Data management and extraction
BUP assists its customers in all the phases of the data management process, from the extraction of knowledge from free text, up to modelling, reconciliation, integration and linking of data.
Knowledge extraction systems
A knowledge extraction system is able to discover and interpret heterogeneous data sources (structured and non-structured data), available in different languages, transforming raw data in knowledge. Extracting knowledge entails, among the others, the strengthening of data search capabilities as well as the construction of complex network of relationships and logical connections among data and documents.
BUP offers consulting services spanning from the comparative assessment of products available on the market, up to the implementation of personalized software based on specific customers’ needs.
(Linked) Open Data production
Open data is public data available in an open format that can be distributed and reused by anyone for any purpose. Specifically, linked open data represents the maximum level of quality of open data that can be achieved thanks to the use of standards for semantic interoperability that facilitate data integration on the web. This allows one to evolve from a partial vision on the data to a systemic one, also exploring unexpected knowledge paths.
BUP offers consulting services and tools to support customers in the automation of all the phases of the (linked) open data management process: from the modelling/production up to the publication, with a specific focus on accessibility and reusability.
Ontology design & maintenance
An ontology is a formal and shared representation of the principal concepts of a domain. It allows one to apply advanced automated reasoning techniques to data that would not be possible with traditional database modelling approaches.
At BUP we talk to experts, collect their requirements, design, engineer the ontologies and we assist customers in the use of these tools within their business processes.
BUP has gained strong experience in the fields of culture, fishing, agriculture, economics and finance, smart cities, scientific data but it is also able to deal with any other domain.
Data reconciliation & integration
Data that we manage every day comes from multiple heterogeneous sources or by stratifications over time of different management. For this reason, data is usually not uniform and must be reconciled; that is, data is to be led back to a consistent representation.
BUP solves this issue by providing automated tools to address data reconciliation and integration processes. Once reconciled, data is structured appropriately to maximize its integration.
BUP provides automated tools for data querying (API REST) and integration, thus facilitating collaboration between different business sectors and b2b processes.
Text understanding and analysis
Thanks to an accurate analysis of any text written in any natural language, BUP helps you to understand within the text the WHO, WHAT, WHY and HOW.
Language detection
Given any text, even very short, written in any natural language, BUP is able to automatically, promptly and accurately identify the language in which the text was written. In particular, BUP is able to recognize the language used in the text among the 100 most popular in the world.
Sentiment analysis
It is a method of analysis that, at real time, allows one to detect the attitude, emotions and opinions expressed in a text.
BUP is able to identify the overall sentiment of a text and detect such moods as anger, sadness happiness, etc., over specific topics and entities in the text. BUP services allow you to recognize the person who expresses an opinion, if present.
Entity recognition & linking
It allows one to identify entities (people, places, products, brands, dates, etc.) quoted in the text and to link them to knowledge bases available on the web (Web of data).
BUP offers services capable of recognizing and linking generic entities included in a written text, and of customizing those entities within a specific domain. The service can be used to understand whether the text mentions a certain entity, for example: your own product or a product of a competitor, a brand, a point of interest, and more.
Text classification
It lets one find the main topics discussed in a text.
BUP offers a classification service for generic documents and / or web pages through the analysis, in addition to textual contents, of the metadata of the documents. Topics are defined in a standard taxonomy that can however be customized according to customer’s specific needs.
For example, the service can be used to define user profiles based on visited web pages so that to successively propose targeted marketing campaigns.
Frame detection
It allows one to identify in a text situations and / or events, and to recognize the involved entities as well as their roles.
Thanks to an inventory of tens of thousands of situations and events, the service detects their occurrences in the text, actually covering all the common situations.
This is made possible by the use of innovative technologies for both the recognition of semantic frames, and the processing of the natural language.
Topic detection
It lets one find the main discussed topics within a text.
BUP offers a service for extracting topics for generic documents and / or web pages through deep text analysis. Unlike text classification, the topics are not predefined and bound to a taxonomy; rather, they are freely extracted according to their relevance. For instance, the service can be used to automatically tag posts, newspaper articles, and more.
Our text understanding and analysis services in action
Follow the example below
