scholarly journals Fine-Grained Emotion Detection in Suicide Notes: A Thresholding Approach to Multi-Label Classification

2012 ◽  
Vol 5s1 ◽  
pp. BII.S8966 ◽  
Author(s):  
Kim Luyckx ◽  
Frederik Vaassen ◽  
Claudia Peersman ◽  
Walter Daelemans
Keyword(s):  
Test Sentence ◽  
Svm Classifier ◽  
Emotion Detection ◽  
Fine Grained ◽  
Suicide Notes ◽  
Probability Estimates ◽  
Single Instance ◽  
Emotion Labels

We present a system to automatically identify emotion-carrying sentences in suicide notes and to detect the specific fine-grained emotion conveyed. With this system, we competed in Track 2 of the 2011 Medical NLP Challenge, 14 where the task was to distinguish between fifteen emotion labels, from guilt, sorrow, and hopelessness to hopefulness and happiness. Since a sentence can be annotated with multiple emotions, we designed a thresholding approach that enables assigning multiple labels to a single instance. We rely on the probability estimates returned by an SVM classifier and experimentally set thresholds on these probabilities. Emotion labels are assigned only if their probability exceeds a certain threshold and if the probability of the sentence being emotion-free is low enough. We show the advantages of this thresholding approach by comparing it to a naïve system that assigns only the most probable label to each test sentence, and to a system trained on emotion-carrying sentences only.

The Value of Precision in Probability Assessment

2019 ◽  
pp. 69-94
Author(s):  
Jeffrey A. Friedman
Keyword(s):  
Foreign Policy ◽  
World Politics ◽  
Intelligence Analysis ◽  
Probability Assessment ◽  
Fine Grained ◽  
Probability Estimates ◽  
Military Planning ◽  
Policy Analysts ◽  
Short Time ◽  
Training Effort

This chapter analyzes a database containing nearly one million geopolitical forecasts. These data show that foreign policy analysts are surprisingly effective at estimating subjective probabilities. Fine-grained distinctions in probability estimates convey meaningful information about world politics, not arbitrary detail. By extension, the chapter shows that common qualitative expressions of uncertainty (including expressions currently recommended for use in intelligence analysis and military planning) systematically degrade the value of foreign policy discourse. The ability of foreign policy analysts to achieve “returns to precision” in probability assessment does not appear to depend on easy questions, short time horizons, or special cognitive attributes. Instead, the value of precision in probability assessment appears to be a generalizable skill that foreign policy analysts can cultivate through training, effort, and experience.

Emotion detection in suicide notes

2013 ◽  
Vol 40 (16) ◽  
pp. 6351-6358 ◽  
Author(s):  
Bart Desmet ◽  
Véronique Hoste
Keyword(s):  
Emotion Detection ◽  
Suicide Notes

scholarly journals LASSA: Emotion Detection via Information Fusion

2012 ◽  
Vol 5s1 ◽  
pp. BII.S8949 ◽  
Author(s):  
Ning Yu ◽  
Sandra Kübler ◽  
Joshua Herring ◽  
Yu-Yin Hsu ◽  
Ross Israel ◽  
...  
Keyword(s):  
Information Fusion ◽  
Sentiment Classification ◽  
Emotion Detection ◽  
Linguistic Features ◽  
Future Directions ◽  
Suicide Notes ◽  
Fusion Approach

Due to the complexity of emotions in suicide notes and the subtle nature of sentiments, this study proposes a fusion approach to tackle the challenge of sentiment classification in suicide notes: leveraging WordNet-based lexicons, manually created rules, character-based n-grams, and other linguistic features. Although our results are not satisfying, some valuable lessons are learned and promising future directions are identified.

Storytelling in multiple contexts

2016 ◽  
Vol 7 (2) ◽  
pp. 297-340 ◽  
Author(s):  
K.K. Luke
Keyword(s):  
Conversation Analysis ◽  
Point Of View ◽  
Fine Grained ◽  
Multiple Contexts ◽  
Sequential Organization ◽  
Single Instance ◽  
The Many ◽  
Favourite Topic

Since Sacks’ pioneering work in the 1970s, storytelling has become a favourite topic of research within conversation analysis. Scholars have examined storytelling from the point of view of sequential organization (Jefferson 1978), participation organization (Goodwin 1984), story co-telling (Duranti 1986, Mandelbaum 1987, Lerner 1992), displays of epistemic statuses (Schegloff 1988), and action formation (M. Goodwin 1982, 1990; Mandelbaum 1993; Beach 2000; Beach & Glenn 2011; Wu 2011, 2012). Work has also been done on the management of storytelling in the context of other, concurrent activities (Goodwin 1984, Goodwin & Goodwin 1992, Mandelbaum 2010, Haddington et al. 2014). The aim of this paper is to apply the many insights that researchers have accumulated since Sacks to the analysis and understanding of a single instance of storytelling in a Cantonese conversation. A detailed, step-by-step unpacking of this story will reveal how the contingencies of an interaction, including the interplay of multiple contexts, may leave fine-grained imprints on the shape and character of a story.

scholarly journals Eavesdropping Whilst You’re Shopping: Balancing Personalisation and Privacy in Connected Retail Spaces

