Detecting concurrency bugs can be challenging due to the intricacies associated with their manifestation. These intricacies correspond to identifying the methods that need to be invoked concurrently, the inputs passed to these methods and the interleaving of the threads that cause the erroneous behavior. Neither fuzzing-based testing techniques nor over-approximate static analyses are well positioned to detect subtle concurrency defects while retaining high accuracy alongside satisfactory coverage. While dynamic analysis techniques have been proposed to overcome some of the challenges in detecting concurrency bugs, we observe that their success is critically dependent on the availability of effective multithreaded clients. Without a priori knowledge of the defects, manually constructing defect-revealing multithreaded clients is non-trivial. In this thesis, we design an approach to address the problem of automatically generate clients for detecting concurrency bugs in multithreaded libraries. The key insight underlying our design is that a subset of the properties observed when the defects manifest in a concur-rent execution can also be observed in a sequential execution. The input to our approach is a library implementation and a sequential testsuite, and the output is a set of multithreaded clients that can be used to reveal defects in the input library implementation. Dynamic defect detectors can execute the clients and analyze the resulting traces to report various kinds of defects including deadlocks, data races and atomicity violations. Furthermore, the clients can also be used by testing frameworks to report assertion violations. We propose two variants of our design – (a) path-agnostic client generation, and (b) path-aware client generation. The path-agnostic client generation process helps in detection of potential bugs present in the paths executed by the input sequential testsuite. It does not attempt to explore newer paths by satisfying path conditions either by modifying the input or by scheduling the threads appropriately. The generated clients are used to expose deadlocks, data races and atomicity violations. Our analysis analyzes the execution traces obtained from executing the input sequential clients and produces a concurrent client program that drives shared objects via library methods calls to states conducive for triggering deadlocks, data races or atomicity violations. For path-aware client generation, our approach explores newer paths that are not covered by the input sequential testsuite to generate clients. For this purpose, we design a directed, iterative and scalable engine that combines the strengths of static and dynamic analysis to help synthesize both multithreaded clients and schedules that violate complex correctness conditions expressed by the developer. Apart from the library implementation and the sequential testsuite as input, this engine also accepts a specification of correctness as input. Then, it iteratively refines each client from the input sequential testsuite to generate an ex-ecution that can break the input specification. Each step of the iterative process includes statically identifying sub-goals towards the goal of failing the specification, generating a plan toward meeting these goals, and merging of the paths traversed dynamically with the plan computed statically via constraint solving to generate a new client. The engine reports full reproduction scenarios, guaranteed to be true, for the bugs it finds. We have implemented prototypes that incorporate the aforementioned ideas and validated them by applying them on 29 well-tested concurrent classes from popular Java libraries, including the latest version of JDK. We are able to automatically generate clients that helped expose more than 300 concurrency bugs including deadlocks, data races, atomicity violations and assertion violations. We reported many previously unknown bugs to the developers of these libraries resulting in either fixes to the code or changes to the documentation pertaining to the thread-safe behavior of the relevant classes. On average, the time taken to analyze a class and generate clients for it is less than two minutes. We believe that the demonstrated effectiveness of our prototypes in helping expose deep bugs in popular Java libraries makes the design, proposed in this thesis, a vital cog in the future development and deployment of dynamic concurrency bug detectors.

Usability aspects of programming languages are often overlooked, yet have a substantial effect on programmer productivity. These issues are even more acute in the field of Inductive Synthesis, where programs are automatically generated from sample expected input and output data, and the programmer needs to be able to comprehend, and confirm or reject the suggested programs. A promising method of Inductive Synthesis, CombInduce, which is particularly suitable for synthesizing recursive programs, is a candidate for improvements in usability as the target language Combilog is not user-friendly. The method requires the target language to be strictly compositional, hence devoid of variables, yet have the expressiveness of definite clause programs. This sets up a challenging problem for establishing a user-friendly but equally expressive target language. Alternatives to Combilog, such as Quine's Predicate-functor Logic and Schönfinkel and Curry's Combinatory Logic also do not offer a practical notation: finding a more usable representation is imperative. This thesis presents two distinct approaches towards more convenient representations which still maintain compositionality. The first is Visual Combilog (VC), a system for visualizing Combilog programs. In this approach Combilog remains as the target language for synthesis, but programs can be read and modified by interacting with the equivalent diagrams instead. VC is implemented as a split-view editor that maintains the equivalent Combilog and VC representations on-the-fly, automatically transforming them as necessary. The second approach is Combilog with Name Projection (CNP), a textual iteration of Combilog that replaces numeric argument positions with argument names. The result is a language where argument names make the notation more readable, yet compositionality is preserved by avoiding variables. Compositionality is demonstrated by implementing CombInduce with CNP as the target language, revealing that programs with the same level of recursive complexity can be synthesized in CNP equally well, and establishing the underlying method of synthesis can also work with CNP. Our evaluations of the user-friendliness of both representations are supported by a range of methods from Information Visualization, Cognitive Modelling, and Human-Computer Interaction. The increased usability of both representations are confirmed by empirical user studies: an often neglected aspect of language design.

