Andrey Kravchenko

University of Oxford | OX · Department of Computer Science

Content-intensive web sites, such as Google or Amazon, paginate their results to accommodate limited screen sizes. Thus, human users and automatic tools alike have to traverse the pagination links when they crawl the site, extract data, or automate common tasks, where these applications require access to the entire result set. Previous approaches,...

Web blocks such as navigation menus, advertisements, and headers and footers are key components of web pages which define not only the appearance of a web page but also the way in which humans interact with different parts of the page. For machines, however, classifying and interacting with these blocks is a surprisingly hard task. Yet, web block c...

The development of state-of-the-art systems in different applied areas of machine learning (ML) is driven by benchmarks, which have shaped the paradigm of evaluating generalisation capabilities from multiple perspectives. Although the paradigm is shifting towards more fine-grained evaluation across diverse tasks, the delicate question of how to agg...

Transformer models play a crucial role in state of the art solutions to problems arising in the field of natural language processing (NLP). They have billions of parameters and are typically considered as black boxes. Robustness of huge Transformer-based models for NLP is an important question due to their wide adoption. One way to understand and i...

Following the recent successful examples of large technology companies, many modern enterprises seek to build Knowledge Graphs to provide a unified view of corporate knowledge, and to draw deep insights using machine learning and logical reasoning. There is currently a perceived disconnect between the traditional approaches for data science, typica...

Robustness of huge Transformer-based models for natural language processing is an important issue due to their capabilities and wide adoption. One way to understand and improve robustness of these models is an exploration of an adversarial attack scenario: check if a small perturbation of an input can fool a model. Due to the discrete nature of tex...

Deep learning models suffer from a phenomenon called adversarial attacks: we can apply minor changes to the model input to fool a classifier for a particular example. The literature mostly considers adversarial attacks on models with images and other structured inputs. However, the adversarial attacks for categorical sequences can also be harmful....

An adversarial attack paradigm explores various scenarios for the vulnerability of deep learning models: minor changes of the input can force a model failure. Most of the state of the art frameworks focus on adversarial attacks for images and other structured model inputs, but not for categorical sequences models. Successful attacks on classifiers...

An adversarial attack paradigm explores various scenarios for vulnerability of machine and especially deep learning models: we can apply minor changes to the model input to force a classifier's failure for a particular example. Most of the state of the art frameworks focus on adversarial attacks for images and other structured model inputs. The adv...

Constructing high-quality generative models for 3D shapes is a fundamental task in computer vision with diverse applications in geometry processing, engineering, and design. Despite the recent progress in deep generative modelling, synthesis of finely detailed 3D surfaces, such as high-resolution point clouds, from scratch has not been achieved wit...

Web blocks are ubiquitous across the Web. Navigation menus, advertisements, headers, footers, and sidebars can be found almost on any website. Identifying these blocks can be of significant importance for tasks such as wrapper induction, assistance to visually impaired people, Web page topic clustering, and Web search among a few. There have been s...

Web blocks such as navigation menus, advertisements, headers, and footers are key components of Web pages that define not only the appearance, but also the way humans interact with different parts of the page. For machines, however, classifying and interacting with these blocks is a surprisingly hard task. Yet, Web block classification has varied a...

Following the recent successful examples of large technology companies, many modern enterprises seek to build knowledge graphs to provide a unified view of corporate knowledge and to draw deep insights using machine learning and logical reasoning. There is currently a perceived disconnect between the traditional approaches for data science, typical...

The web contains a huge amount of data, which can be primarily accessed with the use of web data extraction technology. With increasing complexity of the web development stack and the source code, a web page visual representation rendered by the browser is often the only source reflecting the semantics, functional role, and logical structure of ele...

What if you could turn all websites of an entire domain into a single database? Imagine all real estate offers, all airline flights, or all your local restaurants' menus automatically collected from hundreds or thousands of agencies, travel agencies, or restaurants, presented as a single homogeneous dataset. Historically, this has required tremendo...

Search engines are the sinews of the web. These sinews have become strained, however: Where the web's function once was a mix of library and yellow pages, it has become the central marketplace for information of almost any kind. We search more and more for objects with specific characteristics, a car with a certain mileage, an affordable apartment...

RNA secondary structures play a vital role in modern genetics and a lot of time and eort has been put into their study. It is important to be able to predict them with high accuracy, since methods in-volving manual analysis are expensive, time-consuming and error-prone. Predictions can also be used to guide experiments to reduce time and money requ...

Transitive reduction removes edges from a graph that are implied by transitivity. Computing the minimal transitive reduction of a graph is useful for understanding its structure. It has applications when visualising graphs in software engineering, social networks, databases, web design, etc. Whilst the problem has been studied in theory for a long...