GraphGrail Ai and its Vast Experience in Advanced Software Development
GraphGrail Ai, the world’s first AI-based natural language processing platform with a DApps marketplace is passing through its TGE stage and has already garnered the acclaim of several prominent rating agencies with high rankings. Such results were made possible thanks to the extensive experience that the GraphGrail Ai development team has in implementing AI solutions for businesses and government agencies.
To demonstrate and consolidate the experience of the GraphGrail Ai team for our followers, we have compiled a list of prominent cases that the development team had participated in over the last few years prior to undertaking the implementation of their solutions on blockchain systems.
The GraphGrail Ai team was involved in the development and launch of a service that searches for extremist statements based on the legislation of the Russian Federation for the Rostov Center for Forensic Expertise. The system was used extensively and had successfully detected dangerous publications. In each post of social networks, the Service identified and analyzed several different theses, expressions or reasonings, including nationalistic calls, extremist views, religious contexts, calls for violence, calls for unrest, anti-Semitic views, calls for coups d’états, etc.
The system filtered large data sets and separated posts containing dissatisfaction with government or officials from dangerous statements aiming to induce unlawful actions. As a result, the author of each post is determined along with the author’s geographic location and the category of the post. At that time, we understand the urgency, scrupulousness and social importance of the topic, so our service intelligently solved the problem of complex classification of records with the definition of the probability of the inappropriate nature of each post.
GraphGrail Ai also has experience in launching a service for analysis of banking products. The system found bank products and related problems in their corresponding client reviews, including incompetence of employees, loan delinquency, mortgage loans, rudeness, problems with ATMs, problems with the internet services of bank, erroneous write-offs etc.
The service worked on the principle of selecting data sources and up to date supports the following services — Banki.ru, the Association of Russian Banks, Topbanki.ru, Russian Financier, Federal Financial Bureau, Comparin.ru. The service collected and analyzed data by finding bank products and conveniently categorizing the associated problems. As a result, enriched lists of reviews were formed, making them convenient for further filtering and analysis.
The service was extremely convenient for large banks seeking to track positive / negative statements in their reviews, but also for clearly understanding which product or service the client was dissatisfied with. The service saved time on analytics in the marketing department and reduced waste. Smaller banks seeking new products and services also applied the service to track new products developed by competitors. This decreased the outflow of customers and allowed for rapid responses to market threats.
GraphGrail Ai also has extensive experience in the educational sector as the team was involved in analyzing educational program posts in the VKontakte social network. 289,000 posts in 274 communities, in one way or another related to education, courses, online learning, languages, etc. were analyzed and their results used for developing better quality content
Large brands had also approached the team in the past, asking to conduct analysis of the popularity of cosmetic brands based on the opinions of Runet users. The Praktika marketing agency was the client that had ordered the research. The report contained information on brands that had proven to be most popular, how customers expressed their impressions in the context of positive / negative and how such findings could be useful for brand development.
During the reporting period, a total of 127,624 reviews and posts on social networks and review sites were collected based on select keywords. The report encompassed:
1. 13 728 posts in communities VKontakte
2. 5 594 tweets from Twitter
3. 4,924 posts from Facebook
4. 103 378 posts from specialized review sites
The Russian government was also a major client of the GraphGrail Ai team as the Electoral Committee of the Rostov Region had approached to team to conduct analysis of the mass media sector for various points of interest.
The developed system solved the problem of monitoring compliance with legislation on the internet as it automatically found and downloaded electronic publications from the sites of regional media outlets. After that, intellectual processing of text arrays was carried out according to a specially developed algorithm. The algorithm, using linguistic attributes that identified campaign publications, credited each publication with scores. The more the publication had scored, the more likely it was categorized as agitation and potentially violated the law.
The electoral committee was given a range of services:
1. Monitoring of violations in elections,
2. Data collection from electronic media and news agencies for the analysis of violations of legislation during election campaigns;
3. Prompt notification of suspicious publications and reports in a convenient format (.xls);
4. Analysis of texts on the subject of unlawful agitation;
5. Classification of texts of publications on linguistic grounds;
6. Identification of parties and politicians in texts;
7. Identification in the text of special speeches that indicated agitation, such promises, appeals, insults, etc.
The scheme of work with the electoral committee consisted of the following stages.:
1. Statement of the problem as well as setting of the tasks and criteria for their implementation along with the period of work.
2. Keyword matching. In accordance with the tasks, lists of monitored subjects were drawn up, such as the names of candidates for deputies, political parties, etc.
3. Coordination of sources. The sources set for monitoring were selected, including electronic media and news agencies.
4. The system collected and stores publications in the database and then processed them. It was possible to monitor publications in real time or for a certain period.
5. The final stage was reporting on the findings.
The experience of the GraphGrail Ai team in conjunction with the immense knowledge of the advisors and specialists attracted for the development of the platform are mute testimony to the professionalism with which the team has approached the project. This also stands as a pillar of guarantee to the quality of the final product and its corresponding demand on the market.