A Many-valued Logic for Lexicographic Preference Representation

We introduce lexicographic logic, an extension of propositional logic that can represent a variety of preferences, most notably lexicographic ones. The proposed logic supports a simple new connective whose semantics can be defined in terms of finite lists of truth values. We demonstrate that, despite the well-known theoretical limitations that pose barriers to the quantitative representation of lexicographic preferences, there exists a subset of the rational numbers over which the proposed new connective can be naturally defined. Lexicographic logic can be used to define in a simple way some well-known preferential operators, like "A and if possible B", and "A or failing that B". We argue that the new logic is an effective formalism for ranking query results according to the satisfaction level of user preferences.

The Role of Model Selection in Preference Learning

Learning preferences of an agent requires choosing which preference representation to use. This formalism should be expressive enough to capture a significant part of the agent's preferences. Selecting the right formalism is generally not easy, as we have limited access to the way the agent makes her choices. It is then important to understand how ``universal" particular preference representation formalisms are, that is, whether they can perform well in learning preferences of agents with a broad spectrum of preference orders. In this paper, we consider several preference representation formalisms from this perspective: lexicographic preference models, preference formulas, sets of (ranked) preference formulas, and neural networks. We find that the latter two show a good potential as general preference representation formalisms. We show that this holds true when learning preferences of a single agent but also when learning models to represent consensus preferences of a group of agents.

Learning to Respond with Your Favorite Stickers

Stickers with vivid and engaging expressions are becoming increasingly popular in online messaging apps, and some works are dedicated to automatically select sticker response by matching the stickers image with previous utterances. However, existing methods usually focus on measuring the matching degree between the dialog context and sticker image, which ignores the user preference of using stickers. Hence, in this article, we propose to recommend an appropriate sticker to user based on multi-turn dialog context and sticker using history of user. Two main challenges are confronted in this task. One is to model the sticker preference of user based on the previous sticker selection history. Another challenge is to jointly fuse the user preference and the matching between dialog context and candidate sticker into final prediction making. To tackle these challenges, we propose a Preference Enhanced Sticker Response Selector (PESRS) model. Specifically, PESRS first employs a convolutional-based sticker image encoder and a self-attention-based multi-turn dialog encoder to obtain the representation of stickers and utterances. Next, deep interaction network is proposed to conduct deep matching between the sticker and each utterance. Then, we model the user preference by using the recently selected stickers as input and use a key-value memory network to store the preference representation. PESRS then learns the short-term and long-term dependency between all interaction results by a fusion network and dynamically fuses the user preference representation into the final sticker selection prediction. Extensive experiments conducted on a large-scale real-world dialog dataset show that our model achieves the state-of-the-art performance for all commonly used metrics. Experiments also verify the effectiveness of each component of PESRS.

Preferences, Machine Learning, and Decision Support With Cyber-Physical Systems

Modeling of complex processes with human participations causes difficulties due to the lack of precise measurement coming from the qualitative nature of the human notions. This provokes the need of utilization of empirical knowledge expressed cardinally. An approach for solution of these problems is utility theory. As cyber-physical systems are integrations of computation, networking, and physical processes in interaction with the user is needed feedback loops, the aim of the chapter is to demonstrate the possibility to describe quantitatively complex processes with human participation. This approach permits analytical representations of the users' preferences as objective utility functions and modeling of the complex system “human-process.” The mathematical technique allows CPS users dialog and is demonstrated by two case studies, portfolio allocation, and modeling of a competitive trade by a finite game and utility preference representation of the trader. The presented formulations could serve as foundation of development of decision support tools and decision control.

Party preference representation

Political parties are key actors in electoral democracies: they organize the legislature, form governments, and citizens choose their representatives by voting for them. How citizens evaluate political parties and how well the parties that citizens evaluate positively perform thus provide useful tools to estimate the quality of representation from the individual’s perspective. We propose a measure that can be used to assess party preference representation at both the individual and aggregate levels, both in government and in parliament. We calculate the measure for over 160,000 survey respondents following 111 legislative elections held in 38 countries. We find little evidence that the party preferences of different socio-economic groups are systematically over or underrepresented. However, we show that citizens on the right tend to have higher representation scores than their left-wing counterparts. We also find that whereas proportional systems do not produce higher levels of representation on average, they reduce variance in representation across citizens.

Theoretical Foundations for Preference Representation in Systems Engineering

The realization of large-scale complex engineered systems is contingent upon satisfaction of the preferences of the stakeholder. With numerous decisions being involved in all the aspects of the system lifecycle, from conception to disposal, it is critical to have an explicit and rigorous representation of stakeholder preferences to be communicated to key personnel in the organizational hierarchy. Past work on stakeholder preference representation and communication in systems engineering has been primarily requirement-driven. More recent value-based approaches still do not offer a rigorous framework on how to represent stakeholder preferences but assume that an overarching value function that can precisely capture stakeholder preferences exists. This article provides a formalism based on modal preference logic to aid in rigorous representation and communication of stakeholder preferences. Formal definitions for the different types of stakeholder preferences encountered in a systems engineering context are provided in addition to multiple theorems that improve the understanding of the relationship between stakeholder preferences and the solution space.

