A new command called âUPKW commandâ was created today, and I will consider explaining it.
- I wanted to explain a new feature, but he started talking about âhow it used to beâ or something like that.
- This is an inadequate explanation of the user perspective
- After plowing through, the âwhat is the value?â question verbalized something close to user value.
- I realized that the explanation to the user should bring this to the front of the line more often.
What would you like to see happen in this conversation? A new command called âUPKW commandâ was created today, and I will consider explaining it. What type of UPKW command is that UPKW command? Up until now, when it was running as a bot in an existing chat, such as Slack, the commands were entered by a human as a string. So they were short strings, and basically took no arguments, and were targeted to the keyword selected immediately before the command. The implementation that took arguments was implemented, but it was not used because it was too cumbersome. What kind of Slack is that Slack? A text-based user interface, also known as CUI. What type of user interface is it? In creating the Web version, we decided to create our own user interface, and the idea naturally came to us that it would be a good idea to install the commands as buttons. Is there anything else about its user interface? Even commands that take arguments can be made into buttons for each argument. What type of command is that command? The first command I implemented was the NGKW command, which âremoves keywords used in the previous question when they are keywords you donât want to dig into. Is there anything else about that command? This led to the idea that it would be nice to have a âcommand that tells the user to ask for more of a keyword that was not used in the previous questionâ. Where is that command? They are lined up as buttons below the input fields. So you can separately present âkeywords that were not chosenâ that do not appear in the chat document. What type of keywords are those? Keywords extracted from user input. Is there anything else you can tell us about those keywords? Users implicitly have keywords that they want you to dig into and keywords that they donât. By actually being asked a digging question, what they want you to dig into becomes an output and observable as a response of âNo, donât dig into thatâ or âAsk me more about thisâ. What is it that you value? Software that can do this is valuable, and the UPKW command is valuable because it contributes to this. Is there anything else about that UPKW command? If used, the target keyword is more likely to be used in the question. The information that a user pressed this button is information that the user values the keyword selected by the button more highly than the keyword selected immediately before it, so if a lot of this information is collected, it could be used to improve the keyword selection algorithm. Where are those keywords? This is a button that expresses the feeling that the chosen keyword itself is not bad but the question just before is not so good đ. What type of last minute is that last minute? NGKW Last minute
-
We want to encourage the output of what is not yet clearly output in humans by having software work on it, and software that can do this is valuable, and the UPKW command is valuable because it contributes to this.
- Where is this UPKW command? Keywords are presented in a prominent position below the input field.
-
We want to encourage the output of what is not yet clearly output in humans by having software work on it, and software that can do this is valuable, and the UPKW command is valuable because it contributes to this.
- Where is this UPKW command located? đ
-
We want to encourage the output of what is not yet clearly output in humans by having software work on it, and software that can do this is valuable, and the UPKW command is valuable because it contributes to this.
- What does this UPKW command look like? Mysterious communication, as if you are responding not to the words spoken by the other person, but to the words that were not spoken.
-
We want to encourage the output of what is not yet clearly output in humans by having software work on it, and software that can do this is valuable, and the UPKW command is valuable because it contributes to this.
- Where does this UPKW command come from? When the bot selects keywords to use for a question, it internally scores them and chooses the largest one. At this time, there are keywords with high scores that were not chosen, so they are used Where are those keywords located? Iâm not sure how to do this đ (on second thought, I havenât implemented the server side of this command yet) Where is that command located? UPKW Question
-
When the bot selects keywords to use for a question, it internally scores them and chooses the largest one. At this time, there are keywords with high scores that were not chosen, so they are used
- What type of question is this? Questions to make users verbalize what is bothering them Is there anything else about that question? I think it has the effect of forcing us to move our viewpoints. By looking at things from a different perspective, new insights can be gained.
-
Questions to make users verbalize their inner blur
- Where is this question? đ
-
Questions to make users verbalize their inner blur
- Where is this question? đ
-
Questions to make users verbalize their inner blur
- What does this question seem to be? I used to say like a fish hook, but based on the conversation weâve had so far, itâs not quite the same. Itâs like someone who stands up and takes someone around who is sitting and thinking.
-
Questions to make users verbalize their inner blur
- Where does this question come from? It comes from within this chat system.
-
Questions to make users verbalize their inner blur
- How do you know this question? Displayed on the phone at hand
-
Questions to make users verbalize their inner blur
- This begs the question, what should happen? It is good when users get a new realization, âOh, I see. To do that the question needs to be what needs to happen? They need to be a good balance of questions that are not obvious and easy for the user to answer, but also not misguided and far removed from the userâs interests. The question is can it? Not entirely impossible, but not clear how it can be achieved.
-
We want to encourage the output of what is not yet clearly output in humans by having software work on it, and software that can do this is valuable, and the UPKW command is valuable because it contributes to this.
- What is the relationship between this UPKW command and the question? A command that pulls back ânot that way, this wayâ when the bot starts asking misguided questions. So you get closer to a âgood balanceâ. What is the relationship between that command and the question?
consideration
- If this log is chopped and rearranged in Regroup, it would be a good case study. - Organize logs of listening chats by inscribing them on sticky notes. https://keicho.netlify.app/#talk=yAJxKTwc1ryiUTUbihUA
This page is auto-translated from /nishio/æ°æ©èœăźè§ŁèȘŹăèăă 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. Iâm very happy to spread my thought to non-Japanese readers.