Category Archives

5 Articles

Posted by knomee on

Do whatever you want with your data

KnomeeOpenFormat

A key principle from the quantified self movement is that you should use tools that let you do whatever you want with your data, that is, that do not keep the data locked in some proprietary format.

Knomee stores two types of data on your iPhone:

  • measure data, that you generate each time you record a value from one of your quest.
  • quest meta-data, that describes how the measure data should be interpreted.

Measure data can be exported and imported through CSV files, as shown on the illustration.

  1. to export your data, you press the export button on the "chart view" (small rectangle with a outward arrow), and Knomee produces an email addressed to you with a CSV (comma separated values) file of all the measures from this quest. You can upload them in Excel or any tool of your choice.
    The format is:  date, longitude, latitude, measure target value, measuer first factor value, second factor value, third factor value.
  2. to import your data, you open the measure list interface (second screenshot on the illustration) and you click the import button (small rectangle with an inward arrow). This will open a text zone where you can paste any CSV file that you have copied from your phone. Using these two steps is a way to perform back-up & restore.

Meta-data can be exported and imported as JSON files.  The JSON format is a self-evident description of your quests, its attributes and its trackers.

  1. to export a quest, you open the edit interface and click on the export button. This will generate an email to yourself, with the JSON description of your quest. Remember our promise to use only email as a tool to move data out from the application so that you may control this flow better.
  2. to import a quest, you open the quest list interface (as if you were to select a quest), and you will seen the import button (same, a small rectangle with an inward arrow). This will open a text zone where you can paste the quest description.

What you need to remember is that your data belongs to you (we cannot see it at Knomee) and that it is not stuck in your app, you can get it back whenever you want.

Knomee 2.0 will make sharing quest descriptions easier since we will create an open GitHub with all the Quest Library descriptions. Anyone will be able to add to this list of quests.

 

 

Posted by knomee on

Knomee “sense-making” algorithms just improved !

A new version of Knomee, 1.9, has been made available on the Apple AppStore just before Thanksgiving. It has been almost six months since the previous (1.8) release, we took the complete summer to re-calibrate the insights and forecasting algorithms. After a year in existence, we have accumulated enough experience to improve the robustness and the relevance of Knomee algorithms. Insights, scores and data visualization explanation will be more relevant and robust in the future.

We did not only improve “the engine under the hood”, we have also added a number of significant improvements to make it easier to use Knomee:

  • We have added quests categories : sleep, mood, food, health and activity. It makes navigating and selecting new quests easier. You will recognize the quest category with a small icon that is displayed next to the quest name in your quest list or library. Sleep tracking is on the rise since the amount of evidence that sleep is our #1 tool to improve our well being keeps piling up. There are many apps and devices to help you track your sleep (we like AutoSleep) but there is only Knomee to help you understand why you are getting good or bad sleep numbers.
  • The user interface for quest score has been simplified, and we dropped the “ken score” name that was confusing to most of you. The quest score tells you how “interesting” and “robust” your quest is. A score below 50% says that either you do not have enough data, or that your quest is not insightful, that is, the factors that you are tracking do not seem to play a significant quest towards your goal. Remember that Knomee’s number-one value is to help you find out if there is a relationship between differents aspects of your life that you are tracking. Most often, there is none ! We are “fooled by randomness” to quote Nassim Taleb. A low quest score tells you that you may be looking for sense where there is none.
  • Navigation between the different user interfaces leverages iOS transitions better. The user manual with its tip section has been improved, it should be easier to learn about Knomee while using it.
  • The use of coloring for feedback has both been extended (when you enter a new measure) and improved. The first part is that Knomee signals you whenever you enter a new measure whether you are close or far from the target values that you have set. Knomee uses the same color scheme to indicate “good” to “worse” values, and remember that you can change it in the options. If you are new to Knomee, the “blue” color setup is the simplest to use because of its vivid colors : red is bad and green is good. The second part is what make Knomee different from other tracker apps: as soon as you have enough data, Knomee uses red/green colors to tell you which factor “help” or “play against” your target. You get this coloring in the “Chart user interface” (click on the “eye button”) and with the quest score interface.
  • Insight generation has been improved both in wording and relevance. When you have enough data, and if you have enabled notifications, Knomee will send you insights daily in the form of short notification messages. All this information is available at any time in the “chart” user interface.

