Николенко Сергей Игоревич

Neural Click Models for Recommender Systems

2024 · CHAPTER · en

User response modeling in recommender systems: a survey

2024 · ARTICLE · en

Over the last several decades, recommender systems have become an integral part of both our daily lives and the research frontier at machine learning. In this survey, we explore various approaches to developing simulators for recommendation systems, especially for modeling the user response function. We consider simple probabilistic models, approaches based on generative adversarial networks, and full-scale simulators, and also review the datasets available for the research community.

Поиск редких данных в задаче распознавания лиц на изображениях

2022 · ARTICLE · ru

Одной из основных проблем современных нейросетевых дескрипторов в задаче идентификации лиц является малое число обучающих примеров определенного типа: изображения плохого качества, разный масштаб или освещение, лица детей, пожилых людей, редкие расы. В результате точность распознавания оказывается низкой для входных изображений, не похожих на большинство изображений в наборе данных, используемом для настройки метода извлечения признаков. В работе предлагается способ преодоления такой проблемы за счет автоматического обнаружения нетипичных входных изображений на основе введения предварительного этапа их автоматической отбраковки. Для этого используется специальная свёрточная сеть, обученная на наборе редких данных, которые обрабатывались с помощью известных алгоритмов преобразования изображений. Для повышения вычислительной эффективности решение о наличии редкого изображения принимается на основе того же дескриптора лица, который используется в классификаторе. Экспериментальное исследование подтвердило преимущества в точности предложенного подхода для нескольких наборов данных лиц и современных нейросетевых дескрипторов

Cross-Domain Limitations of Neural Models on Biomedical Relation Classification

2022 · ARTICLE · en

Relation extraction (RE) aims to extract relational facts from plain text, which is essential to the biomedical research field with the rapid growth of biomedical literature and generally large volumes of biomedicine-related text coming from various sources. Numerous annotated corpora and state-of-the-art models have been introduced in the past five years. However, there are no general guidelines about evaluating models on these corpora in single- and cross-domain settings with diverse entities and relation types. We aim to fill this gap for the task of detecting whether a relation holds between two biomedical entities given a text span. In this work, we present a fine-grained evaluation intended to perform a comparative evaluation of four biomedical benchmarks and understand the efficiency of state-of-the-art neural architectures based on Long Short-Term Memory (LSTM) with cross-attention and Bidirectional Encoder Representations from Transformers (BERT) for relation extraction across two main domains, namely scientific abstracts and electronic health records. We present a comparative evaluation of biomedical RE datasets, including the PHAEDRA, i2b2/VA, BC5CDR, and MADE corpora. Our evaluation of BioBERT and LSTM for binary classification shows significant divergence in in-domain and out-of-domain performance, finding an average drop in F1-measure of 34.2% for BioBERT. The cross-attention LSTM model developed in this work exhibits better cross-domain performance, with a drop of only 27.6% in F-measure. © 2013 IEEE.

Near-Zero-Shot Suggestion Mining with a Little Help from WordNet

2022 · CHAPTER · en

In this work we explore the constructive side of online reviews: advice, tips, requests, and suggestions that users provide about goods, venues and other items of interest. To reduce training costs and annotation efforts needed to build a classifier for a specific label set, we present and evaluate several entailment-based zero-shot approaches to suggestion classification in a label-fully-unseen fashion. In particular, we introduce the strategy of assigning target class labels to sentences with user intentions, which significantly improves prediction quality. The proposed strategies are evaluated with a comprehensive experimental study that validated our results both quantitatively and qualitatively. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

The Russian Drug Reaction Corpus and Neural Models for Drug Reactions and Effectiveness Detection in User Reviews

2021 · ARTICLE · en

Drugs and diseases play a central role in many areas of biomedical research and healthcare. Aggregating knowledge about these entities across a broader range of domains and languages is critical for information extraction (IE) applications. To facilitate text mining methods for analysis and comparison of patient’s health conditions and adverse drug reactions reported on the Internet with traditional sources such as drug labels, we present a new corpus of Russian language health reviews. The Russian Drug Reaction Corpus (RuDReC) is a new partially annotated corpus of consumer reviews in Russian about pharmaceutical products for the detection of health-related named entities and the effectiveness of pharmaceutical products. The corpus itself consists of two parts, the raw one and the labeled one. The raw part includes 1.4 million health-related user-generated texts collected from various Internet sources, including social media. The labeled part contains 500 consumer reviews about drug therapy with drug- and disease-related information. Labels for sentences include health-related issues or their absence. The sentences with one are additionally labeled at the expression level for identification of fine-grained subtypes such as drug classes and drug forms, drug indications and drug reactions. Further, we present a baseline model for named entity recognition (NER) and multilabel sentence classification tasks on this corpus. The macro F1 score of 74.85% in the NER task was achieved by our RuDR-BERT model. For the sentence classification task, our model achieves the macro F1 score of 68.82% gaining 7.47% over the score of BERT model trained on Russian data.

Formalization and taxonomy of compute-aggregate problems for cloud computing applications

2021 · ARTICLE · en

Efficient representation of data aggregations is a fundamental problem in modern big data applications, where network topologies and deployed routing and transport mechanisms play a fundamental role in optimizing desired objectives such as cost, latency, and others. In traditional networking, applications use TCP and UDP transports as a primary interface for implemented applications that hide the underlying network topology from end systems. On the flip side, to exploit network infrastructure in a better way, applications restore characteristics of the underlying network. In this work, we demonstrate that both specified extreme cases can be inefficient to optimize given objectives. We study the design principles of routing and transport infrastructure and identify extra information that can be used to improve implementations of compute-aggregate tasks. We build a taxonomy of compute-aggregate services unifying aggregation design principles, propose algorithms for each class, analyze them theoretically, and support our results with an extensive experimental study.

ColocML: machine learning quantifies co-localization between mass spectrometry images

2020 · ARTICLE · en

Motivation Imaging mass spectrometry (imaging MS) is a prominent technique for capturing distributions of molecules in tissue sections. Various computational methods for imaging MS rely on quantifying spatial correlations between ion images, referred to as co-localization. However, no comprehensive evaluation of co-localization measures has ever been performed; this leads to arbitrary choices and hinders method development. Results We present ColocML, a machine learning approach addressing this gap. With the help of 42 imaging MS experts from nine laboratories, we created a gold standard of 2210 pairs of ion images ranked by their co-localization. We evaluated existing co-localization measures and developed novel measures using term frequency–inverse document frequency and deep neural networks. The semi-supervised deep learning Pi model and the cosine score applied after median thresholding performed the best (Spearman 0.797 and 0.794 with expert rankings, respectively). We illustrate these measures by inferring co-localization properties of 10 273 molecules from 3685 public METASPACE datasets.

RecVAE: A new variational autoencoder for top-n recommendations with implicit feedback

2020 · CHAPTER · en

Towards Software-Defined Buffer Management

2020 · ARTICLE · en

Buffering architectures and policies for their efficient management are core ingredients of a network architecture. However, despite strong incentives to experiment with and deploy new policies, opportunities for changing anything beyond minor elements are limited. We introduce a new specification language, OpenQueue, that allows to express virtual buffering architectures and management policies representing a wide variety of economic models. OpenQueue allows users to specify entire buffering architectures and policies conveniently through several comparators and simple functions. We show examples of buffer management policies in OpenQueue and empirically demonstrate its impact on performance in various settings.