ðš [A Financial Hacker's Perspective] The Truth About note's Algorithm and 'Fundamentals' â Superficial Keywords or Essential Read-Through Rates?
Hello everyone. I am Kato-chan, a financial hacker who exposes the dark side of the market using primary information.
I usually dive into the deepest parts of US SEC disclosure documents (10-K and 10-Q) to dissect the hidden incentives of massive capital and the 'market lies' that border on accounting fraud, but today I would like to change my perspective a bit and perform surgery on the 'structure of note's algorithm,' the platform we are currently standing on.
The shallowness of confusing 'labeling' with 'evaluation'
I recently came across an observation that 'note's algorithm is a sloppy mechanism that misfires in response to specific words (like Apple or IBM).' It suggests that machines don't read context and confuse protagonists with supporting roles based solely on superficial character strings.
It is certainly true that if a specific company name is mentioned in an article, the algorithm has a 'mechanical habit' of automatically picking it up as a keyword and linking it to that company's page or tag.
However, that is merely a superficial labeling process of 'which shelf (category) to classify it into.' From my perspective as a professional data analyst, claiming to have 'seen through the structural flaws of the algorithm' based solely on that localized behavior seems a bit premature and shallow.
What would happen if note's algorithm were a simple mechanism that decided official picks based 'only' on the presence of specific keywords without reading the text? Spam articles filled with GAFAM company names or lofty philosophical terms would fill 'note money,' and the platform would have collapsed long ago.
The ultimate, unfalsifiable fact: 'dwell time and read-through rate'
The most important and decisive fundamental for an article to be pushed up to an official, prominent stage like 'note money' is not the presence or absence of keywords.
In corporate analysis, it is possible to temporarily inflate a stock price with good-looking IR materials (poetry). On social media, you can also gather superficial PV (access counts) or 'likes' by lining up beautiful words or acting out attractive attributes.
However, what is weighted most heavily in the deep layer (core) of note's algorithm is not those inflatable numbers, but rather 'dwell time' and 'read-through rate,' which measure how many seconds a reader stays on the article and whether they scrolled to the end and read it thoroughly.
No matter how much you dress up your outward attributes and delude yourself into thinking you've hit a blind spot in the algorithm, hollow, paper-mache writing will cause readers to leave in an instant. Content that is like 'double-entry bookkeeping,' where the public face and the underlying reality diverge, cannot ultimately deceive this cold audit called 'dwell time'.
The reason an article isn't officially picked up (or is picked up extremely rarely) is not because the system didn't pick up the words. It is simply proof of the cruel fact that the article's 'dwell time' and 'read-through rate' are too low.
What the overwhelming law of large numbers proves
I started my blog from zero followers with no social media integration, and in just one week, 17 consecutive articles dealing only with dry financial data were picked up by 'note money,' and the total views quickly exceeded 27,000. Currently, 100 days have passed since I started posting, with over 230 'note money' selections and over 77,000 views!
Looking at the latest weekly data,
I hit 7,301 views in just one week, meaning that on average, readers are devouring my dry reports over 1,000 times a day.
Are these numbers that happen due to the 'whim of an algorithm'?
Please look at these search results. The second article was selected for 'note money,' but the fifth article was picked up immediately after posting in a different category called 'popular articles,' not 'note money'.
This shows that it is not being read simply because it has the 'official stamp of approval.'
Even without the official 'note money' label, the system automatically prioritizes articles with extremely high dwell time and read-through rates as 'popular articles' and delivers them to readers.
The current situation where my name consistently appears in both official picks and popular articles is nothing less than definitive proof that the AI has granted the 'Kato-chan' account the maximum possible 'trust balance' across the entire platform, signaling that 'no matter the category, this account's articles are of the highest quality.'
Furthermore, my articles climb to the top of the 'popular articles' list within just a few minutes to a few hours of publication.
This cannot be explained by 'algorithmic whims' or 'accidental keyword hits.'
The moment an article is released, the system, based on my vast historical data, has learned that 'this account's articles generate abnormal engagement immediately after publication.'
Therefore, the AI instantly detects the overwhelming density and expectation value, pushing it to the forefront of 'popular articles' without waiting for human reviewâin other words, it is irrefutable evidence that the system immediately recognizes my articles as 'valuable primary information that should be delivered to readers with top priority.'
Cold, hard primary information dragged from the deepest parts of EDGAR,
and the logic that exposes corporate lies.
My articles are typically dry reports of 10,000 characters, sometimes exceeding 20,000 characters.
Honestly, the length and weight are enough to make even me, the author, hesitate to re-read them, and it wouldn't be strange for a normal internet article to see a massive drop-off in readers halfway through. Nevertheless, tens of thousands of readers stay glued to the screen, 'reading thoroughly' while following the numbers and facts without dropping off.
