Results 1 to 23 of 23

Thread: I need your help! Please fill in your Socionics type!

  1. #1
    24601 ClownsandEntropy's Avatar
    Join Date
    Mar 2010
    Location
    Melbourne, Australia
    TIM
    LII, 5w6
    Posts
    670
    Mentioned
    24 Post(s)
    Tagged
    0 Thread(s)

    Default I need your help! Please fill in your Socionics type!

    Hey all, I'm going to try to run some machine learning on the posts in the General Discussion board and see if there's any meaningful typing info that comes out of it.

    To do that I need a reliable source of what everyone's socionics type is.

    I've used the spreadsheet we have running as the basis, and from that I've collected the following:
    https://docs.google.com/spreadsheets...vid=1507708861
    Please look through this first sheet and, if any of the typings are incorrect, please update them.

    Also, I don't know some other people's socionics type at all. Please look for your username here and fill it in:
    https://docs.google.com/spreadsheets...#gid=942199603
    If you're reasonably confident, fill in socionics types for people who have left the forum.

    Some final comments:
    1. Please leave all types in the three-letter format, no subtype. E.g. my Socionics type would be LII, not LII-Ne.
    2. If you are uncertain of your type, please leave three question marks instead. I.e. please write ???.
    3. If you don't want your comments analysed, again please leave three question marks.

    Also, to the Admins I guess this isn't really "General Discussion" but I wanted visibility. Feel free to move ths if appropriate...
    Warm Regards,



    Clowns & Entropy

  2. #2
    ouronis's Avatar
    Join Date
    Aug 2014
    TIM
    ref to ptr to self
    Posts
    2,941
    Mentioned
    127 Post(s)
    Tagged
    2 Thread(s)

    Default

    I've wanted to do something like this but I don't have the web crawling skills. What approach do you plan to take on the machine learning side of things?

  3. #3
    Darn Socks DirectorAbbie's Avatar
    Join Date
    Aug 2008
    Location
    Southwest USA
    TIM
    LSE
    Posts
    7,123
    Mentioned
    381 Post(s)
    Tagged
    2 Thread(s)

    Default

    My type was listed as LSI, so I fixed it to LSE.

    Why do you have both Bg and Bionicgoat listed? Aren't they the same person? I listed Bionicgoat as EII. I interact with him a lot and am convinced he's my dual.
    NewBorn STAR was listed as Beta. I updated that to IEI. When I interacted with him, I figured Beta irrational. And others have assured me of IEI.
    Radio/Arsal/Antennae is still around. Why is he listed on the missing page (in triplicate)?
    Aquagraph doesn't post anymore and only visits the chatbox once in the proverbial blue moon. He's been gone longer than BG; should he be moved to the inactives page?
    Spider and limping bird are the same person. Why is she listed twice? She's also still active. She was on this morning.
    Is silke gone? I didn't hear of her leaving. Why is she listed on the missing-persons page?
    Why are Ath/zap/n0ki listed in triplicate?
    Why is SubT listed as a missing person? He's quite active.
    Anon and WorkaholicsAnon and Suz are the same person, so she's listed twice on the first sheet and once on the second.
    Uniden is listed on both sheets.
    7154413 might be polikujm.

    LSE
    1-6-2 so/sx
    Johari Nohari

    Quote Originally Posted by Ritella View Post
    Over here, we'll put up with (almost) all of your crap. You just have to use the secret phrase: "I don't value it. It's related to <insert random element here>, which is not in my quadra."
    Quote Originally Posted by Aquagraph View Post
    Abbie is so boring and rigid it's awesome instead of boring and rigid. She seems so practical and down-to-the-ground.

  4. #4
    Exodus's Avatar
    Join Date
    Jun 2006
    TIM
    LII
    Posts
    8,397
    Mentioned
    325 Post(s)
    Tagged
    0 Thread(s)

    Default

    So I guess this is for people's self-typings?

  5. #5
    Poster Nutbag The Exception's Avatar
    Join Date
    Jun 2005
    Location
    my own personal bubble
    TIM
    LII-Ne
    Posts
    4,097
    Mentioned
    103 Post(s)
    Tagged
    1 Thread(s)

    Default

    I'm listed on the second sheet but not the first one.