[ES] El proceso de ciencia de datos es esencial para extraer valor de los datos. Sin embargo, la parte más tediosa del proceso, la preparación de los datos, implica una serie de formateos, limpieza e identificación de problemas que principalmente son tareas manuales. La preparación de datos todavía se resiste a la automatización en parte porque el problema depende en gran medida de la información del dominio, que se convierte en un cuello de botella para los sistemas de última generación a medida que aumenta la diversidad de dominios, formatos y estructuras de los datos. En esta tesis nos enfocamos en generar algoritmos que aprovechen el conocimiento del dominio para la automatización de partes del proceso de preparación de datos. Mostramos la forma en que las técnicas generales de inducción de programas, en lugar de los lenguajes específicos del dominio, se pueden aplicar de manera flexible a problemas donde el conocimiento es importante, mediante el uso dinámico de conocimiento específico del dominio. De manera más general, sostenemos que una combinación de enfoques de aprendizaje dinámicos y basados en conocimiento puede conducir a buenas soluciones. Proponemos varias estrategias para seleccionar o construir automáticamente el conocimiento previo apropiado en varios escenarios de preparación de datos. La idea principal se basa en elegir las mejores primitivas especializadas de acuerdo con el contexto del problema particular a resolver. Abordamos dos escenarios. En el primero, manejamos datos personales (nombres, fechas, teléfonos, etc.) que se presentan en formatos de cadena de texto muy diferentes y deben ser transformados a un formato unificado. El problema es cómo construir una transformación compositiva a partir de un gran conjunto de primitivas en el dominio (por ejemplo, manejar meses, años, días de la semana, etc.). Desarrollamos un sistema (BK-ADAPT) que guía la búsqueda a través del conocimiento previo extrayendo varias meta-características de los ejemplos que caracterizan el dominio de la columna. En el segundo escenario, nos enfrentamos a la transformación de matrices de datos en lenguajes de programación genéricos como R, utilizando como ejemplos una matriz de entrada y algunas celdas de la matriz de salida. También desarrollamos un sistema guiado por una búsqueda basada en árboles (AUTOMAT[R]IX) que usa varias restricciones, probabilidades previas para las primitivas y sugerencias textuales, para aprender eficientemente las transformaciones. Con estos sistemas, mostramos que la combinación de programación inductiva, con la selección dinámica de las primitivas apropiadas a partir del conocimiento previo, es capaz de mejorar los resultados de otras herramientas actuales específicas para la preparación de datos. / [CA] El procés de ciència de dades és essencial per extraure valor de les dades. No obstant això, la part més tediosa del procés, la preparació de les dades, implica una sèrie de transformacions, neteja i identificació de problemes que principalment són tasques manuals. La preparació de dades encara es resisteix a l'automatització en part perquè el problema depén en gran manera de la informació del domini, que es converteix en un coll de botella per als sistemes d'última generació a mesura que augmenta la diversitat de dominis, formats i estructures de les dades. En aquesta tesi ens enfoquem a generar algorismes que aprofiten el coneixement del domini per a l'automatització de parts del procés de preparació de dades. Mostrem la forma en què les tècniques generals d'inducció de programes, en lloc dels llenguatges específics del domini, es poden aplicar de manera flexible a problemes on el coneixement és important, mitjançant l'ús dinàmic de coneixement específic del domini. De manera més general, sostenim que una combinació d'enfocaments d'aprenentatge dinàmics i basats en coneixement pot conduir a les bones solucions. Proposem diverses estratègies per seleccionar o construir automàticament el coneixement previ apropiat en diversos escenaris de preparació de dades. La idea principal es basa a triar les millors primitives especialitzades d'acord amb el context del problema particular a resoldre. Abordem dos escenaris. En el primer, manegem dades personals (noms, dates, telèfons, etc.) que es presenten en formats de cadena de text molt diferents i han de ser transformats a un format unificat. El problema és com construir una transformació compositiva a partir d'un gran conjunt de primitives en el domini (per exemple, manejar mesos, anys, dies de la setmana, etc.). Desenvolupem un sistema (BK-ADAPT) que guia la cerca a través del coneixement previ extraient diverses meta-característiques dels exemples que caracteritzen el domini de la columna. En el segon escenari, ens enfrontem a la