Why FinSentS?

In financial industry, hundreds of millions of pages are produced daily in SEC and global fillings, broker research, conference calls, investor’s relations presentations, social media, real time news and press releases… across English, French, Chinese, Japanese… It is physically impossible to read all this. And with so many noises, clear truths and facts are often in short supply.

You want to extract useful information from these sources by reading? Ok, but imagine that infinite data VS. limited energy& time, what could happen?

Huge workload will overwhelm you.

So, what should we do?

This is where data analytics comes into play. Data analytic is what takes big data and turns it into something valuable for companies and people, and improve efficiency quickly by transforming unstructured data into structured data.

InfoTrie’s flagship solution FinSentS, based on data mining and text mining techniques, can help you to realize this process, and solve four Vs’ problems of Big data: Volume, Variety, Veracity and Velocity.

What is FinSentS?

InfoTrie’s FinSentS is a cutting edge information portal powered by proprietary Natural Language Processing technology. It is coupled with a low latency algorithmic trading portal.

FinSentS web dashboard indexes in real-time, in a way similar to what Google or Bing does for business news, blogs, and social media feed. It scans millions of sources: websites, blogs, and business news publications and processes premium and private data such as Bloomberg, Reuters or Dow Jones. InfoTrie’s machine/deep learning framework assesses in real time the relevance of each topic/company data to push meaningful content to users. Sentiment scores and other metrics (e.g. volume, buzz) are computed and displayed intuitively.

Seamlessly integrated with FinSentS, InfoTrie’s trading portal allows users to place orders through main brokers, to track their position, to follow 200+ indicators and to implement algorithmic strategies.

How FinSentS Works?

Our engine FinSentS, short for financial sentiment screener analyzes financial news, stock news, professional blogs, social medias like Twitter for over 50 thousand global stocks, FX, commodities globally in real time. Our algorithms automatically generate stock analysis data, sentiment indicators, sentiment scores, buzz scores, trending keywords, all of which could be useful signals in financial analysis and prediction.

FinSentS’s core technology is a hybrid engine with highly sophisticated linguistic algorithms (Natural Language Processing), combined with Machine Learning and Deep Learning which underpins FinSentS whole architecture.

FinSentS can easily and quickly to give you best-in-class text and sentiment analytics, such as trading sentiment, reputation sentiment, stock sentiment and so on. Instead of searching message one by one, monitoring our Sentiment score can make users take action on early market moving information, to gather information from various sources and conduct deeper research from a high perspective and superexcellent timeliness that only FinSentS and our unique data set can provide. FinSentS provides data visualization of deep insights into sentiment through a dashboard delivered through browser or mobile device.

These are the three core components of our IP and technology. Our platform can source multi-lingual content, either information on the web or from private source. We make use of machine learning, predictive analytics to do sentiment analysis for stocks, fx, equities, commodities, people/ topics. We have good research capabilities to keep up with latest techs. (Recently we are going to apply deep learning on news classification and sentiment analysis.) The information we generate is visualized in well-organized dashboard, which can be easily utilized and accessible from desktop and mobile.

1.Real-time data acquisition

Our engine FinSentS, short for financial sentiment screener can monitor financial news, professional blogs, social medias like Twitter for over 50 thousand global stocks, FX, commodities in real time.

2.Run algorithm on the data

Our algorithms automatically classify news and generate sentiment scores, buzz scores, trending keywords, all of which could be useful signals in financial analysis and prediction.

3. Visual results

Metrics such as sentiment, buzz score which are generated from multiple sources are visualized in dashboard for each entity, and for all in the form of treemap, pie chart, world map etc. History of these metrics are downloadable for reporting and statistical analysis functions are provided.

Try FinSentS for Free

Our engine produces huge amount of useful data. The tens of millions of articles analyzed in real time. We have over 15 years of history of sentiment  & buzz,  sentiment indexes for various assets and topics, and market predictions as well. These information can help generate significant alpha and monitor risk in time.The data ingested goes through several steps such as classification/clustering, entity detection, sentiment analysis to generate insights. Latest natural language processing technology and machine learning are used.We make use of open source analytics libraries. The whole processing pipelines are based in a low-latency, scalable and fault tolerant platform powered by the latest open source big data tools, like Spark, Hadoop, Storm. Our system can be deployed on-site so that customer can fully customize sourcing of data, analytics. They can control security, privacy themselves thus it is an easier to maintain and cost efficient solution.

Our platform is highly customizable and open to be defined by users. Users can add new data sources to be integrated into the system and analyzed, and choose only the sources they are interested in. New topics and keywords can be requested and defined by users. All the new sources and topics can be shared and used by all users to enable collaboration.

Our engine can ingest any external unstructured and structured data, in any language.

How is FinSentS Unique?

– Extensive Coverage

Like Google, FinSentS engine has the possibility to massively acquire data from the web, but it is also designed to processing private or premium data. A global coverage of 50,000s assets, topics, companies, and people are tracked in multiple languages in real time, with up to 15 years of history.

If you need to know what people  are really thinking about a company or stock or index, FinSentS’s technology can help you quickly determine the real sentiment behind what people are saying. No linguistic or technical expertise are necessary to understand every message. FinSentS helps you to understand others’ real mind effortlessly.

The FinSentS News Sentiment database offers daily media sentiment indicators for:
– 45,000+ stocks

The FinSentS News Sentiment database currently covers these stocks:

  • 15,000 North American stocks
  •  8,000 European stocks
  • 4,000 Japanese stocks
  • 14,000 Asian stocks excluding Japan
  • 3,000 Australia / New Zealand stocks
  •  1,000 South American stocks

– Major global stock indices
– FX
– Commodities

Each assest has 5 indicators:

– Sentiment Score: a numeric measure of the bullishness / bearishness of news coverage of the stock.
– Sentiment High / Low: highest and lowest intraday sentiment scores.
– News Volume: the absolute number of news articles covering the stock.
– News Buzz: a numeric measure of the change in coverage volume for the stock.

– Real-time, Low-latency and Fault Tolerant

InfoTrie’s superior engine can capture, analyze, and put the large volume and wide variety of real-time sources to use at very high speeds. This feature provides the guarantee for true low-latency service. The system is also fully fault-tolerant.

So, as a result, FinSentS helps keep an eye on what you concern, alerting customers to real-time leads for them to make invest further.  Customers don’t have to watch every trivial information – the most actionable sentiment score finds them directly and quickly.

– Open, Flexible and Scalable

InfoTrie enforces a flexible, secure, and highly scalable infrastructure. The system is both ready to use, with an extensive set of pre-canned analysis, and fully customizable. Advanced users are given the possibility to build and extend features, implement trading algorithms, connect to their in house system and truly leverage our APIs and data.

By using FinSentS, you can get an information edge and save hours every day:

  • Find and get alerted to critical information buried in filings, news, research and your private sources
  • Identify key data points, trends and themes on global companies
  • Never miss any important information but may have been somewhat dangerously overlooked