    My self-typing is in my signature.
    LII-Ne with strong EII tendencies, 6w7-9w1-3w4 so/sp/sx, INxP



  6. #6
    24601 ClownsandEntropy's Avatar
    Join Date
    Mar 2010
    Location
    Melbourne, Australia
    TIM
    LII, 5w6
    Posts
    670
    Mentioned
    24 Post(s)
    Tagged
    0 Thread(s)

    Default

    Thanks for your help so far!

    Here are some answers:
    Quote Originally Posted by ouronis View Post
    I've wanted to do something like this but I don't have the web crawling skills. What approach do you plan to take on the machine learning side of things?
    I'm thinking I'll start by seeing if there are any more frequently used words for introversion or extroversion, and then the same thing for each of the 16 types.
    Then I'm planning on doing some SVM using the frequency of each word - something I'll have to research myself.
    Then I'm planning to look into Word2Vec and use neural networks to see if anything meaningful comes out.
    Any other ideas?

    I've actually already done the scraping for the General Discussion Board. If you're interested I can send you the data or we could collab?
    Indeed, if anyone is reading this in the future and wants the data I can give it to you - future typers may have better machine learning knowledge than me.

    Quote Originally Posted by Director Abbie View Post
    Many questions
    The list of types is from https://docs.google.com/spreadsheets...QF0/edit#gid=0. The "missing sociotypes" are users whose sociotypes are not in that spreadsheet.
    People are listed twice because of some lazy data cleaning - I didn't want to risk accidentally dropping a former username so I just kept all duplicates.
    Thanks for pointing out some duplicates across spreadsheets. That shouldn't have happened...

    Quote Originally Posted by chips and underwear View Post
    I'm listed on the second sheet but not the first one.

    My self-typing is in my signature.
    Thanks! I'l-

    Quote Originally Posted by chips and underwear View Post
    EII - INTj - Dostoyevsky -
    Chips!!!! *shakes fist in the air*
    Warm Regards,



    Clowns & Entropy

  7. #7
    meme hotline Chae's Avatar
    Join Date
    Jul 2016
    TIM
    ethic 3
    Posts
    9,083
    Mentioned
    716 Post(s)
    Tagged
    1 Thread(s)

    Default

    Done, it sounds interesting. Keep us updated!

  8. #8
    Darn Socks DirectorAbbie's Avatar
    Join Date
    Aug 2008
    Location
    Southwest USA
    TIM
    LSE
    Posts
    7,123
    Mentioned
    381 Post(s)
    Tagged
    2 Thread(s)

    Default

    willekeurig is Agarina
    I remember CILi self-typed as IEE, so I added that even though I didn't know him well enough to be fully sure of his typing myself.
    I think Divided and DividedsGhost are the same person, but I don't know his type.
    I added hag and Antennae as self-typed EII, but they could have added themselves.
    crazedrat and rat1 are the same person.
    Jadae has been gone for years, but I know he doesn't self-type as EIE. (So I deleted that.) I'll send you a pm.
    Mediator Kam (page 2) and Director Kam (page 1) are the same person. Sometimes known as Mediator Kam, mostly known as Kamajama, only known as Director Kam for a few weeks.
    I know Limitless is EII.
    Why is Maritsa on page 2?! Why is SisOfNight? (I labeled them EII and IEI, respectively.)
    I remembered soapofsappho being nominated official IEE representative. I hope I remembered correctly. I think that's the same person as SongOfSapphire.
    I put EII as SubT's self-typing.
    I think SyrupDeGem self-typed as IEE, but I don't remember fully...oh, nvm. Gem is listed as IEE on the first page.
    Is ".starpoo {" Starfall?

    LSE
    1-6-2 so/sx
    Johari Nohari

    Quote Originally Posted by Ritella View Post
    Over here, we'll put up with (almost) all of your crap. You just have to use the secret phrase: "I don't value it. It's related to <insert random element here>, which is not in my quadra."
    Quote Originally Posted by Aquagraph View Post
    Abbie is so boring and rigid it's awesome instead of boring and rigid. She seems so practical and down-to-the-ground.