transformació de matrius de dades en llenguatges de programació genèrics com a R, utilitzant com a exemples una matriu d'entrada i algunes dades de la matriu d'eixida. També desenvolupem un sistema guiat per una cerca basada en arbres (AUTOMAT[R]IX) que usa diverses restriccions, probabilitats prèvies per a les primitives i suggeriments textuals, per aprendre eficientment les transformacions. Amb aquests sistemes, mostrem que la combinació de programació inductiva amb la selecció dinàmica de les primitives apropiades a partir del coneixement previ, és capaç de millorar els resultats d'altres enfocaments de preparació de dades d'última generació i més específics. / [EN] Data science is essential for the extraction of value from data. However, the most tedious part of the process, data wrangling, implies a range of mostly manual formatting, identification and cleansing manipulations. Data wrangling still resists automation partly because the problem strongly depends on domain information, which becomes a bottleneck for state-of-the-art systems as the diversity of domains, formats and structures of the data increases. In this thesis we focus on generating algorithms that take advantage of the domain knowledge for the automation of parts of the data wrangling process. We illustrate the way in which general program induction techniques, instead of domain-specific languages, can be applied flexibly to problems where knowledge is important, through the dynamic use of domain-specific knowledge. More generally, we argue that a combination of knowledge-based and dynamic learning approaches leads to successful solutions. We propose several strategies to automatically select or construct the appropriate background knowledge for several data wrangling scenarios. The key idea is based on choosing the best specialised background primitives according to the context of the particular problem to solve. We address two scenarios. In the first one, we handle personal data (names, dates, telephone numbers, etc.) that are presented in very different string formats and have to be transformed into a unified format. The problem is how to build a compositional transformation from a large set of primitives in the domain (e.g., handling months, years, days of the week, etc.). We develop a system (BK-ADAPT) that guides the search through the background knowledge by extracting several meta-features from the examples characterising the column domain. In the second scenario, we face the transformation of data matrices in generic programming languages such as R, using an input matrix and some cells of the output matrix as examples. We also develop a system guided by a tree-based search (AUTOMAT[R]IX) that uses several constraints, prior primitive probabilities and textual hints to efficiently learn the transformations. With these systems, we show that the combination of inductive programming with the dynamic selection of the appropriate primitives from the background knowledge is able to improve the results of other state-of-the-art and more specific data wrangling approaches. / This research was supported by the Spanish MECD Grant FPU15/03219;and partially by the Spanish MINECO TIN2015-69175-C4-1-R (Lobass) and RTI2018-094403-B-C32-AR (FreeTech) in Spain; and by the ERC Advanced Grant Synthesising Inductive Data Models (Synth) in Belgium. / Contreras Ochando, L. (2020). Towards Data Wrangling Automation through Dynamically-Selected Background Knowledge [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/160724 / TESIS

Le code source offre un domaine d'application passionnant des méthodes d'apprentissage en profondeur, englobant des tâches telles que la synthèse, la réparation et l'analyse de programmes, ainsi que des tâches à l'intersection du code et du langage naturel. Bien que les modèles d’apprentissage profond pour le code, en particulier les grands modèles de langage, aient récemment connu un succès significatif, ils peuvent avoir du mal à se généraliser à du code invisible. Cela peut conduire à des inexactitudes, en particulier lorsque vous travaillez avec des référentiels contenant des logiciels propriétaires ou du code en cours de travail. L'objectif principal de cette thèse est d'exploiter efficacement les signaux utiles du contexte disponible afin d'améliorer les performances des modèles de code d'apprentissage profond pour une tâche donnée. En incorporant ces indices contextuels, les capacités de généralisation du modèle sont amplifiées, fournissant des informations supplémentaires non évidentes à partir de l'entrée d'origine et orientant son attention vers des détails essentiels. De plus, l'utilisation d'indices contextuels facilite l'adaptation aux nouvelles tâches et améliore les performances des tâches existantes