 

 

Knomee uses a family of algorithms for analyzing and forecasting self-tracking time series. Stay posted on our web site since we plan to share more about this in 2019. When you use Knomee most of this is implicit : Forecasting is used to animate the sliders and make your self-tracking more efficient, trend analysis is embedded into the mountain icons (the mountain shape tells you about the distance to the target and the weather tells you about the trend), tracker scores are transformed into colors and insights.

When used right, Knomee has the potential to change your life and help you to improve your health significantly - it has happened to this author. This has nothing to do with the technology that is embedded in the Knomee mobile application but everything to do with the fact that behavior change can indeed improve your health and your wellbeing. One of key challenges that the Knomee team faces is to keep simplifying the Knomee user experience so that everyone can enjoy the benefits of a mobile-computer-aided self-tracking. This is why we constantly ask for feedbacks from you, our early-users community.

Posted by knomee on

Privacy by Design – Knomee and GDPR

PrivacyKnomee

Knomee was designed with privacy as our first goal. Knomee is like your private notebook, you can use it to track whatever you want, without worrying about what could happen to your data or who could see them. This is reflected into our Privacy Statement. There is more than simply keeping data and computation on your phone ... Knomee follows "Privacy by design" as defined by GDPR:

  1. Your data is your data, it does not leave your phone unless you ask for it and we cannot see it. Knomee was created with two principles : (1) Self-tracking requires sense not to become rapidly boring  (and self-tracking is good to you, this is a scientific statement) (2) We are all different, self-tracking needs to be fully customizable and becomes extremely personal when relevant. Hence "data privacy" is not a feature, it is the reason for delivering this app, with "nothing on the cloud" and "everything on the phone". Notice that if you lose your phone, your data is lost as well.
  2. All data stored in your phone is visible to you and you can edit it. There is complete consistency between what you see and what Knomee uses for its insights, its forecasts and its analysis. This a great principle from GDPR : you know exactly what the apps store and uses. Note that it makes the app slightly more complex than most trackers. It obviously comes from the ability to customize each quest to you exact liking ... but it is also the reason for the rich data visualization that is available with Knomee.
  3. All data that is stored in Knomee can returned to you if you desire, through an email that contains your data in an open format. Measure data is sent in a CSV file that may be uploaded into any tool, such as Excel (TM). Quests are exported as a human-readable JSON string for better interoperability. This was a strong request from our early users ... and it gives you peace of mind since if you decide to stop using Knomee, you can keep all your accumulated data.
  4. Knomee uses four services from iOS that requires your authorization since they have an impact on your privacy : geolocation, iCloud, notification and HealthKit. Not only Knomee follows strictly Apple's guidelines and asks for your approval before using any of these features, the home screen shows at once if you are using any of these, making it easy to understand what Knomee is doing and to change your mind.

Knomee has a data privacy officer and chief algorithm officer. Our intent is to share our algorithms with the scientific community and to get them published. Although Knomee uses reinforcement machine learning, it also uses a framework (EMLA) that makes all its algorithms safe and auditable.

Posted by knomee on

Knomee Sense-making Algorithm is Grown from your Data

In his famous 1995 book “Out of Control”, Kevin Kelly wrote that smart systems should be “grown, not designed”. He meant that intelligent behavior should emerge from collected data and experience and not engineered in a top down way.

This is exactly how Knomee “smarts” (as in “self-tracking with sense”) have been developed: the pattern-detection algorithm is grown not designed. Knomee holds an “algorithm factory” in your smartphone, that “grows” a specific algorithm from your data, which is unique by construction.

 

(1) Knomee uses AI for Forecast and Statistical Validation

Knomee uses a number of techniques to provide insights and feedback. Most of it is classical statistical lore, but Knomee uses artificial intelligence to craft an algorithm that tries to « understand » your data, which means here to detect a collection of relevant patterns. This algorithm then serves two purposes. First it is use as a « forecasting » oracle. This is useful since it means that when you open Knomee the tracker sliders are usually positioned pretty close to where you would like them to be (10% to 15% error on average). This makes tracking faster ... and fun. This is the most convincing usage of « forecasts »: there is no way that Knomee could predict your future with the small amount of data that you track, but making Knomee « active » makes it faster to use ... and more fun ! Once you have enough data, it is actually amusing to see when Knomee gets it right and when it does not (usually, these are the most interesting self-tracking moments). The second use of this « smart » algorithm is to evaluate the relevance of more classical statistical observation. The scoring that Knomee reports about the influence of factors (tracker, time, location) is a combination of correlation and contribution to the AI insights.

