This article takes a design-driven human-centred approach to Facial Recognition Technology (FRT). In a process of Research through Design, we first generated 120 ways to dodge face recognition, then distilled and tested 50 concepts in the lab. The 19 disguises that successfully bypasses FRT informed the implementation of seven disguises initially tested with 14 white participants walking through a hall, a corridor and a control gate. The control gate led to a larger study (39 participants of different ethnicities) to assess the effectiveness of the disguises in bypassing three open source FR models using three different distance metrics and four backends. We compare our real-life evaluation of design-generated disguises against previous and current computing research: while maliciously crafted digital perturbation attacks work well, they do not capture the complexity of live FRT opening up opportunities for future research.

Human activity recognition (HAR) using smartphone sensors have been recently studied in various applications including healthcare, fitness, and smart home. Their recognition accuracy often depends on high-quality feature design and effectiveness of classification algorithms, where existing work mostly replies on laborious hand-crafted design and shallow feature learning architecture. Recent deep learning techniques demonstrate outstanding effectiveness in performing automatic feature learning and outperform traditional models in terms of accuracy. But their performance is limited by the quality and volumes of available labelled data. It is challenging to achieve accurate multisubject HAR with only smartphone sensing data. This article proposes a novel optimal activity graph generation model incorporating a deep learning framework for automatic and accurate HAR with multiple subjects using only acceleration and gyroscope data. The activity graph generation model presents a multisensory integration mechanism with three-steps sorting algorithms for producing optimal activity graphs containing alignments of neighbored signals in their width and height. Then, we propose a deep convolutional neural network to automatically learn distinguishable features from the graphs for HAR. By leveraging superior presentation of correlations between human activities and neighbored signals alignments via optimal activity graphs, the learned features are endowed with more discriminative power. The experimental evaluation was carried out on several benchmark datasets (i.e., UCI, USCHAD, and UTD-MHAD). The results showed that our approach improved the average recognition accuracy by about 5% when compared with other state-of-the-art HAR methods. Particularly towards multisubject HAR cases (UTD-MHAD dataset with 21 subjects), it achieved up to 10% accuracy gain over other methods. These improvements show the advantage and potential of our method dealing with complex HAR problems with multiple subjects using limited sensing data.

Maintaining smartphone indoor localization systems operational and accurate is challenging. Landmark and WiFi based systems rely on up-to-date reference points and radio maps. These are expensive to refresh for most current systems because of the exhaustive site surveys these rely on. Any physical change in the environment may cause unpredictable transformation in the radio propagation medium or it may affect the typical navigation paths inside the building. Active Learning (AL) involves human intervention in annotating sensor samples collected from the target environment. However, the burden on users to perform this manual task should be minimized to make the approach more user-friendly. We propose a solution that improves AL by involving the users only when sensor data deviates from what is expected at a location. Our solution analyses the confidence of location estimations in the environment. We propose a robust data distribution shift detector that looks at several levels in our location estimation neural network to determine the confidence of predictions. During the training stage, the model builds high confidence for estimations in the areas where data is available. We show that data from outside the training distribution is detected as anomaly, which triggers the AL to prompt the user for her current location as label. These annotated samples and other augmented data are then used to update the location estimation model.

Early rumor detection (ERD) on social media platform is very challenging when limited, incomplete and noisy information is available. Most of the existing methods have largely worked on event-level detection that requires the collection of posts relevant to a specific event and relied only on user-generated content. They are not appropriate to detect rumor sources in the very early stages, before an event unfolds and becomes widespread. In this paper, we address the task of ERD at the message level. We present a novel hybrid neural network architecture, which combines a task-specific character-based bidirectional language model and stacked Long Short-Term Memory (LSTM) networks to represent textual contents and social-temporal contexts of input source tweets, for modelling propagation patterns of rumors in the early stages of their development. We apply multi-layered attention models to jointly learn attentive context embeddings over multiple context inputs. Our experiments employ a stringent leave-one-out cross-validation (LOO-CV) evaluation setup on seven publicly available real-life rumor event data sets. Our models achieve state-of-the-art(SoA) performance for detecting unseen rumors on large augmented data which covers more than 12 events and 2,967 rumors. An ablation study is conducted to understand the relative contribution of each component of our proposed model.