2018 ◽  
Author(s):  
Vasilios Mavroudis ◽  
Michael Veale
Keyword(s):  
Emotion Detection ◽  
Fine Grained ◽  
Detection Systems ◽  
The Law ◽  
To Come ◽  
Legal Frameworks ◽  
Technical Framework ◽  
Confidentiality And Privacy ◽  
Technical Measures ◽  
The Way

Physical retailers, who once led the way in tracking with loyalty cards and ‘reverse appends’, now lag begin online competitors. Yet we might be seeing these tables turn, as many increasingly deploy technologies ranging from simple sensors to advanced emotion detection systems, even enabling them to tailor prices and shopping experiences on a per-customer basis. Here, we examine these in-store tracking technologies in the retail context, and evaluate them from both technical and regulatory standpoints. We first introduce the relevant technologies in context, before considering privacy impacts, the current remedies individuals might seek through technology and the law, and those remedies’ limitations. To illustrate challenging tensions in this space we consider the feasibility of technical and legal approaches to both a) the recent ‘Go’ store concept from Amazon which requires fine-grained, multi-modal track- ing to function as a shop; and b) current challenges in opting in or out of increasingly pervasive passive Wi-Fi tracking. The ‘Go’ store presents significant challenges with its legality in Europe significantly unclear and unilateral, technical measures to avoid biometric tracking likely ineffective. In the case of MAC addresses, we see a difficult-to-reconcile clash between privacy-as-confidentiality and privacy-as-control, and suggest a technical framework which might help balance the two. Sig- nificant challenges exist when seeking to balance personalisation with privacy, and researchers must work together, including across the boundaries of preferred privacy definitions, to come up with solutions that draw on both technology and the legal frameworks to provide effective and proportionate protection. Retailers, simultaneously, must ensure that their track- ing is not just legal, but worthy of the trust of concerned data subjects.

scholarly journals Fingerprint Liveness Detection Based on Fine-Grained Feature Fusion for Intelligent Devices

Mathematics ◽  
2020 ◽  
Vol 8 (4) ◽  
pp. 517
Author(s):  
Xinting Li ◽  
Weijin Cheng ◽  
Chengsheng Yuan ◽  
Wei Gu ◽  
Baochen Yang ◽  
...  
Keyword(s):  
Data Analysis ◽  
Feature Fusion ◽  
Texture Feature ◽  
Fusion Rule ◽  
Support Vector ◽  
Svm Classifier ◽  
Fine Grained ◽  
Liveness Detection ◽  
Fingerprint Liveness Detection ◽  
Intelligent Devices

Currently, intelligent devices with fingerprint identification are widely deployed in our daily life. However, they are vulnerable to attack by fake fingerprints made of special materials. To elevate the security of these intelligent devices, many fingerprint liveness detection (FLD) algorithms have been explored. In this paper, we propose a novel detection structure to discriminate genuine or fake fingerprints. First, to describe the subtle differences between them and take advantage of texture descriptors, three types of different fine-grained texture feature extraction algorithms are used. Next, we develop a feature fusion rule, including five operations, to better integrate the above features. Finally, those fused features are fed into a support vector machine (SVM) classifier for subsequent classification. Data analysis on three standard fingerprint datasets indicates that the performance of our method outperforms other FLD methods proposed in recent literature. Moreover, data analysis results of blind materials are also reported.

scholarly journals Hybrid Approaches to Fine-Grained Emotion Detection in Social Media Data

2020 ◽  
Vol 34 (10) ◽  
pp. 13732-13733
Author(s):  
Annika Marie Schoene
Keyword(s):  
Social Media ◽  
Recurrent Neural Networks ◽  
Emotion Detection ◽  
Problem Statement ◽  
Future Directions ◽  
Social Media Data ◽  
Fine Grained ◽  
Hybrid Approaches ◽  
Research Plan ◽  
Media Data

This paper states the challenges in fine-grained target-dependent Sentiment Analysis for social media data using recurrent neural networks. First, the problem statement is outlined and an overview of related work in the area is given. Then a summary of progress and results achieved to date and a research plan and future directions of this work are given.

scholarly journals A Combined Approach to Emotion Detection in Suicide Notes

2012 ◽  
Vol 5s1 ◽  
pp. BII.S8969 ◽  
Author(s):  
Alexander Pak ◽  
Delphine Bernhard ◽  
Patrick Paroubek ◽  
Cyril Grouin
Keyword(s):  
Machine Learning ◽  
The Other ◽  
Combined Approach ◽  
Emotion Detection ◽  
Suicide Notes ◽  
Clinical Records ◽  
Official Evaluation ◽  
F Measure

In this paper, we present the system we have developed for participating in the second task of the i2b2/VA 2011 challenge dedicated to emotion detection in clinical records. On the official evaluation, we ranked 6th out of 26 participants. Our best configuration, based upon a combination of both a machine-learning based approach and manually-defined transducers, obtained a 0.5383 global F-measure, while the distribution of the other 26 participants’ results is characterized by mean = 0.4875, stdev = 0.0742, min = 0.2967, max = 0.6139, and median = 0.5027. Combination of machine learning and transducer is achieved by computing the union of results from both approaches, each using a hierarchy of sentiment specific classifiers.

Sign in / Sign up

Export Citation Format

Share Document