Exploring the Impact of Single-Character Attacks in Federated Learning Language Classification: Introducing the Novel Single-Character Strike

Federated learning (FL) is a privacy preserving machine learning approach which allows a machine learning model to be trained in a distributed fashion without ever sharing user data. Due to the large amount of valuable text and voice data stored on end-user devices, this approach works particularly well for natural language processing (NLP)...

The Relation Between Words and Definitions in Combined Word-Sentence Space

Methods for learning vector space representations of words have yielded spaces which contain semantic and syntactic regularities. These regularities mean that vector arithmetic operations in the latent space represent meaningful and interpretable relations between words. These word vectors have been so successful in capturing such relations, as...

Predictive Maintenance for Aircraft Systems: Using textual elements as covariates

Unplanned maintenance is a costly factor in aircraft operations. Predictive Maintenance aims at reducing the surprise effect of unplanned maintenance and thereby its associated cost. A variety of statistical models are used to estimate the remaining life, as well as sensors to gauge component condition. The application to statistical models of...

Estimating popularity by sentiment and polarization classification on social media

Mass processing of social media posts has been brought to scientists' attention during the last decade. The massive growth of online social networks, like Twitter and Facebook, have created a need for determining peoples' opinions and moods through these means. This thesis constitutes a research on measuring users' sentiment upon a particular...

Speech-based automatic closed caption alignment

In the Netherlands, four million people watch television programs with closed captions because they are hearing impaired or non-native speakers. Closed captions contain Dutch speech transcriptions and non-speech sound descriptions and are displayed as subtitles. Due to government obligation, the number of television programs that must be closed...

Reasoning about Emotions: An affective natural language processing environment, using lexical relations to measure activation and evaluation, and extracting semantics from natural language

In the field of textual affect sensing many methods have been proposed. These methods vary from keyword spotting techniques, lexical affinity, statistical natural language processing and hand-crafted models. Based on a large scale survey, two profounding theories have been selected for investigation. The first is the proposed work of (Kamps &...