 

(2) How to Grow a Unique Algorithm from your Data 

The emphasis in Knomee is on robustness much more than on precision. In the world of “small time series” (which is precisely why you get with bio-rhythms), high fidelity forecasting is an illusion and the common curse is “overfitting”: trying desperately to see some sense where there is none.

This forecasting algorithm is produced using program synthesis and reinforcement learning. Knomee has crafted an abstract description of meaningful patterns for biorhythm time series (a term algebra) and use randomization techniques to explore the wide space of possible variations. It then selects an evolutionary meta-search method to optimize the programs that better fit (reinforcement) according to their ability to explain the data. The search space includes the set of classical techniques such as k-neighbors or regression, but the evolutionary control protocol is geared at escaping the classical overfitting trap (after all, we never expect you to self-track a large amount of data).

We call the meta-algorithm that runs in your smartphone RIES for Randomized Incremental Evolutionary Search - it is a short-time series variation of techniques that were developed many years ago.  It is part of a method named EMLA (Evolutionary Machine Learning Agents); the « Incremental » specificity of the Knomee implementation is that it is optimized to fit the limited capacity of a smartphone (from a machine learning perspective).

 

(3) This Algorithm is Unique to You because You are Unique 

The RIES "algorithm factory" produces an algorithm that is truly unique because it is grown from your data. This algorithm is born on your phone and stays there. No-one will have access to the set of insights that is embedded into this algorithm. This approach is not meant for scaling or abstracting from multiple individuals.

The most interesting characteristic of EMLA is its ability to avoid false positives and let you know if your data has no relevant or statistically significant insights. This is especially critical for users because we get many of our quests wrong! We believe that we could improve some aspect of our well-being by changing our behavior ... and it simply does not work. As Mark Twain famously quoted « It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so. »

If you track “noise” (random data), Knomee will avoid overfitting and tell you that nothing much can be learned from your self-tracking data. It may does not sound like much, but it is a great feature of Knomee and something that distinguishes it from dubious so-called machine learning applications.

Posted by knomee on

Knomee makes sense from the trackings collected through HealthKit

Self-tracking with your smartphone or connected wearables is simple and efficient, but all you get is data. Self-tracking is a great habit and connected devices make it very simple since we can collect our heart rates, our sleep time or our daily number of steps effortlessly. Thanks to Apple HealthKit, all these devices or apps on our smartphone can produce tracking data continuously that is collected and sorted out. However, all you get eventually is a set of charts, lots of visual representation and a long catalog of measures. At first, it is exciting to look at this freshly collected self-tracking data, either on the device specific app or on Apple Health app, but the interest wears down quickly.

Knomee is the missing link to maximize the value that you may get from your HealthKit connected devices. Knomee is designed to help you understand yourself better through self-tracking. Self-tracking is good – it is a scientific fact – but it is tedious. Connected devices are a great opportunity to self-track:

  •  Fitbit, Garmin, Nokia, Xiaomi … and of course the Apple Watch !
  • Nokia connected scale (ex-withings), sleep monitors, ...
  • Obviously the phone itself (steps count and multiple apps)

Knomee allows you to select imported trackers when you define your quests, that will ensure that Knomee automatically import – if you decide so - the relevant data through HealthKit. Knomee makes it easy to import data from your HealthKit sources automatically:

  •  Knomee knows how to import the daily number of steps, the weight, the average daily heart rate, or the number of sleeping hours of the previous night. This list will be extended with future versions of Knomee.
  •  When you define your own quest by creating trackers, Knomee proposes “imported” as tracker choice that will pick on the four previous types. You will get a warning and will be asked to allow Knomee to import data from HealthKit.
  • Measures obtained from HealthKit are represented with squares (as opposed to circles) to remind you that they were obtained automatically

Knomee embedded “artificial intelligence” agent will help you derive more value from the self-tracking that you get automatically from your iPhone or your connected devices. This is where Knomee is much more than “yet another tracker app":

  • You may see your imported tracker inside a quest, together with other self-tracking data that you think may be linked (like, does what you eat for dinner impact how you sleep ?)
  • Knomee provides you with automated data analysis : you see if some other factors, the time of day, the day of week or the location have an impact on your quest.
  • Knomee helps you to get insights from your data through personalised smart notifications.
  • Because of Knomee “utmost privacy by design” policy, data will never go from Knomee to HeathKit