1)Thesis Title: Data Analysis and Visualization for Transparent User Profiling
Type: BSc/MSc
Description of Topic: When a user provides consent for some of their personal data to be used towards creating their profile and customizing their experience, a natural question arises on whether the user is really informed of what data is being explicitly gathered and what inferences can be drawn from that data. If, for example, a user has chosen to share their geo-location with their smartphone, but has declined to share other demographic information, is the user aware of what stream of data is being gathered by their smartphone, and that this stream of data might reveal indirectly their gender, age, or marital status? How can we most transparently present to the user what data is being gathered, and what inferences can be drawn from that data. Depending on the internship type, the project can be adapted towards the visualization of the gathered data, towards the formal analysis of the inferences that can be drawn, or towards other aspects of the problem that can be discussed and identified with the intern.
 
 
2)Thesis Title: Natural Language Interaction for Machine Coaching
Type: BSc/MSc
Description of Topic: When a human interacts with their machine assistant, the scope and language of their interaction is typically fixed by the developers of the assistant. Even in those cases where the assistant is capable of adapting to the human, the feedback that the human provides to facilitate this process is typically very restricted, and comes in the form of categorizing objects into classes (in support of supervised learning) or reacting to the execution of policies (in support of reinforcement learning). Yet, in real life, the adaptation of a human assistant comes through a much richer form of interaction, where both the assistant and their counterpart engage in a natural language dialogue, with each interlocutor explaining why they chose to take an action and/or why they consider an action to be (in)appropriate. This process of coaching, rather than simple supervision, allows for the much more efficient and robust transfer of knowledge to the assistant. This project seeks to investigate how this process of coaching can be used when the assistant is a machine (see https://www.researchgate.net/publication/334989337_Machine_Coaching ), focusing on how the machine will generate natural language explanations for its internal symbolically-represented knowledge and inferences, and how the natural language explanations received by the human counterpart will be turned into this internal representation. Depending on the internship type, the project can be adapted towards the natural language processing or generation part, towards the formal analysis of the coaching process, or towards the empirical investigation of the effectiveness of a natural language interaction for machine coaching.
 
 
3)Thesis Title: Reconciling “Data for Learning” with “Data for Arguing”
Type: BSc/MSc
Description of Topic: The data-driven view of Artificial Intelligence that is currently at the focus of research and industrial attention can be thought to treat a dataset through a collaborative prism, as a collection whose members have certain commonalities; one wishes to identify these commonalities by embracing the statistics of the collection over the views supported by any single member. Case-based reasoning, on the other hand, treats a dataset through a combative prism, as a loose grouping of individuals that support divergent views; one wishes to resolve the tension between these divergent views by identifying which individual makes a stronger case over the other members of the collection. The project seeks to reconcile these two views of data, by bringing together techniques from machine learning and formal argumentation. Depending on the internship type, the project can be adapted towards the development and analysis of a formal framework for this reconciliation, or towards the design and empirical evaluation of simple heuristics that acknowledge both the statistical and argumentative nature of data.
 
 
4)Thesis Title: Teaching Computation through the Movement of Social Insects
Type: BSc/MSc
Description of Topic: Insects like ants and bees are known to exhibit complex social behavior, not least of which in the way that they move, and this despite their limited and local sensing and decision-making capabilities. Research has shown, in fact, that under appropriate environmental conditions, the behavior of social insects can simulate arbitrary computations (see https://www.researchgate.net/publication/268590027_An_Ant-Based_Computer_Simulator ). The aim of this project is to develop a visual simulator of the movement of social insects, to be used as an educational tool for introducing elementary-school students to the basic notions of computation. Depending on the internship type, the project can be adapted towards the development of the visual and user-experience parts of the simulator, or towards the implementation of a scalable simulation engine that supports the asynchronous movement of thousands of insects.
 
 
5)Thesis Title: Predictive Visual Completion of Simple Sketched Figures
Type: BSc/MSc
Description of Topic: Deep neural-based architectures have emerged in the last decade as a powerful general-purpose substrate for learning directly from raw data, and have been used, in particular, for the predictive completion of images and text. It remains an interesting prospect whether shallower architectures, which presumably require considerably less training data, might be sufficient for the predictive completion of simple sketched figures. The project aims to develop an application for a touch-based device that allows a user to hand-draw simple figures, while the application attempts in parallel to anticipate and visualize the remainder of the figure. Depending on the internship type, the project can be adapted towards investigating theoretically whether simple local learning mechanisms and relatively few training examples suffice for the particular task, or towards the actual development of the application and the empirical evaluation of the efficacy of heuristic learning techniques.
 
 
6)Thesis Title: Teaching Machines to Extract World Knowledge from Text
Type: BSc/MSc
Description of Topic: A point made ad nauseam in the literature is that general-purpose AI systems will need to be able to utilize some form of world knowledge to comprehend the situations that they face. While raw text has been proposed as a potentially-useful source of such knowledge (see https://www.researchgate.net/publication/215991030_A_First_Experimental_Demonstration_of_Massive_Knowledge_Infusion ), the completely autonomous choice of learning material risks derailing the learning process towards sifting through the massive haystack of irrelevant text found, for example, on the Web, while searching for the proverbial needle of useful training material. The project seeks to develop curriculum learning techniques for directing the learning process. Depending on the internship type, the project can be adapted towards empirically demonstrating the effectiveness of curriculum learning over completely autonomous learning, towards the investigation of whether knowledge for a particular domain of interest is, even in principle, learnable from Web text, or towards the identification of appropriate natural language processing techniques for parsing text and extracting knowledge.
 
 
7)Thesis Title: Eliciting and Visualizing Actionable User Preferences
Type: BSc/MSc
Description of Topic: A key feature of personalization is the elicitation and utilization of a user’s preferences to anticipate their future choices. Any form of prompting during the elicitation process, however, might critically affect what information is divulged by the user, suggesting that a passive learning process might be appropriate. On the other hand, theoretical analysis (see https://www.researchgate.net/publication/316279661_Introspective_Forecasting) suggests that passive learning is inappropriate if one wishes to acquire actionable knowledge. The project seeks to reconcile these views by exploring the use of preference elicitation in a real-world setting. Depending on the internship type, the project can be adapted towards identifying a domain in which preference elicitation can be useful and empirically exploring the effectiveness of different learning strategies, or towards the development and analysis of formal elicitation processes that exhibit certain desirable properties.
 
 
8)Thesis Title: Crowdsourced Event Reporting in a Smart City Setting
Type: BSc/MSc
Description of Topic: The “wisdom of the crowd” has been used extensively to elicit reliable information from a collection of potentially unreliable sources, typically relying on the cognitive abilities of humans to compute an answer, recall information, or express an opinion. In the era of the Internet of Things, and with the widespread use of mobile smart devices (such as smartphones, smartwatches, etc.), humans in a crowdsourcing context can also be viewed as sensing devices with an admittedly powerful edge-computing ability to recognize events or situations. This project aims to develop a crowdsourcing application that runs on mobile smart devices and that offers a one-button solution to reporting events in a smart city setting, along with a back-end server application to collect, evaluate, aggregate, and visualize the reported events. Depending on the internship type, the project can be adapted towards the development of the applications and their empirical evaluation, or towards the design and analysis of appropriate incentive schemes and aggregation methods to guarantee the quality of the reporting.
 
 
 The applicants may address any technical or research related inquiries to the Research Office at research@rise.org.cy
 
 
 

Subscribe to our Newsletter


* indicates required