  9. #9
    Humanist Beautiful sky's Avatar
    Join Date
    Jan 2009
    Location
    EII land
    TIM
    EII INFj
    Posts
    25,960
    Mentioned
    669 Post(s)
    Tagged
    6 Thread(s)

    Default

    I type by VI and if you want those types then go ahead and use my typing they are highly consistent with the method I use so if X feature is X type then that is the person's type despite their wanting to be a flower or whatever else emotionally
    -
    Dual type (as per tcaudilllg)
    Enneagram 2w1sw(1w9) helps others to live up to their own standards of what a good person is and is very behind the scenes in the process.
    Tritype 1-2-6 stacking sp/sx


    I'm constantly looking to align the real with the ideal.I've been more oriented toward being overly idealistic by expecting the real to match the ideal. My thinking side is dominent. The result is that sometimes I can be overly impersonal or self-centered in my approach, not being understanding of others in the process and simply thinking "you should do this" or "everyone should follor this rule"..."regardless of how they feel or where they're coming from"which just isn't a good attitude to have. It is a way, though, to give oneself an artificial sense of self-justification. LSE

    Best description of functions:
    http://socionicsstudy.blogspot.com/2...functions.html

  10. #10
    Arete GuavaDrunk's Avatar
    Join Date
    Apr 2011
    Location
    Now in stores near you.
    TIM
    IEI-Fe (9)62 sx/?
    Posts
    1,586
    Mentioned
    108 Post(s)
    Tagged
    1 Thread(s)

    Default

    Updated myself as EIE on first spreadsheet.

    Machine learning: Using neural networks and other such methods very strongly depends on having good positive/negative cases to tune the machine with. I suspect that grammatical constructions will turn out more useful than simple word use; there might be forums where people already use neural networks for such things, and have pre-made algorithms/calibration databases for grammar. Linguistics forums maybe?

    Potentially useful, though afaik it was calibrated for Russian: http://www.tryukraine.com/socionics/...emantics.shtml

    In any case, interesting project
    Reason is a whore.

  11. #11
    Tigerfadder's Avatar
    Join Date
    Jun 2016
    TIM
    ILE
    Posts
    1,305
    Mentioned
    31 Post(s)
    Tagged
    0 Thread(s)

    Default

    but you do not have to apply logic to it. Just run it with big samples and maybe some programming to find corrections in high amount of usage in certain words that is distinguishable for the specific type only. Maybe EIE use "machine" and "whore" more often then the norm for example.

  12. #12
    24601 ClownsandEntropy's Avatar
    Join Date
    Mar 2010
    Location
    Melbourne, Australia
    TIM
    LII, 5w6
    Posts
    670
    Mentioned
    24 Post(s)
    Tagged
    0 Thread(s)

    Default

    Thanks heaps to people so far. I'm gonna give this a final bump before I start using these .
    Warm Regards,



    Clowns & Entropy

  13. #13
    Infinity Persephone's Avatar
    Join Date
    Apr 2014
    Location
    The country of croissants
    Posts
    1,840
    Mentioned
    178 Post(s)
    Tagged
    5 Thread(s)

    Default

    I am featuring there in a triplet form, it feels uncomfortable...


  14. #14

  15. #15
    24601 ClownsandEntropy's Avatar
    Join Date
    Mar 2010
    Location
    Melbourne, Australia
    TIM
    LII, 5w6
    Posts
    670
    Mentioned
    24 Post(s)
    Tagged
    0 Thread(s)

    Default

    Quote Originally Posted by Subteigh View Post
    @ClownsandEntropy I might be able to help fill in any remaining blanks by browsing profiles (signaturesetc.) for self-typings, although having others would mean it would actually get done.
    That is an interesting idea, we could have a divide and conquer approach.
    Warm Regards,



    Clowns & Entropy

  16. #16
    Enlightened Hedonist
    Join Date
    Nov 2005
    Posts
    18,374
    Mentioned
    449 Post(s)
    Tagged
    4 Thread(s)

    Default

    Quote Originally Posted by ClownsandEntropy View Post
    That is an interesting idea, we could have a divide and conquer approach.
    often, it is necessary to go to a forum post to see if a member's type is in their signature or under their avatar space, because the forum software does not show the data by going to the profile (generally, with members who have been inactive for many years): seeing a list of a member's posts counts as a search, which is limited to once every 60 seconds, which makes it especially time-consuming for one or two people to do!