Extensions of logic programming for preference representation

Εξετάζουμε το πρόβλημα της αναπαράστασης προτιμήσεων με τη χρήση επεκτάσεων του λογικού προγραμματισμού. Η αποτελεσματική αναπαράσταση προτιμήσεων είναι ζωτικής σημασίας σε πολλά επιστημονικά πεδία και μπορεί να αποδειχθεί χρήσιμη σε πολλές πραγματικές εφαρμογές. Οι φορμαλισμοί αναπαράστασης προτιμήσεων στη βιβλιογραφία συνήθως εμπίπτουν σε δύο βασικές κατηγορίες: στην ποιοτική προσέγγιση (όπου οι προτιμήσεις εκφράζονται με διμερείς σχέσεις προτίμησης) και στην ποσοτική προσέγγιση (όπου οι προτιμήσεις αναπαριστώνται με τη χρήση αριθμητικών τιμών που εκφράζουν το βαθμό ενδιαφέροντος). Σε αυτή τη διατριβή, προτείνουμε δύο προσεγγίσεις για την έκφραση προτιμήσεων. Η πρώτη προσέγγιση χρησιμοποιεί μια απειρότιμη επέκταση του λογικού προγραμματισμού για την έκφραση ποσοτικών προτιμήσεων, ενώ η δεύτερη προσέγγιση χρησιμοποιεί τον λογικό προγραμματισμό υψηλής τάξης για την έκφραση ποιοτικών προτιμήσεων.Προτείνουμε τη γλώσσα προγραμματισμού PrefLog, μια επέκταση του λογικού προγραμματισμού που χρησιμοποιεί ένα άπειρο σύνολο τιμών αλήθειας για να υποστηρίξει τον ορισμό τελεστών ποσοτικής προτίμησης. Ορίζουμε το συντακτικό και τη σημασιολογία της γλώσσας και προσδιορίζουμε ένα σύνολο από ιδιότητες τις οποίες πρέπει να ικανοποιούν οι διαθέσιμοι τελεστές προτίμησης έτσι ώστε η γλώσσα να έχει καλώς ορισμένη σημασιολογία. Επιπλέον, προτείνουμε μία «από-κάτω-προς-τα-πάνω» τεχνική υλοποίησης για ένα καλώς ορισμένο υποσύνολο της PrefLog που αντιστοιχεί στο προτιμησιακό αντίστοιχο της γλώσσας Datalog. Η εξασφάλιση της ιδιότητας του τερματισμού μιας τέτοιας στρατηγικής δεν είναι προφανής γιατί το σύνολο των τιμών αληθείας και το σύνολο των πιθανών ερμηνειών για τέτοια προγράμματα είναι και τα δύο άπειρα.Προτείνουμε τη χρήση του λογικού προγραμματισμού υψηλής τάξης για την αναπαράσταση ποιοτικών προτιμήσεων. Σε αυτήν την προσέγγιση, σχέσεις, προτιμήσεις μεταξύ πλειάδων, προτιμήσεις μεταξύ συνόλων από πλειάδες και υπολογισμοί σχετικά με προτιμήσεις εκφράζονται στην ίδια γλώσσα υψηλής τάξης. Τα προγράμματα αυτά μπορούν να εκτελεστούν σε πραγματικά συστήματα λογικού προγραμματισμού υψηλής τάξης και η απόδοσή τους μπορεί να ενισχυθεί είτε με γενικές είτε με εξειδικευμένες τεχνικές βελτιστοποίησης. Ανάμεσα σε αυτές, προτείνουμε μια νέα τεχνική μετατροπής λογικών προγραμμάτων υψηλής τάξης σε κλασικά λογικά προγράμματα (πρώτης τάξης) και την εφαρμόζουμε στα προγράμματα της προσέγγισής μας. Τέλος, αποδεικνύουμε την εφαρμοσιμότητα της προσέγγισής μας παρουσιάζοντας μια υλοποίηση και μια πειραματική αξιολόγηση στη γλώσσα λογικού προγραμματισμού υψηλής τάξης HiLog.

Enhanced bottom-up and reduced top-down fMRI activity is related to long-lasting non-reinforced behavioral change

Behavioral change studies and interventions focus on self-control and external reinforcements as means to influence preferences. Cue-approach training (CAT) has been shown to induce preference changes lasting months following a mere association of items with a neutral cue and a speeded response, without external reinforcements. We utilized this paradigm to study preference representation and modification in the brain without external reinforcements. We scanned 36 participants with fMRI during a novel passive viewing task before, after and 30 days following CAT. We pre-registered the predictions that activity in regions related to memory, top-down attention and value processing underlie behavioral change. We found that bottom-up neural mechanisms, involving visual processing regions, were associated with immediate behavioral change, while reduced top-down parietal activity and enhanced hippocampal activity were related to the long-term change. Enhanced activity in value-related regions was found both immediately and in the long-term. Our findings suggest a novel neural mechanism of preference representation and modification. We suggest that non-reinforced change occurs initially in perceptual representation of items, which putatively lead to long-term changes in memory and top-down processes. These findings could lead to implementation of bottom-up instead of top-down targeted interventions to accomplish long-lasting behavioral change.