The scarcity and class imbalance of training data are known issues in current rumor detection tasks. We propose a straight-forward and general-purpose data augmentation technique which is beneficial to early rumor detection relying on event propagation patterns. The key idea is to exploit massive unlabeled event data sets on social media to augment limited labeled rumor source tweets. This work is based on rumor spreading patterns revealed by recent rumor studies and semantic relatedness between labeled and unlabeled data. A state-of-the-art neural language model (NLM) and large credibility-focused Twitter corpora are employed to learn context-sensitive representations of rumor tweets. Six different real-world events based on three publicly available rumor datasets are employed in our experiments to provide a comparative evaluation of the effectiveness of the method. The results show that our method can expand the size of an existing rumor data set nearly by 200% and corresponding social context (i.e., conversational threads) by 100% with reasonable quality. Preliminary experiments with a state-of-the-art deep learning-based rumor detection model show that augmented data can alleviate over-fitting and class imbalance caused by limited train data and can help to train complex neural networks (NNs). With augmented data, the performance of rumor detection can be improved by 12.1% in terms of F-score. Our experiments also indicate that augmented training data can help to generalize rumor detection models on unseen rumors.

Automatic Term Extraction (ATE) deals with the extraction of terminology from a domain specific corpus, and has long been an established research area in data and knowledge acquisition. ATE remains a challenging task as it is known that there is no existing ATE methods that can consistently outperform others in any domain. This work adopts a refreshed perspective to this problem: instead of searching for such a ‘one-size-fit-all’ solution that may never exist, we propose to develop generic methods to ‘enhance’ existing ATE methods. We introduce SemRe-Rank, the first method based on this principle, to incorporate semantic relatedness—an often overlooked venue—into an existing ATE method to further improve its performance. SemRe-Rank incorporates word embeddings into a personalised PageRank process to compute ‘semantic importance’ scores for candidate terms from a graph of semantically related words (nodes), which are then used to revise the scores of candidate terms computed by a base ATE algorithm. Extensively evaluated with 13 state-of-the-art base ATE methods on four datasets of diverse nature, it is shown to have achieved widespread improvement over all base methods and across all datasets, with up to 15 percentage points when measured by the Precision in the top ranked K candidate terms (the average for a set of K ’s), or up to 28 percentage points in F1 measured at a K that equals to the expected real terms in the candidates (F1 in short). Compared to an alternative approach built on the well-known TextRank algorithm, SemRe-Rank can potentially outperform by up to 8 points in Precision at top K , or up to 17 points in F1.

This work explores the usage of Linked Data for Web scale Information Extraction and shows encouraging results on the task of Wrapper Induction. We propose a simple knowledge based method which is (i) highly flexible with respect to different domains and (ii) does not require any training material, but exploits Linked Data as background knowledge source to build essential learning resources. The major contribution of this work is a study of how Linked Data - an imprecise, redundant and large-scale knowledge resource - can be used to support Web scale Information Extraction in an effective and efficient way and identify the challenges involved. We show that, for domains that are covered, Linked Data serve as a powerful knowledge resource for Information Extraction. Experiments on a publicly available dataset demonstrate that, under certain conditions, this simple unsupervised approach can achieve competitive results against some complex state of the art that always depends on training data.

Abstract Measuring lexical semantic relatedness is an important task in Natural Language Processing (NLP). It is often a prerequisite to many complex NLP tasks. Despite an extensive amount of work dedicated to this area of research, there is a lack of an up-to-date survey in the field. This paper aims to address this issue with a study that is focused on four perspectives: (i) a comparative analysis of background information resources that are essential for measuring lexical semantic relatedness; (ii) a review of the literature with a focus on recent methods that are not covered in previous surveys; (iii) discussion of the studies in the biomedical domain where novel methods have been introduced but inadequately communicated across the domain boundaries; and (iv) an evaluation of lexical semantic relatedness methods and a discussion of useful lessons for the development and application of such methods. In addition, we discuss a number of issues in this field and suggest future research directions. It is believed that this work will be a valuable reference to researchers of lexical semantic relatedness and substantially support the research activities in this field.

We address the problem of searching multimedia by semantic similarity in a keyword space. In contrast to previous research we represent multimedia content by a vector of keywords instead of a vector of low-level features. This vector of keywords can be obtained through user manual annotations or computed by an automatic annotation algorithm. In this setting, we studied the influence of two aspects of the search by semantic similarity process: (1) accuracy of user keywords versus automatic keywords and (2) functions to compute semantic similarity between keyword vectors of two multimedia documents. We consider these two aspects to be crucial in the design of a keyword space that can exploit social-media information and can enrich applications such as Flickr and YouTube. Experiments were performed on an image and a video dataset with a large number of keywords, with different similarity functions and with two annotation methods. Surprisingly, we found that multimedia semantic similarity with automatic keywords performs as good as or better than 95% accurate user keywords.