  17. #17
    bye now
    Join Date
    Feb 2012
    Posts
    1,888
    Mentioned
    36 Post(s)
    Tagged
    0 Thread(s)

    Default

    Wouldn't this just end up creating biases for each type, accurate or not, depending on what self-type people choose and how they act on the forum? (which may be different than real life)

    Yes, I know, I'm metaphorically shitting on this right now. Oh well. It's first thing that comes to mind.

  18. #18
    Enlightened Hedonist
    Join Date
    Nov 2005
    Posts
    18,374
    Mentioned
    449 Post(s)
    Tagged
    4 Thread(s)

    Default

    I would think it would have to be weighted to prevent some users having a disproportionate effect, and if possible, it would exclude quoted sections in posts (although if not possible, the inclusion of quotes may represent something significant nonetheless).

  19. #19
    24601 ClownsandEntropy's Avatar
    Join Date
    Mar 2010
    Location
    Melbourne, Australia
    TIM
    LII, 5w6
    Posts
    670
    Mentioned
    24 Post(s)
    Tagged
    0 Thread(s)

    Default

    Quote Originally Posted by strangeling View Post
    Wouldn't this just end up creating biases for each type, accurate or not, depending on what self-type people choose and how they act on the forum? (which may be different than real life)
    Yeah, if this actually works then the tool can only say "most people who talk like you are tend to self-type as X".

    Quote Originally Posted by Subteigh View Post
    I would think it would have to be weighted to prevent some users having a disproportionate effect, and if possible, it would exclude quoted sections in posts (although if not possible, the inclusion of quotes may represent something significant nonetheless).
    Quotes should already be excluded.
    Warm Regards,



    Clowns & Entropy

  20. #20
    Exodus's Avatar
    Join Date
    Jun 2006
    TIM
    LII
    Posts
    8,397
    Mentioned
    325 Post(s)
    Tagged
    0 Thread(s)

    Default

    Could this be adapted to use someone's typing list? It might actually provide support for someone's typings (as opposed to user self-typings, which may be inaccurate), if there was a greater vocabulary correlation.

  21. #21
    ouronis's Avatar
    Join Date
    Aug 2014
    TIM
    ref to ptr to self
    Posts
    2,941
    Mentioned
    127 Post(s)
    Tagged
    2 Thread(s)

    Default

    My preliminary studies reveal that ILIs are most likely to say "I hate everything," IEEs are most likely to use the word "love" or "hug" and ESEs are most likely to say "lol".

    A stronger feature set than simple word frequencies with stopwords removed is needed, for sure.

    Looking for certain grammatical constructions predicted by reinins might be good, but that also sounds like a lot of work.

  22. #22
    Poster Nutbag The Exception's Avatar
    Join Date
    Jun 2005
    Location
    my own personal bubble
    TIM
    LII-Ne
    Posts
    4,097
    Mentioned
    103 Post(s)
    Tagged
    1 Thread(s)

    Default

    Quote Originally Posted by ouronis View Post
    My preliminary studies reveal that ILIs are most likely to say "I hate everything," IEEs are most likely to use the word "love" or "hug" and ESEs are most likely to say "lol".
    I say LOL alot. I also use alot of emoticons. I can't be a real LII.
    LII-Ne with strong EII tendencies, 6w7-9w1-3w4 so/sp/sx, INxP



  23. #23
    ouronis's Avatar
    Join Date
    Aug 2014
    TIM
    ref to ptr to self
    Posts
    2,941
    Mentioned
    127 Post(s)
    Tagged
    2 Thread(s)

    Default

    Some charts I did for this back in 2016.
    https://imgur.com/a/6cyZBo7

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •