Work with us

We are constantly looking for top people who are interested in collaborating with us on the group research topics. This page reports a not exhaustive list of positions that we plan to open in the near future. People interested in any of the positions listed below, or simply in collaborating with us, are strongly encouraged to request more information by sending email to Knowdive -positions@disi.unitn.it.

The diversity of people is one of the characteristic features of the group. We welcome people from different cultures and backgrounds. Depending on the specific topic we welcome students and graduates with backgrounds from various disciplines, including: Computer Science, Artificial Intelligence, Logic, Sociology (Data Science and Statistics), Linguistics, Computational Linguistics, Philosophy, Cognitive and Neuro Science. The positions listed below specify the type of background and competence required for that position.

We have different types of openings, as listed below.

PostDoctoral position. This position usually consists of: (PhD) student coordination, participation and management of Research Projects (e.g., EC or internal projects), personal research, some R&D work. These positions are usually announced for an initial period (e.g., one year) with possible extensions, up to six years. Only people with a PhD are eligible. The calls are continuously announced. They are published in the University call for applications Page.

Research Fellow position. This position usually consists of: R&D coordination, participation and management of Research Projects (e.g., EC or internal projects), R&D work, some personal research. These positions can be of any length. Only people with a Master are eligible. We usually require at least three years of R&D experience after the Master. The calls are continuously announced. They are published in the University call for applications Page.

Research Associate position. This position usually consists of: R&D work, participation in Research Projects (e.g., EC or internal projects), R&D work, some coordination, some personal research. These calls for applications are usually announced for an initial period (e.g., one year) with possible extensions, up to six years. Only people with a Master are eligible. We usually require a strong R&D experience. The calls are continuously announced. They are published in the University call for applications Page.

PhD position. A PhD lasts from a minimum of three years to a maximum of four years. Only people with a masters are eligible. There are two calls per year, usually around February/ March/ April and June/ July. For any applicant there are four possible outcomes: (i) not admitted; (ii) admitted (overall tuition is around 15KE); (iii) admitted with tuition wave; (iv) admitted with tuition wave + three year fellowship (extensible of up to another year) + other benefits. The calls are published in the DISI Doctoral school Web page.

PostGraduate position. This position usually consists of personal research, some R&D work. These positions can last up to one year. They constitute a very good initial experience for anybody interested in applying later for a PhD position. Only people with a masters are eligible. The calls are continuously announced. They are published in the University call for applications Page.

Thesis. We welcome applications from students from any University worldwide. We supervise both Master and Bachelor theses. The requirements for acceptance are provided inside thesis proposals. The duration is coherent with the regulations of the University degree attended by the student. The student is expected to work (almost) full time at the thesis. The student should start working on the thesis when (s)he has passed all the degree exams but one (or two, exceptionally). These positions are not funded.

Internship. This position consists of basic personal research and/or R&D work. The goal is to introduce the student to research and/or technology.We welcome applications from students from any University worldwide. The requirements for acceptance are provided inside internship proposals. There are two types of internship. The first is internships for credit acquisition. These internships have limited duration, with duration expected to be coherent with the program followed by the student. The second is KnowDive internships. The duration of these latter internships is agreed with the student. This position is often integrated with the first type of internships or with the thesis. This position is not funded.

150ore (Student Lab). This position consists of basic software development in support to the group. Only bachelor and master students from the University of Trento can apply to this position. Under this schema, students work within the KnowDive group an amount of hours in the range of (150 – 400) hours per year. This position can be iterated for multiple years. See the dedicated University Web Page for details about applications, work and economics.

Research and Development. This position consists of basic software development in support to the group. The work is expected to be of small size, usually to be finished in 1-3 months. This work must be done independently and outside the University premises. This position is assigned directly to the relevant person. This position cannot be taken more than once per year. The students of the University of Trento cannot hold this position.

Currently, we have the following openings.

No position available. For any information, please contact knowdive-positions@disi.unitn.it

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and billions of data). Example of successful knowledge graphs are the Google knowledge graph graph or the Facebook knowledge graph.
The goal of this project is to generate a library of knowledge graphs to be reused in general web applications. These knowledge graph will be compliant with the most common standards used in the Web, e.g., iCal for events, Inspire and OpenStreetMaps for locations and facilities, Google GTFS for transportation modes. The input data will be collected from the Web, e.g., DBPedia, Geonames, OpenStreetMaps. The knowledge graphs will be generated based on the iTelos methodology.

