Learn Japanese with JapanesePod101.com

View topic - Question regarding rikaichan returns

Question regarding rikaichan returns

Have a textbook or grammar book that you find particularly helpful? What about a learning tip to share with others?

Question regarding rikaichan returns

Postby Snowflake » Wed 09.21.2011 9:08 pm

Rikaichan is a very helpful when it gives only one answer to a highlighted word. It becomes confusing to me when it presents multiple answers. How do I know which is correct? It is particularly confusing when none of the answers relate to any of the other ones. It's very discouraging :cry: .
Snowflake
 
Posts: 201
Joined: Thu 01.24.2008 7:04 pm
Location: New England, USA
Native language: English

Re: Question regarding rikaichan returns

Postby furrykef » Wed 09.21.2011 9:50 pm

You pretty much have to go on context as well as what you already know about Japanese.
Founder of Learning Languages Through Video Games.
Also see my lang-8 journal, where you can help me practice Japanese (and Spanish, and Italian!)
User avatar
furrykef
 
Posts: 1572
Joined: Thu 01.10.2008 9:20 pm
Native language: Eggo (ワッフル語の方言)
Gender: Male

Re: Question regarding rikaichan returns

Postby Snowflake » Wed 09.21.2011 10:32 pm

Thank you, Furrykef. I find that even more discouraging, but I do appreciate the reply.
Snowflake
 
Posts: 201
Joined: Thu 01.24.2008 7:04 pm
Location: New England, USA
Native language: English

Re: Question regarding rikaichan returns

Postby furrykef » Thu 09.22.2011 6:56 am

Well, if you're looking for encouragement, I'll tell you this: I rarely find it a problem. Sometimes I wind up with a large number of entries, none of which seem to fit the context, but it happens infrequently enough that it's no trouble to just ask somebody. (And there are lots of places to ask: here; WordReference, ##japanese on freenode; and my current choice for when ##japanese is being unhelpful: japanese.stackexchange.com...)

So if you're having trouble now, you probably won't have so much as you advance. After all, I'm still only JLPT N3. (I haven't taken the test, but it's very obvious to me that this is my skill level from the sample questions for the N3 and N2 tests.)

Oh yeah, one tip -- do be sure you have your cursor on the first character of the word or phrase. Otherwise you might be looking the wrong thing up.
Last edited by furrykef on Fri 09.23.2011 1:18 pm, edited 1 time in total.
Founder of Learning Languages Through Video Games.
Also see my lang-8 journal, where you can help me practice Japanese (and Spanish, and Italian!)
User avatar
furrykef
 
Posts: 1572
Joined: Thu 01.10.2008 9:20 pm
Native language: Eggo (ワッフル語の方言)
Gender: Male

Re: Question regarding rikaichan returns

Postby jimbreen » Thu 09.22.2011 7:47 am

furrykef wrote:Well, if you're looking for encouragement, I'll tell you this: I rarely find it a problem. Sometimes I wind up with a large number of entries, none of which seem to fit the context, but it happens infrequently enough that it's no trouble to just ask somebody.


If rikaichan gives a heap of possibilities, you have possibly hit on a case where the word/phrase is not the JMdict/EDICT dictionary, which "only" has 160,000 entries (195,000 counting variants.) In that case, once you work out what the word is, please add it to the dictionary.
(http://www.edrdg.org/jmdictdb/cgi-bin/e ... jmdict&c=1)

Jim
User avatar
jimbreen
 
Posts: 165
Joined: Tue 06.27.2006 2:09 am
Location: Melbourne, Australia

Re: Question regarding rikaichan returns

Postby Snowflake » Thu 09.22.2011 11:55 am

Again, thank you both. Much appreciated.

I have never seen a step-by-step explanation of the Rikaichan readouts for whole words. I've seen an explanation for kanji but not for words. I'll use the sentence below (found elsewhere on TJP) to illustrate.

