Reuters Corpus Volume I (RCV1) is an archive of over 800,000 manually categorized newswire stories recently made available by Reuters, Ltd. for research purposes. Use of this data for research on text categorization requires a detailed understanding of the real world constraints under which the data was produced. Drawing on interviews with Reuters personnel and access to Reuters documentation, we describe the coding policy and quality control procedures used in producing the RCV1 data, the intended semantics of the hierarchical category taxonomies, and the corrections necessary to remove errorful data. We refer to the original data as RCV1-v1, and the corrected data as RCV1-v2. We benchmark several widely used supervised learning methods on RCV1-v2, illustrating the collection’s properties, suggesting new directions for research, and providing baseline results for future studies. We make available detailed, per-category experimental results, as well as

Computing semantic relatedness of natural language texts requires access to vast amounts of common-sense and domain-specific world knowledge. We propose Explicit Semantic Analysis (ESA), a novel method that represents the meaning of texts in a high-dimensional space of concepts derived from Wikipedia. We use machine learning techniques to explicitly represent the meaning of any text as a weighted vector of Wikipedia-based concepts. Assessing the relatedness of texts in this space amounts to comparing the corresponding vectors using conventional metrics (e.g., cosine). Compared with the previous state of the art, using ESA results in substantial improvements in correlation of computed relatedness scores with human judgments: from r =0.56 to 0.75 for individual words and from r =0.60 to 0.72 for texts. Importantly, due to the use of natural concepts, the ESA model is easy to explain to human users. 1

The aim of this paper is to address recognition of natural human actions in diverse and realistic video settings. This challenging but important subject has mostly been ignored in the past due to several problems one of which is the lack of realistic and annotated video datasets. Our first contribution is to address this limitation and to investigate the use of movie scripts for automatic annotation of human actions in videos. We evaluate alternative methods for action retrieval from scripts and show benefits of a text-based classifier. Using the retrieved action samples for visual learning, we next turn to the problem of action classification in video. We present a new method for video classification that builds upon and extends several recent ideas including local space-time features, space-time pyramids and multichannel non-linear SVMs. The method is shown to improve state-of-the-art results on the standard KTH action dataset by achieving 91.8 % accuracy. Given the inherent problem of noisy labels in automatic annotation, we particularly investigate and show high tolerance of our method to annotation errors in the training set. We finally apply the method to learning and classifying challenging action classes in movies and show promising results. 1.

Open source development projects typically support an open bug repository to which both developers and users can report bugs. The reports that appear in this repository must be triaged to determine if the report is one which requires attention and if it is, which developer will be assigned the responsibility of resolving the report. Large open source developments are burdened by the rate at which new bug reports appear in the bug repository. In this paper, we present a semi-automated approach intended to ease one part of this process, the assignment of reports to a developer. Our approach applies a machine learning algorithm to the open bug repository to learn the kinds of reports each developer resolves. When a new report arrives, the classifier produced by the machine learning technique suggests a small number of developers suitable to resolve the report. With this approach, we have reached precision levels of 57 % and 64 % on the Eclipse and Firefox development projects respectively. We have also applied our approach to the gcc open source development with less positive results. We describe the conditions under which the approach is applicable and also report on the lessons we learned about applying machine learning to repositories used in open source development.

The past decade has seen a tremendous growth in the amount of experimental and computational biomedical data, specifically in the areas of Genomics and Proteomics. This growth is accompanied by an accelerated increase in the number of biomedical publications discussing the findings. In the last few years there is a lot of interest within the scientific community in literature-mining tools to help sort through this abundance of literature, and find the nuggets of information most relevant and useful for specific analysis tasks. This paper

We enhance machine learning algorithms for text categorization with generated features based on domain-specific and common-sense knowledge. This knowledge is represented using publicly available ontologies that contain hundreds of thousands of concepts, such as the Open Directory; these ontologies are further enriched by several orders of magnitude through controlled Web crawling. Prior to text categorization, a feature generator analyzes the documents and maps them onto appropriate ontology concepts, which in turn induce a set of generated features that augment the standard bag of words. Feature generation is accomplished through contextual analysis of document text, implicitly performing word sense disambiguation. Coupled with the ability to generalize concepts using the ontology, this approach addresses the two main problems of natural language processing—synonymy and polysemy. Categorizing documents with the aid of knowledge-based features leverages information that cannot be deduced from the documents alone. Experimental results confirm improved performance, breaking through the plateau previously reached in the field. 1

Text document clustering plays an important role in providing intuitive navigation and browsing mechanisms by organizing large amounts of information into a small number of meaningful clusters. The bag of words representation used for these clustering methods is often unsatisfactory as it ignores relationships between important terms that do not co-occur literally. In order to deal with the problem, we integrate background knowledge --- in our application Wordnet --- into the process of clustering text documents.

This paper provides background on record linkage methods that can be used in combining data from a variety of sources such as person lists business lists. It also gives some areas of current research.

We study an approach to text categorization that combines distributional clustering of words and a Support Vector Machine (SVM) classifier. This word-cluster representation is computed using the recently introduced Information Bottleneck method, which generates a compact and efficient representation of documents. When combined with the classification power of the SVM, this method yields high performance in text categorization. This novel combination of SVM with word-cluster representation is compared with SVM-based categorization using the simpler bag-of-words (BOW) representation. The comparison is performed over three known datasets. On one of these datasets (the 20 Newsgroups) the method based on word clusters significantly outperforms the word-based representation in terms of categorization accuracy or representation efficiency. On the two other sets (Reuters-21578 and WebKB) the word-based representation slightly outperforms the word-cluster representation. We investigate the potential reasons for this behavior and relate it to structural differences between the datasets.

ROC and DET curves are often used in the field of person authentication to assess the quality of a model or even to compare several models. We argue in this paper that this measure can be misleading as it compares performance measures that cannot be reached simultaneously by all systems. We propose instead new curves, called Expected Performance Curves (EPC). These curves enable the comparison between several systems according to a criterion, decided by the application, which is used to set thresholds according to a separate validation set. A free sofware is available to compute these curves. A real case study is used throughout the paper to illustrate it. Finally, note that while this study was done on an authentication problem, it also applies to most 2-class classification tasks.