On 5-7 July, Exactpro took part in the 7th AIST Conference on the Analysis of Images, Social Networks, and Texts. Iosif Itkin, Exactpro co-CEO and co-Founder, gave a talk in the Startups and Industry section of the event.
The 7th AIST took place at Moscow Polytechnic University. The academic part of the conference program featured the following topics:
- Natural Language Processing
- Social Network Analysis
- Analysis of Images and Video
- Optimization Problems on Graphs and Network Structures
- Analysis of Dynamic Behaviour though Event Data
On 7 July, Iosif Itkin, Exactpro co-CEO and co-Founder, opened the Industrial Applications sessions day with a talk on the latest software testing trends, as well as some future projections.
The AIST hackathon, sponsored by a cryproanalytics startup Santiment.net, ran parallel to the conference and was dedicated to the analysis of сryptocurrency-related data. The participants were provided with over 2 months’ worth of pricing and volumes data, relevant blockchain transaction data, as well as around 3 million Twitter, Reddit and other chat and news messages. Analysing this data, the teams were asked to resolve such tasks as:
- Finding a correlation between the pricing/volumes and social media discussions,
- Observing whether separate wallets are able to start market rise/fall, and how they can be found,
- Observing how trends affect news sources, wallets and prices,
- Classifying/grouping chats and the participant behaviour patterns based on the evolving news landscape,
- Creating an aggregate history of the coin, taking into account all the aforementioned factors.
The hackathon teams were mentored by the researchers and practitioners participating in the academic part of the conference. Four Exactpro specialists represented the company at the hackathon. “It was an exciting opportunity to be among such experienced Data Science professionals, and it was an honor to be among the teams who made it to the defence panel”, says Andrey Smirnov, Java Developer, Exactpro.
AIST conference is aimed at computer scientists and industry professionals whose research interests involve applications of data mining and machine learning techniques to various problem domains: image processing, analysis of social networks, and natural language processing. AIST boasts its interdisciplinary nature, bridging the gap between researchers and practitioners. AIST proceedings, in the form of selected papers, are published with Springer.