図書館には本がどのくらいありますか

For 本, Rikaichan gives two definitions: "origin" and "book". In this case, since I know 図書館 is "library", I can make an educated guess that, in this sentence, 本 is book(s). Why wouldn't "book" be the first option? Moving on to the next "word" (が), there are 6 options, the first of which is "moth" and is shown along with a kanji. "Moth" doesn't seem to fit with "library" and "book", and, if I continue reading the Rikaichan return, I see the second option is the standalone kana が, exactly as it is in the sentence. Given what I know of Japanese (which isn't as much as I'd like, obviously), I assume the second option is my best bet.

The next word, どのくらい, is more troublesome. It appears that there are two words with kanji which are pronounced どのくらい. Given that the author of the sentence wrote the word in kana, how would I know that the definition is correct? Is there another word which is pronounced どのくらい and is typically written in kana only?

And finally, the readout for the last word か shows か as the bottom option. I'm confused as to why it wouldn't be the first option. Isn't question marker the most common use of a standalone か?

I'm sorry to ask so many questions but understanding how Rikaichan results are organized has been troubling me for at least 2 years. Try as I might, I have never found an answer. I really like Rikaichan but I think I could get much more out of it if I understood it better. Am I the only person who doesn't understand? Or is there a walkthrough or diagram somewhere that I just haven't found?

I thank you for bearing with me through this.
Snowflake
 
Posts: 201
Joined: Thu 01.24.2008 7:04 pm
Location: New England, USA
Native language: English

Re: Question regarding rikaichan returns

Postby Hyperworm » Thu 09.22.2011 2:28 pm

Snowflake wrote:And finally, the readout for the last word か shows か as the bottom option. I'm confused as to why it wouldn't be the first option. Isn't question marker the most common use of a standalone か?
This is strange. I'm using Chrome at the minute and so I'm using Rikaikun instead of Rikaichan, but what I get is a big list of kanji words read か, with the hiragana one at the top. Since I've actually highlighted the hiragana か and not a kanji, this is unsurprising. If Rikaichan and Rikaikun differ in implementation in this regard, that would surprise me - I don't see why Rikaichan wouldn't be able to put it in the proper order, unless they want you to read the list bottom-up.

Snowflake wrote:For 本, Rikaichan gives two definitions: "origin" and "book". In this case, since I know 図書館 is "library", I can make an educated guess that, in this sentence, 本 is book(s). Why wouldn't "book" be the first option?
Picking out the correct word definition would require performing semantic analysis (meaning analysis) of the sentence. This is a very difficult area of computing which even commercial machine translation programs have immense problems with (and are in fact abandoning in favor of a statistics-based approach to translation). Rikaichan doesn't have such a feature; it just performs word segmentation (even that is a little tricky) and leaves you to work out the rest.

Snowflake wrote:Moving on to the next "word" (が), there are 6 options, the first of which is "moth" and is shown along with a kanji. "Moth" doesn't seem to fit with "library" and "book", and, if I continue reading the Rikaichan return, I see the second option is the standalone kana が, exactly as it is in the sentence. Given what I know of Japanese (which isn't as much as I'd like, obviously), I assume the second option is my best bet.
Again, this is a little odd - Rikaikun lists subject particle が first for me.
On the other hand, I don't see why a user would even highlight が in this sentence, or か as mentioned above. Rikaichan isn't a tool for enabling people who don't know Japanese to read Japanese by simply mousing over the words in sequence - it's a tool for making it easier and faster to read Japanese when you have a smaller vocabulary. You still have to know/learn the grammar, and you may still need to puzzle over a sentence's meaning for a bit until you're familiar enough with the language that all you need to know is the meaning of the particular nouns and verbs.
With が in hiragana following a noun like that, it should be fairly obvious (to the user, not to the software) that it is the subject particle.