en effectuant des prédictions plus contextuelles. Pour y parvenir, nous présentons un cadre général comprenant deux étapes : (a) l'amélioration du contexte, qui implique l'enrichissement de l'entrée avec un contexte de support obtenu grâce à l'identification et à la sélection d'indices contextuels pertinents, et (b) la prédiction à l'aide du contexte amélioré, où nous exploitez le contexte de support combiné aux entrées pour faire des prédictions précises. La thèse présente quatre articles qui proposent diverses approches pour ces étapes. Le premier article divise le problème standard de la programmation par exemples en deux étapes: (a) trouver des programmes qui satisfont des exemples individuels (solutions par exemple) et, (b) combiner ces solutions par exemple en tirant parti de leurs états d'exécution de programme pour trouver un programme qui satisfait tous les exemples donnés. Le deuxième article propose une approche pour sélectionner des informations ciblées à partir du fichier actuel et les utiliser pour adapter le modèle de complétion de code à un contexte local jamais vu précédemment. Le troisième article s'appuie sur le deuxième article en tirant parti des indices contextuels de l'ensemble du répertoire de code à l'aide d'un ensemble de requêtes ({\it prompts}) proposées suggérant l'emplacement et le contenu du contexte particulièrement utile à extraire du répertoire. Nous proposons un cadre pour sélectionner la requête la plus pertinente, qui est ensuite utilisée pour demander à un modèle de langage de code de générer des prédictions pour le reste de la ligne de code suivant un curseur positionné dans un fichier. Le quatrième article prolonge le troisième article en proposant un cadre qui apprend à combiner plusieurs contextes divers à partir du répertoire. Nous montrons que la formation de modèles de language de code plus petits de cette manière fonctionne mieux ou à égalité avec des modèles beaucoup plus grands qui n'utilisent pas le contexte du répertoire de code. / Source code provides an exciting application area of deep learning methods, encompassing tasks like program synthesis, repair, and analysis, as well as tasks at the intersection of code and natural language. Although deep learning models for code, particularly large language models, have recently seen significant success, they can face challenges in generalizing to unseen code. This can lead to inaccuracies especially when working with repositories that contain proprietary software or work-in-progress code. The main focus of this thesis is to effectively harness useful signals from the available context such that it can improve the performance of the deep learning models of code at the given task. By incorporating these contextual cues, the model's generalization capabilities are amplified, providing additional insights not evident from the original input and directing its focus toward essential details. Furthermore, the use of contextual cues aids in adapting to new tasks and boosts performance on existing ones by making more context-aware predictions. To achieve this, we present a general framework comprising two stages: (a) Context Enhancement, which involves enriching the input with support context obtained through the identification and selection of relevant contextual cues, and (b) Prediction using the Enhanced Context, where we leverage the support context combined with the input to make accurate predictions. The thesis presents four articles that propose diverse approaches for these stages. The first article breaks the standard problem of programming by examples into two stages: (a) finding programs that satisfy individual examples (per-example solutions) and, (b) combining these per-example solutions by leveraging their program execution states to find a program that satisfies all given examples. The second article proposes an approach for selecting targeted information from the current file and using it to adapt the code completion model to an unseen, local context. The third article builds upon the second article by leveraging contextual cues from the entire code repository using a set of prompt proposals that govern the location and content of the context that should be taken from the repository. We propose a framework to select the most relevant prompt proposal context which is then used to prompt a large language model of code to generate predictions for the tokens in the rest of the line following the cursor in a file. The fourth article extends the third article by proposing a framework that learns to combine multiple diverse contexts from the repository. We show that training smaller models of code this way performs better or at par with significantly larger models that are not trained with repository context.