ordering search results by relevance

I just registered pleco dict for the iphone, and am comparing it with dian hua. I have downloaded the cc-cedict, but can't figure out how to order results by relevance. For instance, if I search for "acid", there are nearly 100 results, with items like "aminobenzoic acid", "amino acid", "saturated fatty acid", etc. showing up before plain old "acid". In dian hua, the first entry is a match on the full word "acid" rather than words that contain "acid", so the first result is "酸", but I can then look further down the list if necessary. How can I get pleco dict to order results so that full matches appear before matches that contain the input as a part of the word. Another example would be searching for "rain". I would expect the first result to be 雨, but the first entry is instead "rain on Mt Ba (idiom)", followed by "shower(rain)" and then "rainstorm" and "seek shelter from wind and rain(idiom)" with plain old "rain" or 雨 WAYYYY down on the list.
 

Eggwind

举人
andrewstrauss said:
I just registered pleco dict for the iphone, and am comparing it with dian hua. I have downloaded the cc-cedict, but can't figure out how to order results by relevance. For instance, if I search for "acid", there are nearly 100 results, with items like "aminobenzoic acid", "amino acid", "saturated fatty acid", etc. showing up before plain old "acid". In dian hua, the first entry is a match on the full word "acid" rather than words that contain "acid", so the first result is "酸", but I can then look further down the list if necessary. How can I get pleco dict to order results so that full matches appear before matches that contain the input as a part of the word. Another example would be searching for "rain". I would expect the first result to be 雨, but the first entry is instead "rain on Mt Ba (idiom)", followed by "shower(rain)" and then "rainstorm" and "seek shelter from wind and rain(idiom)" with plain old "rain" or 雨 WAYYYY down on the list.
Well, I think the problem is that you don't have an English-Chinese dictionary (CC-CEDICT is Chinese-English), so Pleco is defaulting to a full-text search, which comes with downsides such as what you write here. You can download the free LDC, or you can buy a dictionary. I've got all but two of the dictionaries offered by Pleco and would seriously recommend the commercial dictionaries over the free ones.
 
Eggwind said:
andrewstrauss said:
I just registered pleco dict for the iphone, and am comparing it with dian hua. I have downloaded the cc-cedict, but can't figure out how to order results by relevance. For instance, if I search for "acid", there are nearly 100 results, with items like "aminobenzoic acid", "amino acid", "saturated fatty acid", etc. showing up before plain old "acid". In dian hua, the first entry is a match on the full word "acid" rather than words that contain "acid", so the first result is "酸", but I can then look further down the list if necessary. How can I get pleco dict to order results so that full matches appear before matches that contain the input as a part of the word. Another example would be searching for "rain". I would expect the first result to be 雨, but the first entry is instead "rain on Mt Ba (idiom)", followed by "shower(rain)" and then "rainstorm" and "seek shelter from wind and rain(idiom)" with plain old "rain" or 雨 WAYYYY down on the list.
Well, I think the problem is that you don't have an English-Chinese dictionary (CC-CEDICT is Chinese-English), so Pleco is defaulting to a full-text search, which comes with downsides such as what you write here. You can download the free LDC, or you can buy a dictionary. I've got all but two of the dictionaries offered by Pleco and would seriously recommend the commercial dictionaries over the free ones.
My understanding was that dian hua used this same database, or at least that is what michael love said in a comment on this post:
http://www.sinosplice.com/life/archives/2009/12/19/pleco-for-iphone-is-out#comment-511344
If dian hua is able to display the results from this dictionary by relevance, why can't pleco?
 

mikelove

皇帝
Staff member
Relevance sort should be added soon, hopefully in the next minor / bug-fix update.

The data's certainly there and it's easy enough for us to use it to sort search results, but basically the problem is that we decided very late in development to make Pleco available as a free app. Before then, we were planning to charge something for the basic version of Pleco and include both a C-E and an E-C dictionary in that version, which meant there was no need to optimize our full-text search feature around people using it to look up Chinese translations of English words; it was instead designed more to look up all of the vocabulary relevant to a particular subject (wedding toasts, say), or to look up very uncommon words that wouldn't show up in more than a couple of entries (words that wouldn't always show up in the E-C dictionary, in other words).

So we've been steadily improving E-C fulltext search since then, fixed a bunch of bugs in it in the 2.0.5 update and should hopefully be adding relevance sorting in 2.0.6. In the meantime, though, the LDC database is free and might be a better option anyway since it saves you the trouble of scrolling through the results; you just look up a word, tap on it to view its translations, and then can tap on any of those to bring up the corresponding C-E entry in a popup window with Pinyin / detailed definition / etc.
 
mikelove said:
Relevance sort should be added soon, hopefully in the next minor / bug-fix update.

The data's certainly there and it's easy enough for us to use it to sort search results, but basically the problem is that we decided very late in development to make Pleco available as a free app. Before then, we were planning to charge something for the basic version of Pleco and include both a C-E and an E-C dictionary in that version, which meant there was no need to optimize our full-text search feature around people using it to look up Chinese translations of English words; it was instead designed more to look up all of the vocabulary relevant to a particular subject (wedding toasts, say), or to look up very uncommon words that wouldn't show up in more than a couple of entries (words that wouldn't always show up in the E-C dictionary, in other words).

So we've been steadily improving E-C fulltext search since then, fixed a bunch of bugs in it in the 2.0.5 update and should hopefully be adding relevance sorting in 2.0.6. In the meantime, though, the LDC database is free and might be a better option anyway since it saves you the trouble of scrolling through the results; you just look up a word, tap on it to view its translations, and then can tap on any of those to bring up the corresponding C-E entry in a popup window with Pinyin / detailed definition / etc.

Good to know.....thanks for the info.....I will keep an eye out for 2.0.6, and download the LDC in the meantime to play around
 
Top