Snowflake wrote:The next word, どのくらい, is more troublesome. It appears that there are two words with kanji which are pronounced どのくらい. Given that the author of the sentence wrote the word in kana, how would I know that the definition is correct? Is there another word which is pronounced どのくらい and is typically written in kana only?
What I get is:
Rikaikun wrote:どの位 どのくらい
何の位 どのくらい
(n,uk) how long; how far; how much; (P)
The uk here signifies usually kana. uK is "usually Kanji". So from that, the kana appearance shouldn't surprise you.
The list of codes used by EDICT/Rikaichan can be found here.
Also, since this definition points towards どのくらい having a fundamental grammatical meaning rather than being an ordinary noun/verb/adjective, if you don't recognize it, you should look it up in a proper grammar resource rather than relying on Rikaichan's definition to understand it. Then you should find you're able to interpret it well in context and find out if it fits or not.
Also, since it's the longest match available (the rest of the entries only match どの alone), that's a third sign that it's probably right (this can fail though).

...

Incidentally, Rikaikun did fail me in one area with regard to entry ordering: hovering 図書館 gives me
Rikaikun wrote:図書館 ずしょかん
(n,ok) library; (P)
図書館 としょかん
(n) library; (P)
even though "ok" means obsolete kana usage, and therefore としょかん is the more common reading and should be listed first.
fun translation snippets | need something translated?
BTC@1KMZXgoWiDshQis5Z7feCx8jaiP4QAB2ks
User avatar
Hyperworm
 
Posts: 493
Joined: Tue 11.20.2007 2:26 pm
Native language: English
Gender: Male

Re: Question regarding rikaichan returns

Postby Snowflake » Thu 09.22.2011 4:33 pm

Hyperworm wrote: (...)
Snowflake wrote:For 本, Rikaichan gives two definitions: "origin" and "book". In this case, since I know 図書館 is "library", I can make an educated guess that, in this sentence, 本 is book(s). Why wouldn't "book" be the first option?
Picking out the correct word definition would require performing semantic analysis (meaning analysis) of the sentence. This is a very difficult area of computing which even commercial machine translation programs have immense problems with (and are in fact abandoning in favor of a statistics-based approach to translation). Rikaichan doesn't have such a feature; it just performs word segmentation (even that is a little tricky) and leaves you to work out the rest.


Then, it sounds like when I'm shown several definitions, I'm pretty much on the right track by asking myself "given what I already know about the sentence, which seems to be the best fit".


Hyperworm wrote:
Snowflake wrote:Moving on to the next "word" (が)...


My primary reason for including the が example was to ask why the particle explanation wasn't the first option. I agree with you -- normally I shouldn't Rikaichan a が. For the purposes of the post, however, I was "next-wording" through the entire sentence to see what Rikaichan would return. The return for が was perplexing.


Hyperworm wrote:What I get is:
Rikaikun wrote:どの位 どのくらい
何の位 どのくらい
(n,uk) how long; how far; how much; (P)
The uk here signifies usually kana. uK is "usually Kanji". So from that, the kana appearance shouldn't surprise you.
The list of codes used by EDICT/Rikaichan can be found here.


That list is definitely a piece of the puzzle I was missing. Hehe, I guess you could say it was my missing link! (/groan ;)) Thank you ever so much; it's now bookmarked!

Hyperworm wrote:Also, since it's the longest match available (the rest of the entries only match どの alone), that's a third sign that it's probably right (this can fail though).


Yes, I was thinking that might be the case. Thank you for the confirmation.


Hyperworm wrote:Incidentally, Rikaikun did fail me in one area with regard to entry ordering: hovering 図書館 gives me
Rikaikun wrote:図書館 ずしょかん
(n,ok) library; (P)
図書館 としょかん
(n) library; (P)
even though "ok" means obsolete kana usage, and therefore としょかん is the more common reading and should be listed first.


And interestingly enough, Rikaichan, lists them in the opposite order. としょかん is first and ずしょかん is second.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

I can't tell you, Hyperworm, Jim Breen and Furrykef, how much I appreciate your patience and answers. Thank you again!
Snowflake
 
Posts: 201
Joined: Thu 01.24.2008 7:04 pm
Location: New England, USA
Native language: English

Re: Question regarding rikaichan returns

Postby phreadom » Thu 09.22.2011 6:50 pm

furrykef wrote:(And there are lots of places to ask: here; WordReference, #japanese on freenode; and my current choice for when #japanese is being unhelpful: japanese.stackexchange.com...)


