Conversation log 2019-07-27-2

• 1> Let’s talk.
• 0> What would you like to see happen in this conversation?
• 1> I’ve replaced the selection part of the questionnaire with machine learning and I’ve made sure to save the logs in an easy to read format and I’d like to test it.
• 0> Is there anything else about that question statement?
• 1> When the questionnaire was created 5 years ago, it was “the one with the highest score is randomly selected with high probability”. In the process of porting it to Python3 and connecting it to Mattermost this time, once the ”
• 0> How do you know that question statement?
• 1> (mistake, I sent it in the middle of a sentence)
• 1> continued, once I changed it to “in order of highest score without randomness”. However, there was still no mechanism to change the score for each question text. Now, machine learning has enabled us to obtain an “appropriateness” score for each combination of question text and keywords.
• 0> What do you value?
• 1> Asking the right questions at the right time encourages verbalization, and that verbalization is valuable.
• 0> Is there anything else about that question?
• 1> Ask questions in a straightforward, logical manner, distributing emotions and such. It’s like a mirror.
• 0> What type of question is that?
• 1> For keywords in sentences used by humans, this is done to make them more specific.
• 0> What type of keywords are those?
• 1> Very general meaning.
• 0>

  • continued, once we changed it to “in order of highest score without randomness”. However, there was still no mechanism to change the score for each question text. Now, machine learning has made it possible to obtain an “appropriateness” score for each combination of question text and keywords.

  • What type of questionnaire is this?
• 1> We are using a set of “questions with as little interpretation by the questioner as possible,” which is used in a technique called clean language. This project started with the idea that it is very important in this method that the questioner does not include his/her own interpretation, and if that is the case, it would be better to have a program that does not have his/her own interpretation to begin with.
• 0>

  • continued, once we changed it to “in order of highest score without randomness”. However, there was still no mechanism to change the score for each question text. Now, machine learning has made it possible to obtain an “appropriateness” score for each combination of question text and keywords.

  • What does this question statement seem to be?
• 1> fish hook. Something that hooks and catches something in the speaker that has not yet been verbalized. Not valuable in itself, but valuable in the fish caught.
• 0> How do you know its verbalization?
• 1> (That’s the thing.)

(After a while)

• 1> Currently, there is no way to detect that a verbalization has occurred, so the conversation with this program continues on the sly with no end in sight. In fact, when the speaker is able to verbalize something new, we should switch to the phase of proceeding toward the end.
• 0> Is there anything else you can tell us about that verbalization?
• 1> When good verbalization occurs, the speaker may express emotions such as “Oh, I see! etc., so hopefully we can pick up on that.
• 0> Is there anything else you can tell us about that speaker?
• 1> It might be easier to do if it were possible to get something other than a string of text, such as the momentum of keyboard tapping, facial expressions, or heart rate, but it would lose the ease of use. It would be rather nice to be able to use it on the move with a smartphone.
• 0> What is the relationship between the question statement and the verbalization?
• 1> The questionnaire stimulates the user to think about things from a new angle, which leads to the verbalization of things that have not been verbalized before.
• 0> Where is that question text?
• 1> When the user tries to type, just above the
• 0> What type of user is this user?
• 1> Users who see the value in making the language
• 0> What type of verbalization is that?
• 1> Hmmm, well, I suppose that what the user already knows from previous experience, but is unable to express in words, or is not even aware that he/she knows, becomes an object of linguistic thought when expressed in words, and can be communicated to others, observed by oneself, or critiqued. and can be communicated to others, observed by oneself, and criticized.
• 0> Is there anything else about that user?
• 1> Assumed users are those who find value in the communication and observation of such thoughts to others, and in their own criticism of their own thoughts.
• 0> Is a questionnaire the same as a verbalization? Are they different?
• 1> Question text is blank, and the language is to fill in the blanks.
• 0> What happens to the verbalization at the time of that question statement?
• 1> NG
• 0> What happens to the question statement at the time of its verbalization?
• 1> NG
• 0> What is the connection between that question statement and the verbalization?
• 1> User has time to think
• 0> How do you know the user?
• 1> NG
• 0> What is the relationship between the user and the question text?

---

This page is auto-translated from /nishio/会話ログ2019-07-27-2 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.