The instrument SemantiCase

What is that

SemantiCase is a simple but powerful software procedure for analyze all types of open texts through NLP models, reducing effort and complexity of the analysis. Semanticase integrates models of Natural Language Processing and Understanding, machine learning, statistics and IT.

By integrating into its back-end of complexes mathematical-probabilistic models of semantic representation, allows each operator to start the analysis through a user-friendly system, delegating the intervention of the subject matter expert, not in the onerous phase of data classification and organization, but directly in the most crucial phase of interpreting the results. Through the dashboard, a representation of the data is available facilitating the analysis of qualitative phenomena and the reporting to support decision-making processes.

The semantic analysis of texts, in fact, makes it possible to integrate the usual decision-making processes, making usable and enhancing qualitative information that is generally difficult to grasp, if not through an expensive processing.

The open text, in addition to specific contents, is also characterized by connotations on attitudes, experiences and personal opinions. The text immediacy and complexity offers the writer a broader perimeter of expression and allows for an authentic communication of meanings. Being able to grasp all the texts senses and specificities through comparison and synthesis methods is the mission of Semanticase.

Making the information capital of texts analyzable and usable in organizational processes becomes strategic for organizations in order to evolve their processes.


How it's growing

Semanticase is updated by combining the procedure and robustness of semantic data analysis with the ability to generate LLMs.
With the integration of generative AI, its ability to understand textual data and the activity of exploration, interrogation and interpretation of the results is consolidated. 

Semanticase stands out for models specifically designed and checked with respect to:  

  • the import of multiple sources and the valorisation of the text as "data" with specific pre-processing tools, 
  • no-bias analysis of data with a wide parameterization of available models to be adjusted based on the data, 
  • scalability of models from small to large texts, 
  • the activation of generative Ai on specific phases of data analysis 
  • data visualization now integrated with Ai Generativa. 

Today Semanticase It represents a rich and parameterizable pipeline, which enhances various NLP algorithms, integrating the LLMs in the analysis and visualization steps in a controlled way, with a parameter setting adequate to keep bias and hallucinations under control. 

Semanticase It allows use the most interesting open source LLMs (LLaMA, Mixtral, Solar) according to the project characteristics and continuing to constantly monitor and test the emergence of new models. All analyzes take place on dedicated servers, without using APIs towards third-party services, with certainty of cost and confidentiality. 

Semanticase it remains a procedure that aims to be at the service of the data, with a user friendly process, adaptive to the analysis needs, but not automated.  

We believe in the combination of knowledge, in the possibility of understanding the operating logic of the models and we want to offer data owners and analysts specific functions and tools to work with the results of the analyzes to understand them, question them and bring them back into the work processes. 

From today more on Semanticase:  

  • topic titling (with generative Ai) 
  • summary of the contents of the topic (with generative Ai) 
  • detailed representation of the topic (sub-topic model) 
  • new organization of results with the possibility of inserting comments and notes 

Semanticase for HR - Human Resources

Semantic analysis is applied in the HR field in the various phases of Human Resource management, bringing to light the meanings expressed in open texts and making a contribution to knowledge to assume efficient decision-making behaviors.

Some application examples:

  • TALENT ACQUISITION: semantic analysis of CVs and preparation of the final shortlist (ranking on ideal profile), screening of evaluations for the meta-evaluation of the selection system.
  • DEVELOPMENT & TRAINING: chatbot and semantic engine to support informal and social learning processes (on course content and / or internal social networks), analysis of learning and skills evaluation papers, analysis of questionnaires or needs analysis interviews, satisfaction with training.
  • PERFORMANCE & CAREERS: motivation monitoring surveys, interviews, performance evaluation of leaders.
  • PEOPLE ANALYTICS: climate surveys, internal communication analysis, work group analysis, informal communication networks and trend topics and influencer analysis.

Semanticase for digital learning

It's possible to use Semanticase for various uses in the digital learning field.

Some examples are:

  • Online tutor: indexing of all course contents and integration with user support chatbots;
  • Periodic analysis of social interaction,
  • Analysis and correction of papers and exercises,
  • Survey analysis and needs analysis.
  • Indexing of didactic material to support self-study processes.