Keywords: Knowledge graph, open data, Web Standards.
Academic profile: Bachelor or Master student in computer science.
Prerequisites (mandatory): Strong programming skills (e.g., Java), basic knowledge of ER Models, UML models.
Prerequisites (optional): —
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and billions of data). Example of successful knowledge graphs are the Google knowledge graph or the Facebook knowledge graph.
The goal of this project is to generate a library of knowledge graphs to be reused in general web applications. These knowledge graph will be compliant with the most common standards used in the Web and in apps (i.e., the ones used to store contacts in a smart phone). The input data will be collected from the Web, e.g., DBPedia, Wikipedia. The knowledge graphs will be generated based on the iTelos methodology.

Keywords: Knowledge graph, open data, Web Standards.
Academic profile: Bachelor or Master student in computer science
Prerequisites (mandatory): Strong programming skills (e.g., Java), basic knowledge of ER Models, UML models.
Prerequisites (optional): —
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Nowadays, the large majority of people uses cell phones and wearable devices in general (e.g., smart watches). This gives the possibility to collect sensor data from these devices and then learn behavioral patterns. In turn this information can be used to help and support people in their everyday activities.
One example of behavioral patterns is the person social context, namely the people a person usually interacts with. This type of pattern is very important for many reasons, not least in order to avoid infections (e.g., from the Corona Virus) but also to learn about the social behavior of a person (for instance the time a person spends with his friends). This information can be learned from, e.g., location information (e.g., GPS, wireless networks), proximity information (e.g., Bluetooth, both normal and low power, or speaker recognition). This information can also be inferred from the information in the agenda when available and the contact list.
The goal of this project is to develop a system for discovering proximity information. The system will be privacy and GDPR compliant and it will have to be integrated inside a more general platform which collects various types of data. This project will consist of: (i) assessment of the state of the art and selection of the best performing open source algorithm, (ii) its implementation inside a larger system, (iii) evaluation and tuning. Depending on the availability and interest of the student, this project could be further extended with a 150ore project (only for students or the University of Trento who have applied to this program) and, finally, into a thesis.

Keywords: Sensor streams, pervasive systems, machine learning, Human machine interaction.
Academic profile: Master or Bachelor student in computer science.
Prerequisites (mandatory): Strong programming skills, knowledge of Machine learning.
Prerequisites (optional): Basic knowledge of IOS or Android.
Reference research area: Human-Machine Symbiosis.
Starting date: based on student availability.

Nowadays, the large majority of people uses cell phones and wearable devices in general (e.g., smart watches). This gives the possibility to collect sensor data from these devices and then learn behavioral patterns. In turn this information can be used to help and support people in their everyday activities.
One example of behavioral patterns is the person location change in time and his/ her transportation modality. This type of pattern is very important as the physical location is a strong indicator of daily routines.
This information can be learned from, e.g., location information (e.g., GPS, wireless networks), proximity information, accelerometer, and so on. This information can also be inferred from the information in the agenda when available.
The goal of this project is to develop a system for discovering the people mobility patterns. The system will be privacy and GDPR compliant and it will have to be integrated in a more general platform which collects various types of data. This project will consist of: (i) assessment of the state of the art and selection of the best performing open source algorithm, (ii) its implementation inside a larger system, (iii) evaluation and tuning.
Depending on the availability and interest of the student, this project could be further extended with a 150ore project (only for students or the University of Trento who have applied to this program) and, finally, into a thesis.

Keywords: Sensor streams, pervasive systems, machine learning, Human machine interaction.
Academic profile: Master or Bachelor student in computer science.
Prerequisites (mandatory): Strong programming skills. Knowledge machine learning.
Prerequisites (optional): Basic knowledge of IOS or Android.
Reference research area: Human-Machine Symbiosis.
Starting date: based on student availability.

