What’s in the package?
News SentimentThrough the API, clients can access data from 50,000 credible sources of different ranking and market impact. Our system targets the precise information that affects the particular crypto-assets that interest you.
Social SentimentWe process Twitter, Telegram, Reddit, Github, Discord, and most recently Youtube, as well as exchange chat groups. For instance, CryptoMood detects 500,000,000 tweets every day, on average, 45 million of them are detected to be crypto-related. After intelligent filtering, our AI deals with around 25,000 of them on a daily basis
Historical DataWe can provide API users with up to 3 months of data. You can retrospectively analyse the market impact of different events in the crypto-industry on the assets that you track.
Custom IndicatorsAPI users are provided with up to 3 months of data, enabling retrospective analysis of the market impact of crypto-industry events on different tracked assets.
AI that learnsWe constantly improve and update our algorithms to provide this data as quickly and accurately as feasibly possible. You can be sure that this is the most accurate sentiment data available on the market.
How can you use the
You can connect the API to your trading bot to optimise your trading strategy.
Alternatively, you can opt to conduct independent research on price movement,
prediction, accuracy, and specified models, including fundamental analysis. The
detailed instructions are here.
CryptoMood’s data can be passed to your data analysts and interpreted in order to combine different models and make custom made analytics for your users.
Available API plans
We offer different models based on your needs if just you or your company wants to get the highest quality data through our API.
|Historical data||1 month||3 months||Unlimited||Unlimited|
|Frequency of data*||Daily||Hourly||Minutely||Minutely|
|Indicators||Social & News
|API calls||60 per minute||120 per minute||240 per minute||Unlimited|
|Cryptos||ETH, BCH, LTC||Top 10||Top 20||All|
|Business Support||N/A||N/A||Priority email / chat||Priority email / chat|
|FREE||$ 39.90 / month||$ 249.50 / month||On request|
|FREE||$ 383.00||$ 2,395.00||On request|
CryptoMood’s core is Inspired by the architecture of Google and Netflix, where, like ourselves, technologies are developed in-house. CryptoMood’s NLP and NLU engine (LangProc) is unique, it’s the one of its kind in the market. The methods CryptoMood uses to aggregate, group, and rank market data is the result of many years of research and cooperation with two European Universities and senior professionals in AI, Machine Learning, and Natural Language Processing.
Our API is powered by a lightning-fast language analytics engine. Everything that can be tracked on the internet, and is labelled, can be processed by CryptoMood’s technology. Similar technologies are being used in the advertising industry (AI targeting), sports betting industry, and political science.
Why are particular posts negative or positive?
Example 1: Positive tweet from an influencer
The algorithm detects the mention of the new technology and infrastructure development update. As well as that, it takes into account the social weight of the influencer in the community based on how many followers they have and comments they receive. Overall, positive sentiment is given.
Example 2: Negative article
In this example, the fear is multiplied by the credibility of the source, which is calculated from page ranking and other parameters like the number of the related comments in the article.
Our Data Mining Engine processing time ranges from 1 to 100 milliseconds, depending on the length and weight. After about 1 second, the item gets to the data pipeline to be analysed, filtered, and then distributed to the client. [data acquisition] -> [processing] -> distribution = max 1 second
Intelligent evaluation of the source
The data scraped by CryptoMood’s Mining Engine goes to LangProc distribution pipe, where it is split into: Base, Author, Content, Publication date, Sentiment, Sentiment, Market impact, Multidimensional Sentiment, Locations, Headline.
We use our built-in-house sentiment scoring technology with VADER (Valence Aware Dictionary and sentiment Reasoner) method to measure whether the text is positive or negative. Our technology is also known as ‘Opinion Mining’, a method used to analyse conversations, opinions, and viewpoints (all in the form of tweets) in order to understand business strategy, political analysis, and to assess public actions.
Once it is analysed, the text in its entirety is given a sentiment score. The way we measure sentiment is CryptoMood’s secret sauce, but basically the score is a result of the sentiment multiplied by the impact. A detailed description of the scraping process can be found in the API documentation.
Content cleared from noise and spam
Content cleared from noise and spam
We use statistical methods to detect if the piece of content is spam or not. It is based on text and language analysis, language, and source rank. The most innovative aspect of our spam detection filter in social media is the usage of social graphs - the system that analyses what position a certain person holds in the community. Spammers profiles usually have no value, so CryptoMood easily filters them out from a decent newsfeed.