Semanticase for Customer Care

Semantic analysis can be used in the context of customer assistance for the systematic study of the flow of commercial and technical complaints sent away by customers and / or requests for assistance (also including voice messages). It is possible to analyze the complaint topics, the trends over time, the sentiment index and investigate the predictive words of churn / termination.

Semanticase for document management

Application of semantic analysis techniques for the analysis of corporate documentation, also in order to integrate semantics into document management processes.

An example of application: compliance of documents

Application of semantic analysis techniques aimed at checking the compliance of the security risk assessment documentation (defined on a regulatory and internal regulation level).
The compliance analysis is conducted by measuring the degree of adequacy / semantic proximity for each risk area through the expression of a probabilistic score with respect to the defined system of rules. This indication can be integrated into document analysis prioritization processes.

Simple workflow to start the analysis

Rich choice of analysis options and reports to be created


Integrated consultation of the results

Import from database or file

Import open text from a database field or from file folders with possible OCR of scanned documents.

Import from whatsapp

Import open text and voice from one or more groups, broadcast lists or whatsapp numbers.

Audio import

Import open text from audio files and voice messages via speech-to-text.

Analysis of meaningful words

Identify the most popular words by frequency and rank them according to the different states of a variable.

Exclusive word analysis

Identify the distinctive and exclusive words for each state of a structured variable that accompanies the texts

Sentiment and Emotions Analysis

Analyze the polarization of opinions, their trend over time and in relation to covariates; emotions’ analysis.

Identification of issues

Identify transversal themes in texts through the construction of stochastic models of the co-occurrences of words

Variations in content

Studying the characteristic words of a theme in relation to the state of the covariates associated to the text.

Linear and mixed effects trends

Studying issues based on descriptive co-variations using both linear and mixed-effect models

Reading with AI

Labelling automatic and summaries of topical with AI generative

GenAi Sub-clustering

Detailed analysis of topical and reading with Generative AI


Year tooltation during the analysis for reporting creation

Writing is always hiding something,
so that it is later discovered

Italo Calvino - If a Traveler on a Winter's Night (1979)


Le solutions


Free Trial








Data Analysis Service

Basic configuration: SI

Type of report:
based on the volume of data

Data processing:
based on the processing data volume

Specialist consultancy for the interpretation of the results:
based on the number of reports


Software license

Setup: SI

Type of report:
based on the data volume

Data processing: 
based on the processing data volume

Training: SI

Operator License: 1

Viewing License: 5

Basic fee Months:
duration of your choice (from the month, on a quarterly, half-yearly or annual basis)

Upon request, integration of the results with:
WEBSEM month
Monthly chatbot

Le thesis

Il team

The research and development team is made up of internal staff and researchers.

Senior Partner - Learning & Innovation Expert

Daniela Pellegrinic

Expert in complex learning projects, didactic innovation, design of LMS platforms, tutoring processes. Today in Piazza Copernico is involved in innovation projects in the training and human resources fields, including learning analytics systems, adaptive learning and the application of semantic analysis and sentiment analysis models to the HR world.

Innovation Advisor and Senior Data Science Researcher

Mario Santoro

Researcher at IAC-CNR in Rome, graduated in Physics with a doctorate in Mathematics. He is an expert in Natural Language Processing (NLP), sentiment analysis, topic modeling, word embedding and data science. He is responsible for supporting innovation from a scientific point of view, taking care of the modeling, feasibility and industrialization of the project.

Research Developer

Marcello Pucci

Software engineer with significant experience in development, devops and systems management. Technical skills from the java world to open source, Linux / Unix. It supports from a technical point of view the evolution of the tool and the integrations with third parties, with an approach to constant improvement in the context of solid programming.

Senior Data Scientist

Sarah Zuzzi

Senior Data Scientist, studies and interprets large amounts of data to obtain useful information on which a company can base its strategic actions. Through the processing of Big Data, the data scientist is able to make the information hidden in the data understandable, and to transform the data into new knowledge and opportunities.

Lorenzo Pozzi

Lorenzo Pozzi

Data Scientist specializing in Machine Learning (ML) and Natural Language Processing (NLP). In Piazza Copernico I apply models for language analysis in order to understand their meaning and extract useful information for solid business growth.

Write us

Get in touch with research teams to evaluate the feasibility of your project