Note organization for my research projects

[serbring]

Hi all,

I'm a researcher in mechanical engineering. I like taking note on every paper I read or the result of most of the analysis I carry out on my daily work. I use MS Onenote to type down notes and I believe it's a great tool. I'm searching for advices on the organization of notes. I'm trying to follow the GTD approach but I feel it's not the best solution for accadamic jobs. How do you organize your notes? Which kind of SW do you use?

Thanks

 

[jedishrfu]

What is the GTD approach?

I think ordering by date is best. I don't use any particular software but instead write my notes using LibreOffice editor or something similar whatever makes sense. I then create desktop folders for active projects and place the notes inside along with photos or pdfs ... In active project folders are placed in a desktop archives folder by year and month depending on how old it might be.

Its not very organized but I did feel that date time approach is probably the best with keywords embedded in the file name or folder name.

I even wrote a python script at one point to help organize things by dates in folders like:

Desktop/archives/email/2016-email/01-january/04-monday/note-for-some-project.txt

It allowed me to search on day name, month name and year name and project name. Further scans could pull out documents by looking for keywords inside the documents.

 

[serbring]

What is the GTD approach?

I think ordering by date is best. I don't use any particular software but instead write my notes using LibreOffice editor or something similar whatever makes sense. I then create desktop folders for active projects and place the notes inside along with photos or pdfs ... In active project folders are placed in a desktop archives folder by year and month depending on how old it might be.

Its not very organized but I did feel that date time approach is probably the best with keywords embedded in the file name or folder name.

I even wrote a python script at one point to help organize things by dates in folders like:

Desktop/archives/email/2016-email/01-january/04-monday/note-for-some-project.txt

It allowed me to search on day name, month name and year name and project name. Further scans could pull out documents by looking for keywords inside the documents.

Hi,

thanks for your reply. I used to use libreoffice, but I realized it was impractical, many files for many different projects and to search the information was hard, instead in onenote I have everything in a place and I can quickly search for anything and paste anything for any source. The date time approach is rather interesting, I shortly tried it but I felt it does it doesn't fit with research purposes because research projects do not always progress linearly in time. Moreover, the result of a research project is a paper, therefore I came up with something which is similarly organized as paper. I use the following block note for each project:
Page 1: Target -> Clear description of the result I want to achieve
Page 2: Theory -> all the paper I read organized as a summary with the hyperlink to articles or books. I use it for the introduction of paper
Page 3: Results -> all the results of my analysis with the description of the methods and material
Page 4: Ref file -> Since the file of my simulation might be many I write down a description of each file I used in the analysis. Useful when I have to recall these files after some time.

Moreover I have a unique to do list with a tag for each project.

Even if this approach seem to be fine, I realized is a bit messy especially for big project because I have a long list of information and I have difficulties to organize them. Maybe the result page might be organized in date and time. What do you write in you note? A table like the following?
Date | time | Note | File link |

Do you use any tag to retrieve information? I'm open to any suggestion.

 

[jedishrfu]

One idea I toyed with was using markdown or something similar (I had my own markdown before there was a markdown) to construct notecard files organized by date and time with keywords extracted from the note. I would construct a list of indexing keywords and if the note used one then it became associated with it.

My goal was to have a command like tool that could instantly capture a note log it and display selected notes in time order relevant to some keywords I typed in. Later, I started developing a plugin for eclipse as I did most of my development in eclipse and then I wanted to capture science/math notes and web page references... as part of my note stream (a kind of personal blog).

At the time, I had several concurrently running projects (actually I always do) and while notetaking isn't linear, it is monotonically increasing in time. I could use grep and find to locate all notes in time order taken for a given set of keywords. The user would provide the time or I'd use the last changed time of the file to automatically store it in my note filesystem structure.

Each note was like a 3x5 card small paragraphical format in separate files using markdown like syntax, easy to edit and update.

Currently, though I have a transfer script that takes my files and drops them into my notes filesystem using dated directory naming.as I described in a previous post and I have a frontend script to vi that selects files based on keywords encoded in the filename. Matchng files are displayed in a numbered list. To view a file I type the number and its displayed in vi. Its real handy for doing java projects in a system environment where you don't have the developer tools like netbeans or eclipse.available. The vi frontend can select files if the full path contains any of the string snippits and can then further select the file if it contains additional snippits.

As an example,

vm java Server -Abstract -find main

Give me all "java" files that contain the string "Server" but not the string "Abstract" in their filename and only if the file uses the string "main" inside. For my notes I used the suffix "jot" as the file extenson and so:: vm jot keyword ... -find keyword ... would find my relevant notes..

I still think about it and feel that a blogging approach is probably the best where the keywords filter out blog posts that aren't relevant. Perhaps NodeJS server app with a backing database (or grails) would be a viable strategy for the tool with the browser providing latex, svg and other image display capabilities.

 

[serbring]

Here is the description of gettng things done https://en.wikipedia.org/wiki/Getting_Things_Done. It's an interesting approach which I'm trying to adopt.

One idea I toyed with was using markdown or something similar (I had my own markdown before there was a markdown) to construct notecard files organized by date and time with keywords extracted from the note.

How do you extract the keywords from the note?

 

[jedishrfu]

In my note system, I was able to search for any word or string in my notes using a combination of find and grep. By keyword, I would define a keyword line in a given note and could build a composite list of keywords collected across my notes. From the list I was able to locate other notes that used those keywords but didn't specify them in the keyword line of some note.

For example, I would write some note on GR and not specify GR as a keyword in it. Later on, I'd write a new note where I specified GR as an important keyword. The note system on a search could scan for all notes finding that keyword somewhere within the text and now associating that note with that keyword. Its a kind of implicit keyword, identified in one note and now other notes using that keyword would be associated with it too.