The reason my articles are valued is not because I hacked specific keywords, but simply because the system accurately evaluated the fundamentals of 'deep, long engagement (reality)' that you, the readers, have generated.
The category related to 'money (note money)' is particularly high-risk for the platform (as it is prone to fraudulent information products and illegal investment solicitation).
Therefore, the algorithm applies a weight to prioritize articles from 'accounts that have repeatedly achieved high read-through rates in the past, have no compliance violations, and have satisfied readers' over one-off articles from new, unknown accounts.
This is precisely the 'trust balance'.
Postscript: Anatomy of the Financial Hacker's Algorithmã
(Why do my articles keep getting selected for 'note money'?)
Let's go a step further and dissect the reality of this platform's 'review process' from the perspective of a data analyst.
You might wonder, 'Is note money all selected by human staff through visual inspection?' but it is physically impossible to check the massive number of articles posted every day 100% manually. Given the structure of the system, it is extremely natural to view this as a 'highly advanced hybrid of AI and humans (staged screening).'.
â First Screening: Detection of 'Anomalies' by Algorithm
The first net is the aforementioned 'dwell time' and 'read-through rate.' The system ignores superficial numbers like follower counts and monitors the reality of traffic in real-time, such as 'how long readers stay and whether they read to the end.' Only articles that hit 'abnormal engagement' significantly above the baseline here are sent to the next step.
â¡ Second Screening: Visual Check by Curators (Humans)
The 'money and business' domain, in particular, is a delicate category where fraudulent information and compliance violations are easily mixed in. For high-engagement articles that have passed the AI's first net, human eyes enter here for the first time to perform a final check on the 'quality of information' and 'alignment with the platform's philosophy'.
⢠The Ultimate Algorithm Hack: 'Trust Score (Trust Balance)'
This is the biggest reason why I have been selected over 230 times in the 100 days since I started posting. By repeatedly releasing high-quality primary information (SEC disclosure analysis) that hits overwhelming dwell times every time and consistently clears human reviews, the 'Trust Score (Domain Authority)' (the accumulation of trust in an account (domain) based on past performance and the resulting algorithmic preferential treatment) on the system reaches its limit.
As a result of the system and management learning that 'articles from this account always have high engagement and are of the highest quality,' a VIP route (express pass) has likely been built where 'once an article passes the AI's primary filter, human checks are simplified and it is immediately moved to pick-up'.
If you just line up random words and happen to get picked up once by hitting a 'bug' in the AI's primary filter, that is just noise. A trust score will never be accumulated.
Primary information of 10,000 to 20,000 characters written with overwhelming passion, and tens of thousands of readers reading it thoroughly. This 'series of unfalsifiable facts' is the only key to prying open the VIP route in the deepest part of the system.
In fact, even for me, with my track record, not all articles are selected for 'note money.' There are naturally articles that are not chosen. However, from a data analyst's perspective, this is exactly what proves the 'reliability' of this algorithm.
If it were a 'sieve' mechanism where specific accounts were automatically picked up 100% of the time, this platform would have collapsed long ago.
The fact that there are articles that are not selected is nothing but proof that the system performs a cold 'audit' of 'dwell time' and 'read-through rate' for each article every day, judging the quality of that moment.
Finally, I would like to add the biggest reason why I trust this platform and continue to write.
The truly wonderful thing about the note platform is its 'overwhelming fairness,' which does not favor only influencers with tens of thousands of followers or speakers who gather 'likes' through organized voting.
Even if you start from complete obscurity with 'zero followers' and 'no SNS integration,' if the content of the article is genuine and of a quality worth taking up the reader's dwell time, the platform will accurately detect it and push it to the front. There are few other fair stages in the current internet world where you can compete purely on the 'power of content (fundamentals).'
That is why there is no need for superficial keyword hacks or unsightly excuses. If you polish the content, the system and the readers will definitely find it.
Regardless of whether it gets on 'note money' or not, the reality that it is pulled to the top of 'popular articles' immediately after publication and continues to reach an average of 1,000 viewers per day proves the soundness of note and the correctness of my dissection technique.
Now, let's put an end to dealing with fruitless noise. Today, while paying my respects to this fair and wonderful stage, I will once again dive into the deep sea of EDGAR, my true main battlefield.
Real facts crush superficial techniques with sheer force.
Well, that's enough platform analysis; let's return to the sea of EDGAR, my true main battlefield.
There is only one watchword. 'Crush shallow, false information!'
(Disclaimer: This text is for informational purposes only and does not constitute solicitation for specific stocks, investment advice, or a guarantee of the completeness of financial analysis. Investing involves risks. Please always make final investment decisions at your own responsibility.)
#KatoChanAnalysis #NoteAlgorithm #DataAnalysis #Fundamentals #PrimaryInformation #FinancialLiteracy