Topic
Annotation
About: Annotation is a research topic. Over the lifetime, 6719 publications have been published within this topic receiving 203463 citations. The topic is also known as: note & markup.
Papers published on a yearly basis
Papers
More filters
••
01 Jan 2017
TL;DR: The paper describes the development of a corpus from social media built with the aim of representing and analysing hate speech against some minority groups in Italy and introduces the issues related to data collection and annotation.
Abstract: English. The paper describes the development of a corpus from social media built with the aim of representing and analysing hate speech against some minority groups in Italy. The issues related to data collection and annotation are introduced, focusing on the challenges we addressed in designing a multifaceted set of labels where the main features of verbal hate expressions may be modelled. Moreover, an analysis of the disagreement among the annotators is presented in order to carry out a preliminary evaluation of the data set and the scheme. Italiano. L’articolo descrive un corpus di testi estratti da social media costruito con il principale obiettivo di rappresentare ed analizzare il fenomeno dell’hate speech rivolto contro i migranti in Italia. Vengono introdotti gli aspetti significativi della raccolta ed annotazione dei dati, richiamando l’attenzione sulle sfide affrontate per progettare un insieme di etichette che rifletta le molte sfaccettature necessarie a cogliere e modellare le caratteristiche delle espressioni di odio. Inoltre viene presentata un’analisi del disagreement tra gli annotatori allo scopo di tentare una preliminare valutazione del corpus e dello schema di annotazione stesso.
57 citations
•
TL;DR: This paper proposed an active learning-based strategy, called CEREALS, in which a human only has to hand-label a few, automatically selected, regions within an unlabeled image corpus.
Abstract: State of the art methods for semantic image segmentation are trained in a supervised fashion using a large corpus of fully labeled training images. However, gathering such a corpus is expensive, due to human annotation effort, in contrast to gathering unlabeled data. We propose an active learning-based strategy, called CEREALS, in which a human only has to hand-label a few, automatically selected, regions within an unlabeled image corpus. This minimizes human annotation effort while maximizing the performance of a semantic image segmentation method. The automatic selection procedure is achieved by: a) using a suitable information measure combined with an estimate about human annotation effort, which is inferred from a learned cost model, and b) exploiting the spatial coherency of an image. The performance of CEREALS is demonstrated on Cityscapes, where we are able to reduce the annotation effort to 17%, while keeping 95% of the mean Intersection over Union (mIoU) of a model that was trained with the fully annotated training set of Cityscapes.
57 citations
01 Jan 2008
TL;DR: An extensible trajectory annotation model, which is oriented on the notion of episodes and allows a clear separation of semantic and physical trajectory information is introduced and a program to support trajectory annotation independent of the recorded location is developed.
Abstract: The analysis of mobile behavior is elementary for applications such as city planning, optimization of public transport or mobile communications. However, many applications in the mobility domain require a semantic interpretation of movement information. While physical trajectories can readily be recorded using Global Positioning System (GPS) technology, the semantic interpretation of the data is still a great research challenge. However, GPS trajectories can be utilized to facilitate the manual semantic annotation process, and thus render tedious interviews and manual mobility records unnecessary. In order to realize this schema in practice, two requirements are necessary. First, a conceptual annotation model, and second a graphical user interface which takes advantage of the rich information hidden in physical trajectory data. In this paper, we introduce an extensible trajectory annotation model, which is oriented on the notion of episodes and allows a clear separation of semantic and physical trajectory information. We implemented our model and developed a program to support trajectory annotation independent of the recorded location. We present our program architecture and various features that facilitate the annotation process. 1 MOTIVATION
57 citations
•
IBM1
TL;DR: In this article, the authors present a method for teaching a user to implement an application in a web-based environment on a computer system, which includes providing predetermined applications and presenting an annotation page that includes one or more annotations describing a predetermined application.
Abstract: A method performed in a web-based environment on a computer system teaches a user to implement an application. The method includes providing predetermined applications and presenting an annotation page that includes one or more annotations descriptive of a predetermined application. Each annotation includes keyword links, annotation links, and detail of implementation of the application. The method includes permitting the user to select a link in an annotation. If the user selects a keyword link, reference documentation associated with that keyword is presented. If the user selects an annotation link, another annotation descriptive of another source file of a predetermined application is presented.
57 citations
•
IBM1
TL;DR: In this paper, a method for document annotation includes providing a contextual framework extending over multiple electronic documents, where annotations are input to a computer by a user with respect to two or more of the documents, while the user views the documents on a computer screen.
Abstract: A method for document annotation includes providing a contextual framework extending over multiple electronic documents. Annotations are input to a computer by a user with respect to two or more of the documents, while the user views the documents on a computer screen. Each of the annotations is stored together with respective context information in a memory of the computer, the context information including multiple fields defined by the framework. The annotations are sorted with respect to the two or more documents according to one or more of the multiple fields of the respective context information, and the sorted annotations are output for viewing by the user.
57 citations