- Written on 2015-07-13, discussing how to help my writing in the past combine with my current thinking.
- There is talk of Similar document search by word vector and a Wiki system that would later be called Scrapbox. #warming up
-
When I reread my past writings for some reason, I sometimes find myself thinking, âOh, this is what I wrote, this can be applied to what I am thinking about now,â and I wonder if this process can be enhanced by software or methodology.
-
There is a similar but different âOh, I think Iâve written about this before, but where?â.
-
We know it exists, we know where it exists, we know where it is.
-
I can see that it might be relevant, I can see that it is relevant, I can see the content.
-
Search engines will tell you where to find documents that may be relevant by searching for keywords that may be relevant.
-
What do you do when you donât realize that âif you search, youâll find something that might be relevant?â What if you donât know the keywords? - Hideki Kondo It is convenient to always talk to people and they will associate themselves with it!
-
How does the person make the association? If we can find that out, a computer might be able to read everything and make the associations for us.
-
Tatsuo Yamashita Similar document search is good, isnât it?
- To put it in detail, the documents are term-vectored in advance, and the similarity of all combinations is calculated. When a document is presented, several documents with high similarity to it are also presented.
-
In the âwhen I present a document, I present a document that is highly similar to itâ approach, you need to input a new verbalization of âwhat I am thinking aboutâ in order to realize that it is useful for what I am thinking about right now. I was thinking âchicken and egg if that verbalization is what you want to supportâŠâ but when I think about it, you could just raise a few keywords. Maybe thatâs the right approachâŠ
-
Tatsuo Yamashita Yes, yes. As for input, just make your thoughts into word vectors (i.e., multiple keywords) instead of sentences!
- If you are Mr. Nishio, why donât you say something along the lines of âintellectual production techniques by word2vecâ or âidea support by word2vecâ or something along those lines?
-
I thought about something like that, but I was bothered by the similarity between âBag of Vectorsâ when the words are vectorsâŠ
-
Tatsuo Yamashita Why donât you just do it all?
- There are some papers on word2vec to doc2vec, so I was wondering if there is a way to make it work. Well, maybe it will work in a wild way, such as averaging all the words in a document and making it the vector of the document.
-
Toshiyuki Masui Why not Wiki?
-
-
How can we use the Wiki to enhance the âprocess of realizing that what we were thinking about in the past is useful for what we are thinking about nowâ?
-
Toshiyuki Masui Gyazz, you will see pages that use the same keywords, which may remind you of old ideas.
-
I see, same keywords. Are those keywords extracted by morphological analysis or something? Are they added by humans?
-
Toshiyuki Masui A human being will put it on.
- Itâs not so much the keywords as the page name. Since there is no distinction between page name and keywords.
-
I see. What I was envisioning now was a blog post or a post to Facebook, but even now, porting it to a wiki might be effective in strengthening future recall.
-
Toshiyuki Masui I have several thousand pages of my own, so I am warming to the idea.
-
By the way, what is this graphical thing?
- Kohei Okubo How about evernoteâs function to suggest past related articles?
This page is auto-translated from /nishio/éć»ăźèȘćăźæžăç©ăšä»ăźæèăźç”ćæŻæŽ 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.