Which wordlist endpoints would be useful for you?

2»

Comments

  • @AmosDuveen

    Hi Amos. Was wondering if there is any news on a WordList endpoint that includes all forms of words (not just base forms)? Is there a rough ETA? Cheers.

  • AmosDuveenAmosDuveen Member, Administrator, Moderator admin

    Hi @lkneebone489,

    Nothing to announce yet; there are a few other developments happening first before we get to this.

  • flurbiusflurbius Member

    Hi
    Im working on software that will help EFL teachers prepare class material. Something that would be useful to me would be an endpoint that provided a bridge between the sense/subsense id provided by the entries endpoint and the ids provided by the thesaurus endpoint. I would use this to link the definitions from entries to the synonyms and antonyms from the thesaurus.

    eg if I hit it with a value like id=m_en_gbus0422850.057 it returned something like
    [
    ...
    entries {
    id: m_en_gbus0422850.057
    },
    thesaurus {
    id: "id2d60209a-b30e-4af2-89e4-9d4c01b07f0d"
    }
    ...
    ]

    I did see another thread where the reason for this deficiency was explained, but it looks as though there is still no solution. I do realise that this is not a trivial problem, just thought I would chime in to let you know that there is at least one other person who would like to see this solved.

  • SimoneSimone Administrator admin

    Hi @flurbius
    Many thanks for your suggestion - feedback like this, especially when you detail an example of what you need, is very helpful for us to inform future decisions about the API (although I can't promise that this particular endpoint will be included).
    And it always helps to know more than one person is interested in the same thing.
    I'm passing your suggestion to the development team.

    You mentioned your project is to help EFL teachers, so I wanted to take the chance to mention that we have a community for English learners as well - in case you are a teacher yourself or know teachers for whom it may be useful: https://forum.oxforddictionaries.com/en

  • AmosDuveenAmosDuveen Member, Administrator, Moderator admin

    Hi @flurbius,

    In theory, this should already be possible; we have the data to hand already but there are a few complications.

    Firstly, these IDs have never been searchable and I don't think that is going to be fixed any time soon; the only searchable element is the headword and I think that is a limitation of the system we have to work with. I would love to be able to search by ID, but that's just me being a data geek; accumulating real world scenarios will help us to build a business case for making such changes, if we can.

    Secondly, you'd think it ought to be possible to make two calls: one to the dictionary, one to the thesaurus and compare the IDs within the two outputs to match the senses. However, since the dictionary data undergoes significant conversion work, including editorial changes regarding sense/homograph boundaries, I don't think this would be as easy as we would like to assume.

    The release of the 2018 data later this year should bring the entry-level parts of the ID (before the point) back into line but the sense-level part (after the point) may still not match up due to the various conversions that happen to the dictionary data to optimize it for digital use (it would be lovely if they did match, but I won't be holding my breath).

    A more likely solution would be to surface corresponding IDs. Like I said at the start of this reply, we do have the data available to us, however, finding a way to fit it into the data model is quite a complex task and would then require us to reconvert the data. This is a very big job that would have to be balanced against many other competing priorities.

  • GarryWoodGarryWood Member

    Hello,

    Just wanted to add a vote in favor of a 'is this word valid?' endpoint, if it was kept simple and fast to respond it could be very useful.

    Thanks.

  • SimoneSimone Administrator admin

    Hi @GarryWood
    Noted!
    Thanks for letting us know what would be useful for you.

  • lkneebone489lkneebone489 Member

    @Simone
    Hi Simone. I am still wondering about a new "WordList" endpoint that includes all forms of words, not just base-forms. Is this in the works or planned? Cheers.

  • SimoneSimone Administrator admin

    Hi @lkneebone489
    I'm not sure there are any updates on this, but let me ask about the latest plans - I'll get someone to reply to you, bear with me!

  • AmosDuveenAmosDuveen Member, Administrator, Moderator admin
    edited July 25

    We have been looking at wordlist options and have some proposals in early stage development right now; I won't say too much more here but the wheels are in motion.

  • lkneebone489lkneebone489 Member
    edited July 25

    Thanks for letting me know.

    The features I am personally looking for are:
    -All forms of words, not just base-forms
    -The option to fetch "GB" or "US" languages
    -The exclusion options that are in the current WordList endpoint. Ie, exclude_senses=grammaticalFeatures=proper,abbreviation (also excluding all forms of these words, not just base-form. Exclude "Acronyms" aswell)
    -Very rare/uncommon words are not strictly required

  • shahoodshahood Member

    Hi,

    Prefix should be an option that can be turned on or off and we should be able to give a search string.
    The objective is to retrieve word list of all words/phrases that start from or contain the search string. Currently we can get only those that 'start from' the prefix string.

  • SimoneSimone Administrator admin

    Your suggestion was noted, @shahood (I can't promise it will be implemented, but it will be considered).
    Thanks for your feedback!

2»
Sign In or Register to comment.