This time you’ll get 2 returns instead of 46, and both are articles written by Michael about Avocets. Very importantly, the first mentioned result shows clearly (by reading the context) that this result contains what you are looking for. Compare the results below from the two different searches – both searches found the same correct CBN issue, but the first gives no indication that the desired article is within.
1….Search on: [ avocet “Michael lenz” ] (Result #6 of 46)
[PDF]canberra bird notes – Canberra Birds
canberrabirds.org.au/wp-content/canberra-bird-notes/CBN-42-2-final.pdf
Jul 2, 2017 – On 6 Nov 2016 Michael Lenz reported to the COG Chatline that he had observed. Whiskered Terns (Childonias hybrida) and Red-necked Avocet (Recurvirostra novaehollandiae) on Lake George. Four days later Garry Moffitt observed a Red- necked Avocet in an ephemeral marsh on the Hoskinstown Plain …
2….Search on: [ avocet AROUND(10) “Michael lenz” ] (Result #1 of 2)
[PDF]canberra bird notes – Canberra Birds
canberrabirds.org.au/wp-content/canberra-bird-notes/CBN-42-2-final.pdf
Jul 2, 2017 – AVOCET AT LAKE BATHURST. MICHAEL LENZ. 117/50 Ellenborough Street, Lyneham ACT 2602, Australia michael.lenz.birds@gmail.com. The status of the Red-necked Avocet (Recurvirostra novaehollandiae) in COG’s. Area of Interest is described as “non-breeding vagrant” (Canberra Ornithologists.
Other operators that may be occasionally useful.
| – | Put minus (-) in front of any term (including operators) to exclude that term from the results. This is not as useful as it seems. Say you are looking for an article you remember that was about Black Swans. You might search on [ black swan ] but find your results are cluttered with every Annual Bird Report (resulting in a total of 92 results). It seems possible that you can remove those by changing your search to [ “black swan” -“annual bird report” ] and indeed this reduces the results to 46 issues. However, you have also removed any issue that happens to include any reference at all to “Annual Bird Report”. Those words will occur in documents that are not themselves Annual Bird Reports but happen to mention them in passing. They may include the issue you are looking for and excluding them is clearly not what you want. The operator ‘-‘ only works when the word or phrase you are excluding will never occur in desired issues of CBN.
|
||
| * | An asterisk (*) acts as a wild-card and will match on any word. eg: [ “cog * survey” ] will return ‘cog waterbird survey’, ‘cog woodland survey’, ‘cog garden bird survey’ etc | ||
| ~ |
Include synonyms. This seems to be unreliable, and synonym inclusion is supposed to be the default now.
|
||
| .. |
|
An example to demonstrate refinement of search techniques
Assume we are searching for a review of a book on eagles by Stephen Debus.
- If we search on [ debus ] we’ll find 28 returns covering 28 issues but no idea which issue includes the desired book review.
- We may try to improve the situation by adding another search term, perhaps [ book debus ]. This doesn’t help much, probably because every issue that mentions ‘debus’ also has the word ‘book’ somewhere else in the same issue. (The correct result turns out to be 5th out of 27)
- We could instead search on [ debus “book review” ] which reduces the field to 9 results. The correct result can be picked easily by reading the returned context notes. (1st of 9 results).
- If you know more information, the search can often be narrowed by including that extra information. eg1: Add the author of the review i.e. [ “book review” debus veerman ] In this case we achieve no real improvement, though still easily identified (2nd of 8 results).
- If you knew the title of the book you hit the jackpot with only a single return using [ “Australasian Eagles and Eagle-like Birds” debus ] (1 of 1 result).
- But the most efficient search using the least information is [ debus AROUND(30) “book review”]. This gives a single result with very clear context information without needing to know anything but the book’s author. (1 of 1 result).
(A reminder – this search translates to: the word debus occurs within 30 words of the phrase “book review”).