# HW4 TEC#10 DAOnysious
# TEC Praise System Analysis

### Measuring system healthiness generally
Our main interest is to understand how much TEC Commons is in line with its MVV and Ostrom's principles. In order to test this, it is necessary to analyse the effectiveness of the ongoing peer-to-peer "praise" system. In particular, we provide a text analysis of the motivations declared by users related to each "praise" given.

### Research questions

Q1. What is the most praised (real-life) public activity?
Q2. What is the most praised social media activity?
Q3. What is the most praised TEC ordinary activity?
Q4. What is the most praised Working Group?
Q5. What is the most praised TEC Internal Project?
Q6. What type of actions are most praised?
Q7. What is the most praised External Project?

### Analysis

1. Import, filter, and cleaning Data
2. Text Analysis: exploration
3. General WordCloud
Combine all praise descriptions into one big text
and create a big cloud to see which ngrams are most common.
4. CountVectorizer
Unigrams & bigrams & trigrams
5. TfidfTransformer
Unigrams & bigrams & trigrams
A useful transformation that is often applied to the bag-of-word
encoding is the so-called term-frequency inverse-document-frequency
(tf-idf) scaling, which is a non-linear transformation of the word
counts.

Tf-idfs are a way to represent documents as feature vectors. tf-idfs can be understood as a modification of the raw term frequencies (tf); the tf is the count of how often a particular word occurs in a given document. The concept behind the tf-idf is to downweight terms proportionally to the number of documents in which they occur. Here, the idea is that terms that occur in many different documents are likely unimportant or don't contain any useful information for Natural Language Processing tasks such as document classification. If you are interested in the mathematical details and equations, see this [external IPython Notebook](http://nbviewer.jupyter.org/github/rasbt/pattern_classification/blob/master/machine_learning/scikit-learn/tfidf_scikit-learn.ipynb) that walks you through the computation.

Q1. **Praises for real-life public activities**

% praises for Parties: 1.223955900214893 %
% praises for Educational activities: 1.0744651032420818 %
% praises for Conference: 0.48584509016163696 %

Q2. **Praises for social media activities**

% praises for Twitter contributions: 17.864150238250957 %
% praises for Discord contributions: 1.5976828926469215 %
% praises for Discord contributions: 0.21489302064841634 %
% praises for YouTube contributions: 2.074184807997758 %
% praises for Memes: 1.5229374941605156 %
% praises for Social media: 0.05605904886480426 %

Q3. **Praises for TEC ordinary activities**

% praises for meeting/call: 20.12519854246473 %
% praises for Hatch: 8.829300196206672 %
% praises for Survey: 0.8221993833504625 %
% praises for Forum: 2.092871157619359 %
% praises for MVV: 0.6633654115668505 %
% praises for Book Club: 1.1398673269176867 %
% praises for Gratitude: 2.475941324862188 %
% praises for Conflict Management: 0.6259927123236476 %
% praises for other TEC ordinary activities: 6.05437727739886 %

Q4. **Praises in Working Group activities**

% praises for Stewards: 1.9153508362141456 %
% praises for SoftGov: 5.39101186583201 %
% praises for Legal: 3.148649911239839 %
% praises for Transparency: 2.1022143324301594 %
% praises for Omega: 0.08408857329720638 %
% praises for Gravity: 3.1860226104830422 %
% praises for Parameters: 3.6251518265906757 %
% praises for Comms: 6.811174437073718 %
% praises for Commons Swarm: 0.8221993833504625 %
% praises for Parameters: 0.2616088947024199 %
% praises for Communitas: 0.0 %

Q5. **Praises for TEC Internal Projects**

% praises for SourceCred: 1.3080444735120993 %
% praises for Trusted Seed: 0.4951882649724376 %
% praises for cadCAD: 1.3080444735120993 %
% praises for Commons Stack: 3.260768008969448 %
% praises for Commons Simulator: 0.7287676352424554 %
% praises for Swiss Membership: 0.9623470055124731 %