Nowadays, the large majority of people uses cell phones and wearable devices in general (e.g., smart watches). This gives the possibility to collect sensor data from these devices and then learn behavioral patterns. In turn this information can be used to help and support people in their everyday activities.
One example of behavioral patterns is the physical activities of a person during the day (e.g., walking, running, jogging, exercising). This type of pattern is very important and relevant for the well being of people, and a necessary condition for dealing with certain diseases.
This information can be learned from, e.g., mobility information, accelerometer, and so on.
This information can also be inferred from the information in the agenda when available.
The goal of this project is to develop a system for discovering the people mobility patterns.
The system will be privacy and GDPR compliant and it will have to be integrated in a more general platform which collects various types of data. This project will consist of: (i) assessment of the state of the art and selection of the best performing open source algorithm, (ii) its implementation inside a larger system, (iii) evaluation and tuning.
Depending on the availability and interest of the student, this project could be further extended with a 150ore project (only for students or the University of Trento who have applied to this program) and, finally, into a thesis.

Keywords: Sensor streams, pervasive systems, machine learning, Human machine interaction
Academic profile: Master or Bachelor student in computer science.
Prerequisites (mandatory): Strong programming skills. Knowledge machine learning.
Prerequisites (optional): Basic knowledge of IOS or Android.
Reference research area: Human-Machine Symbiosis.
Starting date: based on student availability.

Nowadays, the large majority of people uses cell phones, where phones have high quality cameras. The goal of this project is to develop a system for computing the amount of ingested carbohydrate based on a photo of the food to be eaten. The system will be privacy and GDPR compliant and it will have to be integrated in a more general platform which collects various types of data. This project will consist of: (i) assessment of the state of the art and selection of the best performing open source algorithm, (ii) its implementation inside a larger system, (iii) evaluation and tuning.
Depending on the availability and interest of the student, this project could be further extended with a 150ore project (only for students or the University of Trento who have applied to this program) and, finally, into a thesis.

Keywords: Computer vision, supervised learning, human machine interaction.
Academic profile: Master or Bachelor student in computer science.
Prerequisites (mandatory): Strong programming skills. Knowledge of computer vision.
Prerequisites (optional): Basic knowledge of IOS or Android.
Reference research area: Human-Machine Symbiosis.
Starting date: based on student availability.

Nowadays, the large majority of people uses cell phones and wearable devices in general (e.g., smart watches). This gives the possibility to collect sensor data from these devices and then learn behavioral patterns. At the same time, many medical and health related devices now exist which collect health data and provide them via, e.g., low power blue- tooth. The integration of these two types of information is very important as it allows to help users in taking care of their well-being.
The goal of this project is to develop a system which allow to collect data from any (medical) device able to interact via low power blue-tooth. The project will concentrate on Diabetes medical devices. The system will be privacy and GDPR compliant and it will have to be integrated in a more general platform which collects various types of data. This project will consist of: (i) assessment of the state of the art and selection of the best performing open source algorithm, (ii) its implementation inside a larger system, (iii) evaluation and tuning.
Depending on the availability and interest of the student, this project could be further extended with a 150ore project (only for students or the University of Trento who have applied to this program) and, finally, into a thesis.

Keywords: sensor streams, medical devices, low power blue-tooth.
Academic profile: Master or Bachelor student in computer science.
Prerequisites (mandatory): Strong programming skills. Knowledge of IOS or Android.
Prerequisites (optional): Basic knowledge of Human – Computer interaction.
Reference research area: Human-Machine Symbiosis.
Starting date: based on student availability.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpore the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER.
The goal of this work is to generate words and corresponding meanings, or validating existing words for your favorite language. This work will be done by taking a reference language known by you (typically, but not necessarily English) and generating or validating how words in the reference language are translated in your favorite language. This work will be done manually using excel sheets.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master or Bachelor student in any degree in the Humanities.
Prerequisites (mandatory): Deep knowledge of two languages.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.
Notes:

this project can be done in remote.
The credits are certified by the Group coordinator.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpora where the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER.
The goal of this work is to generate lexical gaps in your favorite language. Lexical gaps are words which denote concepts in one language which do not have a corresponding word in another language. Thus, for instance “agritour” is a word in Italian which denotes a specific type of hotel, cannot be translated in English. Dually, the English word “to bike” does not have a corresponding word in Italian. This work will be done by taking a reference language known by you and generating words which exist in your favorite language and not in the reference language and vice versa. This work will be done manually using excel sheets.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master or Bachelor student in any degree in the Humanities.
Prerequisites (mandatory): Deep knowledge of two languages.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.
Notes:

this project can be done in remote.
The credits are certified by the Group coordinator.

