Quantified Self

Quantified Self refers to the activity of analyzing one's state by sensing and tracking the physical and mental state in an individual's daily activities and quantifying it, researching ways to improve the quality of life, and applying it to real life. It was introduced by Gary Wolf and Tom Kelly, and received attention during a Ted Talk in 2010. Since then, a community has been formed as a movement and has been developed through the QS website. Their mission is:

To improve quality of life by generating and sharing knowledge on Quantified Self (QS)

In this repository, I collect various data about myself and share the codes and insights I used for analysis.

Blog

Personal Assistant Kino Part 1 - Overview
Personal Assistant Kino Part 2 - Basic Structure of Chatbot Skill & Scheduler
Personal Assistant Kino Part 3 - Manage tasks and record data as easily as possible
Personal Assistant Kino Part 4 - Smart Feed that automatically saves frequently read articles
Quantified Self Part 5 - Data visualization and dashboard with KPI
Quantified Self Part 6 - Quantitative expression of a productive day and 4 years of data story

Data Source

This is a list of the types of data currently collected and the apps and wearable devices used.

Record

Record is used to collect and analyze various data about myself.

Data	Value	Description	Integration	Note
Happy	1 point ?, 2 points ?, 3 points ?, 4 points ?, 5 points ?	Happiness index at the time the question was asked
Attention	1 point ?, 2 points ?, 3 points ?, 4 points ?, 5 points ?	Concentration index when completing one task
Sleep	start time, end time	sleep time data	fitbit	python-fitbit
Productivity	Comprehensive RescueTime, Github, Toggl, Todoist	Productivity composite score	See below
-Task	Toggl ID, start time, end time, category, work content, concentration	data for one task	Toggl, Todoist, Trello	TogglPy, todoist-python, py-trello
- RescueTime	Productivity Score	Time-tracking tools for websites, apps, etc.	RescueTime
- Github	weekly commit	Number of commits on Github	Github	PyGithub
Repeat Task	Exercise, BAT, Diary	Activities repeated every day (Exercise, study organization, diary)
Total Score	overall score	Comprehensive score considering Attention, Productive, Happy, Sleep, and Repeat Task

Log

Log is a list of data for automation through ML learning.

Data	Value	Description	Note
Message	time, command text	Command text given to kino (Chatbot)
RSS Feed	Whether to save Category, Title, Pocket	Data on various Articles If you want to read an article in detail, save it to Pocket.	feedparser, pocket, python-twitter

Components

It is largely composed of the following four components.

Chatbot (kino-bot)
Scheduler
Webhook
Dashboard

Chatbot (kino-bot)

Implemented based on Slack
Multilingual support: based on Korean and English templates
Integrate with Giphy : To provide fun outside of the set template
Skill : You can register skills and set triggers by implementing them yourself.

Register Skills

Currently, a total of 27 skills are implemented.

? air_quality : Air quality forecast. (can use only Korea airkoreaPy)
✍️ attention_question : Attention survey after do task.
✍️ attention_report : Attention Report.
? bus_stop : Bus arrival information. (can use only Korea (gbus api))
? forecast : Weather forecast. (using darksky)
github_commit : Check Github push count.
? happy_question : Happiness survey.
? happy_report : Happiness Report.
? honeyjam : Easter Egg - Korea Azae Humor (using honeyjam).
jenkins_build : Build a registered project for Jenkins.
? kanban_sync : Todoist's tasks and Kanban board's card Syncing.
:thinking_face: keep_idea : Keep idea in Trello board's inbox list.
:scales: maxim_nietzsche : Nietzsche's Maxim.
:thinking_face: remind_idea : Remind Trello's inbox card randomly pick.
? rescuetime_efficiency : RescueTime Efficiency Chart
? samhangsi : I am thinking about the Samhangsi with the kor ballad! (using char-rnn-tensorflow)
send_message : Send a text message.
? today_briefing : Today Briefing - brief Todoist tasks
today_summary : Today summary - todoist_feedback , toggl_report , rescuetime_efficiency , happy_report , attention_report , github_commit
todoist_feedback : Feedback from Todoist activity.
? todoist_remain : Show todoist's remaining tasks.
? toggl_checker : Toggl time checker Every 30 minutes.
toggl_report : Toggl task Report.
⌚ toggl_timer : Toggl Timer.
? total_chart : Overall chart - productivity weekly, happiness, overall score chart.
? total_score : Overall score - Productivity (RescueTime, Github Commit, Todoist, Toggl), Mean happiness, mean attention, Exercise, Diary.
? translate : Language translation using Naver Papago api.

Scheduler

schedule based
Registered skills can be set to operate at a specified time.
- eg) Let me know about fine dust every hour, and give me a daily briefing at 8 o'clock every day.

Webhook

Implemented with AWS API Gateway + Lambda through serverless
Webhook for linking Automation Tools such as IFTTT or Zapier (delivered directly to Bot)

Dashboard

Implemented based on plotly/dash and plotly/plotly.py

Daily Schedule

Daily Habit

Daily Summary

Weekly Task Report

Prerequisites

Slack
Python 3.6

Quick Start

First, install the requirements.

pip install -r requirements.txt

Next, minimal setup is required. (config.yml)

 bot :
  MASTER_NAME : 
  BOT_NAME : Kino
  LANG_CODE : en
  TRIGGER :
    - hey kino
    - 키노야
  ONLY_DIRECT : false   // text startswith Trigger or @kino, or Direct Message
  GIPHY_THRESHOLD : 85  // all responses are random pick number (1~100) to use giphy

slack :
  TOKEN : 
  channel :
    DEFAULT : " #general "

Next, if you run the command below, the Bot will run.

python main.py

Integration with ML/DL

In the future, I would like to add more ML/DL-based functions to enable smarter automation.

CLaF: Open-Source Clova Language Framework
- OpenQA: DrQA is a system for reading comprehension applied to open-domain question answering. The system has to combine the challenges of document retrieval (finding the relevant documents) with that of machine comprehension of text (identifying the answers from those documents).