Q6. **Praises for Actions**

% praises for Communication: 31.888255629262822 %
% praises for Passive Participation: 37.746426235634864 %
% praises for Leadership: 7.297019527235355 %
% praises for Active Participation: 7.530598897505373 %
% praises for Care: 24.180136410352237 %

Q7. **Praises for External Projects**

% praises for Netlify: 0.02802952443240213 %
% praises for 1Hive: 1.7378305148089321 %
% praises for Gitcoin: 0.5699336634588433 %
% praises for Gitbook: 0.3830701672428291 %
% praises for Giveth: 0.8315425581612632 %
% praises for Metaspace: 0.6540222367560496 %
% praises for Graviton: 0.5979631878912454 %
% praises for Balancer: 0.10277492291880781 %
% praises for PrimeDAO: 0.13080444735120994 %
% praises for locallyowned: 0.018686349621601418 %
% praises for Pictosis: 0.02802952443240213 %
% praises for Honeyswap: 0.018686349621601418 %
% praises for Uniswap: 0.009343174810800709 %

### Analysis conclusion
Q1. **Praises for real-life public activities**
High attention on celebration and education.

Q2. **Praises for social media activities**
Strong presence in Twitter and YouTube.

Q3. **Praises for TEC ordinary activities**
High effort invested on internal meetings.

Q4. **Praises in Working Group activities**
High effort invested in communication (Comms) and governance (SoftGov).

Q5. **Praises for TEC Internal Projects**
High effort invested in Commons Stack.

Q6. **Praises for Actions**
High effort invested in Passive Participation, Communication, and Care.

Q7. **Praises for External Projects**
High effort invested in 1Hive.

### Zoom out again, does this analysis suggest anything about system health?

The provided analysis gives an overview of the priorities that TEC Commons as a decentralised organisation is expressing. The importance of celebration (Q1), governance(Q4), and conflict resolution/care activities (Q6) seems to validate the declared [cultural practices](https://docs.google.com/presentation/d/1N4M7VhMrlwD0vhMCmYNwbAYHTH_cmoTIjYgWNtBa_Bw/edit?usp=sharing). The internal and external projects seem to validate the declared [MVV](https://docs.google.com/presentation/d/1N4M7VhMrlwD0vhMCmYNwbAYHTH_cmoTIjYgWNtBa_Bw/edit?usp=sharing) (Q5, Q7). Finally, social media communication strategy seems to be mainly focused on Twitter (Q2), and mostly internal meetings are rewarded as ordinary activities (Q1).

### Future research and project ideas

**Research ideas**
1. Relative importance by category and individual praise
2. Alignment with DAO strategic objectives - qualitative validation process is necessary (e.g. survey analysis)
3. Further explore repeating patterns (network motif) in the praise network - are collective strategies significantly emerging? If yes, which type?
4. Check if the level of personal engagement is correlated with the number of "praises" received (e.g. test autocorrelation of praises per user); we could expect different correlations per each category of praises previously defined.
5. Presence of communities and clusters (using network analysis), and their specialization.
6. More sophisticated algorithmic reward maturity assessment model ala Zargham where algorithmic/automated is state of art and basic is ad hoc/manual, where process steps/elements are then assessed - from 1) tracking activities that may constitute relevant contribution (manual tracking low maturity -> automatic as high maturity), through 2) assignment of praise/reward (ad hoc -> automated) to validation (ad hoc -> auto) to allocation (ad hoc -> continous) to integrated (aligned to financial situation / bonding curve / treasury). Scoring could be measured in "zarghams" :-)

**Project ideas**

1. Define Accountability - i.e. praising and receiving praising (consequences of people's actions)
2. How to measure value added by the praise system to Stakeholders?
3. Algorithmic governance maturity assessment model

### Supplementary Material
[Jupyter Notebook](https://www.dropbox.com/s/qfshpcy0mz82q5f/HW_4_DAOnysius.ipynb?dl=0)