Nowadays a large and growing percentage of information is stored in various multimedia formats. In order for multimedia information to be efficiently utilised by users, it is very important to add suitable metadata. In this paper we will present AKTiveMedia, a tool for enriching multimedia documents with semantic information.

In this demo paper, we present AKTiveDoc, a tool for supporting sharing and reuse of knowledge in document production (e.g. writing) and use (e.g. reading).

Comparative evaluation of Machine Learning (ML) systems used for Information Extraction (IE) has suffered from various inconsistencies in experimental procedures. This paper reports on the results of the Pascal Challenge on Evaluating Machine Learning for Information Extraction, which provides a standardised corpus, set of tasks, and evaluation methodology. The challenge is described and the systems submitted by the ten participants are briefly introduced and their performance is analysed.

No abstract available.

The new frontier of research on Information Extraction from texts is portability without any knowledge of Natural Language Processing. The market potential is very large in principle, provided that a suitable easy-to-use and effective methodology is provided. In this paper we describe LearningPinocchio, a system for adaptive Information Extraction from texts that is having good commercial and scientific success. Real world applications have been built and evaluation licenses have been released to external companies for application development. In this paper we outline the basic algorithm behind the scenes and present a number of applications developed with LearningPinocchio. Then we report about an evaluation performed by an independent company. Finally, we discuss the general suitability of this IE technology for real world applications and draw some conclusion.

No abstract available.

Amilcare is a tool for Adaptive Information Extraction (IE) designed for supporting active annotation of documents for the Semantic Web (SW). It can be used either for unsupervised document annotation or as a support for human annotation. Amilcare is portable to new applications/domains without any knowledge of IE, as it just requires users to annotate a small training corpus with the information to be extracted. It is based on (LP)2, a supervised learning strategy for IE able to cope with different texts types, from newspaper-like texts, to rigidly formatted Web pages and even a mixture of them[1][5].Adaptation starts with the definition of a tag set for annotation, possibly organized as an ontology. Then users have to manually annotate a small training corpus. Amilcare provides a default mouse-based interface called Melita, where annotations are inserted by first selecting a tag from the ontology and then identifying the text area to annotate with the mouse. Differently from similar annotation tools [4, 5], Melita actively supports training corpus annotation. While users annotate texts, Amilcare runs in the background learning how to reproduce the inserted annotation. Induced rules are silently applied to new texts and their results are compared with the user annotation. When its rules reach a (user-defined) level of accuracy, Melita presents new texts with a preliminary annotation derived by the rule application. In this case users have just to correct mistakes and add missing annotations. User corrections are inputted back to the learner for retraining. This technique focuses the slow and expensive user activity on uncovered cases, avoiding requiring annotating cases where a satisfying effectiveness is already reached. Moreover validating extracted information is a much simpler task than tagging bare texts (and also less error prone), speeding up the process considerably. At the end of the corpus annotation process, the system is trained and the application can be delivered. MnM [6] and Ontomat annotizer [7] are two annotation tools adopting Amilcare's learner.In this demo we simulate the annotation of a small corpus and we show how and when Amilcare is able to support users in the annotation process, focusing on the way the user can control the tool's proactivity and intrusivity. We will also quantify such support with data derived from a number of experiments on corpora. We will focus on training corpus size and correctness of suggestions when the corpus is increased.

This paper proposes a robust approach to parsing suitable for Information Extraction (IE) from texts using finite-state cascades. The approach is characterized by the construction of an approximation of the full parse tree that captures all the information relevant for IE purposes, leaving the other relations underspecified. Sequences of cascades of finite-state rules deterministically analyze the text, building unambiguous structures. Initially basic chunks are analyzed; then clauses are recognized and nested; finally modifier attachment is performed and the global parse tree is built. The parsing approach allows robust, effective and efficient analysis of real world texts. The grammar organization simplifies changes, insertion of new rules and integration of domain-oriented rules. The approach has been tested for Italian, English, and Russian. A parser based on such an approach has been implemented as part of Pinocchio, an environment for developing and running IE applications.