You mean ##Japanese. (or even #nihongo, but apparently that's not as active)

#Japanese is invite only.
猿も木から落ちる
User avatar
phreadom
Site Admin
 
Posts: 1762
Joined: Sun 01.29.2006 8:43 pm
Location: Michigan, USA
Native language: U.S. English (米語)
Gender: Male

Re: Question regarding rikaichan returns

Postby jimbreen » Thu 09.22.2011 7:57 pm

furrykef wrote:You pretty much have to go on context as well as what you already know about Japanese.


Absolutely. Word-by-word or worse, kanji-by-kanji, translation is a horrible waste of time.

I don't use rikaichan, but I think it does some very silly things. For example, consider the simple passage "静かでした". When you hover over the 静か it gives the correct reading & meaning. But when you move the pointer along to the か it tells you all sorts of things about か (deer, mosquitoes, etc.) This is both wrong and silly in the context of that passage. Having correctly identified 静か it should have skipped on. Similarly, it makes a hash of でした, even offering 下 and 舌 for the した. This is not an aide to reading; it is a source of confusion unless you know what it's all about.

In WWWJDIC's text glossing function, which I know is often clunky and occasionally inaccurate, I go to the other extreme and avoid giving translations unless I'm fairly sure of them. For 静かでした it simply gives:
* 静か 静か(P); 閑か 【しずか】 (adj-na) quiet; peaceful; (P); ED
* でした (v) was ; KD

Jim
User avatar
jimbreen
 
Posts: 165
Joined: Tue 06.27.2006 2:09 am
Location: Melbourne, Australia

Re: Question regarding rikaichan returns

Postby furrykef » Fri 09.23.2011 1:20 pm

jimbreen wrote:If rikaichan gives a heap of possibilities, you have possibly hit on a case where the word/phrase is not the JMdict/EDICT dictionary, which "only" has 160,000 entries (195,000 counting variants.) In that case, once you work out what the word is, please add it to the dictionary.

I think usually when this happens to me it's when a word is written in kana and there's some absurd number of homonyms for it. It happens to me every now and then because I deal with a lot of all-kana text (specifically from video games). So it isn't necessarily a problem with EDICT. Take せんせい for instance. Of course this will almost always mean "teacher", but on one of those rare occasions it might not, you have to work out which of the six definitions it is. If you understand most of the rest of the passage, this should be trivial, but if not, you could be in trouble.

This sort of problem can also arise when I think I have the gist of the passage but in fact I have it wrong, so I'm trying to fit the words into my screwed-up context and none of them will fit. (Corollary: If none of the words seem to fit your context, question your assumptions about the context.)

Of course, neither EDICT nor Rikaichan could sort this sort of thing out for you.

jimbreen wrote:This is both wrong and silly in the context of that passage. Having correctly identified 静か it should have skipped on.

I actually find this helpful. This way you can easily deal with the occasions where the match is a false positive and it matches a longer string than it should. With 静か there is unlikely to be a false positive because 静 is not an independent word, but there are other cases where it can happen. It's admittedly contrived, but imagine 青いろ. If Rikaichan did it your way, it'd match あおい and then there'd be a hanging ろ, and you'd have no way of seeing that いろ = color because it thinks the い goes with 青い. (Of course, if Rikaichan were really smart, it could find 青色 and provide the definition for that... but that'd be asking a bit much.)

There are of course cases much less contrived than that, I'm just too sleepy to think of one right now. ;)
Founder of Learning Languages Through Video Games.
Also see my lang-8 journal, where you can help me practice Japanese (and Spanish, and Italian!)
User avatar
furrykef
 
Posts: 1572
Joined: Thu 01.10.2008 9:20 pm
Native language: Eggo (ワッフル語の方言)
Gender: Male


Return to Learning Materials Reviews & Language Learning tips

Who is online

Users browsing this forum: No registered users and 9 guests