3. Centrum Rafinacji Informacji
Spółka z o.o.
We offer analysis - REFINING, large information resources - BIG Data for
a given purpose (e.g. prediction of results of general elections,
monitoring of technological trends, information about Poland in the
world, assessment of the company, product).
Our Advantages:
• 8 years of experience,
• Large number of experts: IT, identification and collection of source
materials, analysis of large resources
4. Information sources - the world
Identification, Collection, Monitoring and
Analysis System - SIGMA
• Business
• Science
• State & Governance
• Patents
• Tenders
• Stock Market
• Web 2.0
e.g. the most common research areas related to
technologies and innovations:
1. Engineering (plus: “electrical engineering”; “electronic engineering”;
engineering AND technology)
2. Computer science
3. Business (plus: management; business AND management;
business AND accounting; management AND accounting; business AND
economics)
4. Chemistry (plus: multidisciplinary AND chemistry, physical AND
chemistry)
5. Materials science (plus: “materials science” AND
multidyscyplinary)
5. Refining
• Information osmosis. Machine learning (shallow and deep).
• Identification of sources related to the subject of research.
• Selection and collection of publications containing materials related
to the highlighted / detailed subject matter, subject of research.
• Analysis
• Quantitative (TF-IDF statistics)
• Bigram
• Clusters (LDA method)
• Prediction
8. Machine learning
The bigram graph is intended to present the results of the analysis of sentiment pairs. The thickness of the arrow is proportional
to the frequency of the given conjunction of sentiments. The direction of the arrow indicates the sequences of words that the
word in the pair first occurred and which followed. Connection chains are joined pairs of words. They show the occurrence of
sentiment sequences.
The length of the arrow has no substantive interpretation. It aims to clearly present data so that individual elements do not
overlap. The distance of nodes is not used to draw conclusions about the connection of sentiments.
Example 3
9. Electromobility and shipbuilding
(01.2017-10.2018)
normalized abundance with a linear and exponential function with the best least-squares
fitment. The linear function is marked with a blue continuous line. The exponential function
is drawn with a red dotted line.
Example 4
Industry 4.0 (01.2017-10.2018)
10. Refining vs official
results of parliamentary
and presidential
elections 2015
Example 5
1879
1522 1540 1546
1757
560
1863
1534
1682
1574 1650
522
0
500
1000
1500
2000
2015-05-18 2015-05-19 2015-05-20 2015-05-21 2015-05-22 2015-05-23
Duda Komorowski
13. The Power of information refining
• Complete source materials.
• Transparency of the methods and calculations used.
• Full control over analytical procedures.
• The system is scalable, there is no limit to the number of
automatically identified, collected and analysed information.
• Competitively priced in operation.