Cause-effect is the main mechanism by which we represent how things happen in timeand what is consequential to what. For instance, students get a degree because they pass a certain number of exams.
The goal of this project is to develop an ontology modeling cause-effect. The model will be presented in terms of a knowledge graph featuring both part-of and isa relations. The model will be used to validate to the UniTn Digital University knowledge graph.

Keywords: Knowledge representation, knowledge graphs, ontologies.
Academic profile: Master student in computer science, master student in philosophy.
Prerequisites (mandatory):

Computer Science student: Basic knowledge of ER Models, UML models;
Philosophy student: Husserl work on Ontology, basic knowledge of Philosophy of language.

(optional): Basic knowledge of description logics, RDF and OWL.
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Computer Science Student: Basic knowledge of ER Models, UML models;
Mathematics Student: Mathematical Logics.

Prerequisites (optional): Basic knowledge of description logics, RDF and OWL.
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpora where the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER. Among many, a core difficulty in this task is the well-known problem of untranslatability across languages.
This thesis is part of a long-term project whose goal is the definition of a general methodology for translating words from one language to another. The thesis will consist of refining the existing methodology and validating it in case study consisting of the translation of a set of words from one language to another language.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master student in any degree in the Humanities.
Prerequisites (mandatory): Deep knowledge of two languages, basics of linguistics and translation.
Prerequisites (optional): Basic knowledge of computational linguistics.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpora where the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER. Among many, a core difficulty in this task is the well-known problem of untranslatability across languages.
This thesis is part of a long-term project whose goal is the definition of a general methodology for generating lexical gaps across languages. Lexical gaps are words which denote concepts in one language which do not have a corresponding word in another language. Thus, for instance “agritour” is a word in Italian which denotes a specific type of hotel, cannot be translated in English. Dually, the English word “to bike” does not have a corresponding word in Italian. The thesis will consist of refining the existing methodology and validating it as a part of a case study on a language selected by the student.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master student in any degree in the Humanities.
Prerequisites (mandatory): Deep knowledge of two languages, basics of linguistics and translation.
Prerequisites (optional): Basic knowledge of computational linguistics.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpora where the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER.
This thesis will consist of generating a corpus which should be of at least 10,000 words for your favorite language. This corpus should integrate the words for that language which are already available. This project will consist of collecting words from online resources, e.g., online dictionaries, and integrating them with those already available. A first complication comes from the fact that the same word has multiple meanings and, dually, from the fact that there are different words, i.e., synonyms, which have the same meaning. A second complication comes from the fact that words must be organized in a Directed Acyclic Graph which models the different levels of generality of words. Thus, for instance “vehicle” has a meaning which is more general than that of “car”.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master student in Computer Science.
Prerequisites (mandatory): Deep knowledge of two languages, good programming skills.
Prerequisites (optional): Basic knowledge of computational linguistics.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.
Notes: this project can be done in remote with help of a local supervisor.

The world is globalized and online. Machine translation, namely the ability for a machine to translate automatically text from any language to any other language, is one of the core functionalities which are needed to build the society online of the future. One core enabling technology is the availability of large corpora of words, equipped with an abstract representation of their meaning codified as, for instance in the Princeton Wordnet. Of specific interest are the corpora where the words from one language are mapped to the words in another language as, for instance in the UKC system CITE ONE WORLD 7000 LANGUAGES PAPER.
This thesis will consist of generating a set of words for your favorite language, via crowdsourcing. Crowdsourcing is a technique by which people around the world are asked for help in solving a given problem, in this case the problem of providing words in a given language. The new words should integrate the words for that language which are already available. The student will be given an informal specification of the words to be generated for the selected language. The student will have (i) to design the given task, (ii) to define the incentive mechanism, (iii) to implement the algorithm, (iv) to validate the results and (v) to consolidate the technique in the DataScientia platform so that it will be usable also after the the end of thesis.

Keywords: language translation, computational linguistics, machine translation.
Academic profile: Master student in Computer Science.
Prerequisites (mandatory): Deep knowledge of two languages, good programming skills.
Prerequisites (optional): Basic knowledge of computational linguistics.
Reference research area: Language Diversity, DataScientia Linguarena Initiative.
Starting date: based on student availability.

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and sometimes billions of data). Example of successful knowledge graphs are the Google knowledge graph or the Facebook knowledge graph.
The goal of this project is to develop a user interface which allows to provide multiple visualizations of knowledge graphs, in particular; timelines, maps and abstract graphs. The user interface will allow to search inside among the nodes of a knowledge graph, to explore the surroundings of a node, to select the nodes to be visualized and so on.

