Record matching is an essential step in duplicate detection as it identifies records representing same real-world entity. Supervised record matching methods require users to provide training data and therefore cannot be applied for web databases where query results are generated on-the-fly. To overcome the problem, a new record matching method named Unsupervised Duplicate Elimination (UDE) is p...

Annually, web search engine providers spend more and more money on documents ranking in search engines result pages (SERP). Click models provide advantageous information for ranking documents in SERPs through modeling interactions among users and search engines. Here, three modules are employed to create a hybrid click model; the first module is a PGM-based click model, the second module in a d...

Supervised learning algorithms have been widely applied in tracking-by-detection based methods for object tracking in recent years. Most of these approaches treat tracking as a classification problem and solve it by training a discriminative classifier and exhaustively evaluating every possible target position; problems thus exist for two reasons. First, since the classifier describes the commo...

Recently deep neural networks have shown many successful applications in different domains. For this lesion segmentation task, we utilize a deep convolutional neural network with 5 layers in a sliding window fashion to create a voxel-based classifier. We evaluate our system with Dice similarity, misclassification rate and area under the ROC curve. Based on experimental results our proposed CAD ...

This research describes the COD (Classifier Output Difference) distance metric for finding similarity between hypotheses and learning algorithms. This metric is a tool which can be used to measure how similar two hypotheses are. It goes beyond simple accuracy comparisons and provides insights about fundamental differences between learning models. This paper describes how COD works and shows how...

Electronic health records (EHRs) contain important clinical information about patients. Efficient and effective use of this information could supplement or even replace manual chart review as a means of studying and improving the quality and safety of healthcare delivery. However, some of these clinical data are in the form of free text and require pre-processing before use in automated systems...

This paper describes AUEB’s participation in TAC 2009. Specifically, we participated in the textual entailment recognition track for which we used string similarity measures applied to shallow abstractions of the input sentences, and a Maximum Entropy classifier to learn how to combine the resulting features. We also exploited WordNet to detect synonyms and a dependency parser to measure simila...

Lots of work has been done on speech and speaker recognition. Many technologies were developed for the analysis of speech waveforms. Musical instrument recognition is an important aspect of music information retrieval system. In this paper we analyzed features for musical instruments recognition and a brief study on music similarity. Music similarity search is done by using mel-frequency cepstr...

This paper describes our approach to the SemEval-2017 shared task of determining question-question similarity in a community question-answering setting (Task 3B). We extracted both syntactic and semantic similarity features between candidate questions, performed pairwise-preference learning to optimize for ranking order, and then trained a random forest classifier to predict whether the candida...

We develop a machine-learned similarity metric for Windows failure reports using telemetry data gathered from clients describing the failures. The key feature is a tuned callstack edit distance with learned costs for seven fundamental edits based on callstack frames. We present results of a failure similarity classifier based on this and other features. We also describe how the model can be dep...