Keywords: Knowledge graph, data analytics visualization, user interfaces
Academic profile: Master student in computer science.
Prerequisites (mandatory): Strong programming skills using front-end languages (e.g., Java Script), basic knowledge of ER Models, UML models.
Prerequisites (optional): Knowledge of Angular JS
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and sometimes billions of data). Example of successful knowledge graphs are the Google knowledge graph or the Facebook knowledge graph.
One problem with knowledge graphs is their size which complicates their management, navigation and exploration. The goal of this project is to develop a user interface which will allow to selectively prune, at different levels of details those elements of the knowledge graph which are considered irrelevant to a specific task.

Keywords: Knowledge graph, data analytics visualization, user interfaces
Academic profile: Master student in computer science
Prerequisites (mandatory): Strong programming skills using front-end languages (e.g., Java Script), basic knowledge of ER Models, UML models.
Prerequisites (optional): Knowledge of Angular JS
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and billions of data). Example of successful knowledge graphs are the Google knowledge graph or the Facebook knowledge graph.
The goal of this project is to generate a library which will allow to generate a knowledge graph from a set of input datasets and then to manipulate it in order to make it more efficient. Examples of functions which will need to be implemented will be of two kinds. The first will allow to compute analytics about the knowledge graph itself, the second will allow to perform a set of operations on selected parts of the graphs aimed at making it more efficient.

Keywords: Knowledge graph, data analytics, entity search.
Academic profile: Master student in computer science
Prerequisites (mandatory): Strong programming skills (e.g., Java), basic knowledge of ER Models, UML models, knowledge of Javascript
Prerequisites (optional): —
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Knowledge graphs are the main technology for representing and integrating large scale data (in the order of millions and billions of data). Example of successful knowledge graphs are the Google knowledge graph or the Facebook knowledge graph.
The goal of this project is to define a set of processes, and corresponding roles, which must be followed by a Data Scientist in order to generate a Knowledge Graph from a set of input data sets. The specification will be made using Use-case diagrams, class diagrams and activity diagrams. The results will be validated by rationally reconstructing a large scale data integration projects developed in the past.

Keywords: Knowledge graph, UML, ER models
Academic profile: Master student in computer science
Prerequisites (mandatory): Knowledge of ER Models and UML models
Prerequisites (optional): —
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

We all generate thousands of media assets, e.g., photos, videos, audios. We store them in various places, e.g., the hard disk of the PC, the cloud. A lot of applications exist which allow to manage them which provide various functionalities, e.g., automatic tagging, face recognition, media search. However, despite all these functionalities, it is still hard to find the right media element(s), e.g., the right photo or set of photos, at the right moment.
The goal of this project is to define a general facility which allows for an easy management of media independently of where they are and independently of the specific organization of data. Based on this, the system will support the easy and focused exploration and visualization of media. With a few clicks the user will be able to find what (s)he wants and visualize it on the screen.
This project can be scaled at various levels of complexity and will eventually lead to an application and possibly to its commercialization.

Keywords: Knowledge graph, data analytics, entity search.
Academic profile: Master student in computer science
Prerequisites (mandatory): Strong programming skills (e.g., Java).
Prerequisites (optional): —
Reference research area: Knowledge Diversity.
Starting date: based on student availability.

Cause-effect is the main mechanism by which we represent how things happen in time and what is consequential to what. For instance, getting infected by the Coronavirus is a consequence of having been in proximity of a person which was already infected.
The goal of this project is to develop an ontology of cause effect. The model will be presented in terms of a knowledge graph featuring both part-of and isa relations. The model will be applied in the Health Domain, aligned with the main standards (e.g., Snowmed, ICD, FHIR), and used to drive the integration of heterogeneous data with a dedicated application.

Keywords: Knowledge representation, knowledge graphs, Data integration, Data adaptation, ontologies.
Academic profile: Masters in computer science.
Prerequisites (mandatory): Strong programming skills (e.g., in Java), basic knowledge of ER Models, UML models.
Prerequisites (optional): Basic knowledge of description logics, RDF and OWL.
Reference research area: Knowledge Diversity..
Starting date: based on student availability.

No position available. For any information, please contact knowdive-positions@disi.unitn.it