SMAFED: Real-Time Event Detection in Social Media Streams

Interactions via social media platforms have made it possible for anyone, irrespective of physical location, to gain access to quick information on events taking place all over the globe. However, the semantic processing of social media data is complicated due to challenges such as language complexity, unstructured data, and ambiguity. In this paper, we proposed the Social Media Analysis Framework for Event Detection (SMAFED). SMAFED aims to facilitate improved semantic analysis of noisy terms in social media streams, improved representation/embedding of social media stream content, and improved summarisation of event clusters in social media streams. For this, we employed key concepts such as integrated knowledge base, resolving ambiguity, semantic representation of social media streams, and Semantic Histogram-based Incremental Clustering based on semantic relatedness. Two evaluation experiments were conducted to validate the approach. First, we evaluated the impact of the data enrichment layer of SMAFED. We found that SMAFED outperformed other pre-processing frameworks with a lower loss function of 0.15 on the first dataset and 0.05 on the second dataset. Secondly, we determined the accuracy of SMAFED at detecting events from social media streams. The result of this second experiment showed that SMAFED outperformed existing event detection approaches with better Precision (0.922), Recall (0.793), and F-Measure (0.853) metric scores. The findings of the study present SMAFED as a more efficient approach to event detection in social media.

Out-of-vocabulary but not meaningless: Evidence for semantic-priming effects in pseudoword processing

Non-arbitrary phenomena in language, such as systematic association in the form-meaning interface, have been widely reported in the literature. Exploiting such systematic associations previous studies have demonstrated that pseudowords can be indicative of meaning. However, whether semantic activation from words and pseudowords is supported by the very same processes, activating a common semantic memory system, is currently not known. Here, we take advantage of recent progresses from computational linguistics models allowing to induce meaning representations for out-of-vocabulary strings of letters via domain-general associative-learning mechanisms applied to natural language. We combined these models with data from priming tasks, in which participants are showed two strings of letters presented sequentially one after the other and are then asked to indicate if the latter is a word or a pseudoword. In Experiment 1 we re-analyzed the data of the largest behavioral database on semantic priming, while in Experiment 2 we ran an independent replication on a new language, Italian, controlling for a series of possible confounds. Results were consistent across the two experiments and showed that the prime-word meaning interferes with the semantic pattern elicited by the target pseudoword (i.e., at increasing estimated semantic relatedness between prime word and target pseudoword, participants’ reaction times increased and accuracy decreased). These findings indicate that the same associative mechanisms governing word meaning also subserve the processing of pseudowords, suggesting in turn that human semantic memory can be conceived as a distributional system that builds upon a general-purpose capacity of extracting knowledge from complex statistical patterns.

Segmentation of semantic and perceptual boundaries affects associative memory and risk of false memories

Temporally structured sequences of experiences, such as narratives or life events, are segmented in memory into discrete situational models. In segmentation, contextual shifts are processed as situational boundaries that temporally cluster items according to the perceived contexts. As such, segmentation enhances associative binding of items within a situational model. One side effect of enhanced associative processing is increased risk of false recollections for not-presented, semantically related items. If so, do boundaries facilitate false recollections, or does segmentation protect against them? In two experiments, we introduced situational shifts in word sequences in the form of semantic and perceptual boundaries, with semantic relatedness between words or the frame color around a word changing on a regular basis. After encoding, we tested participants’ associative memory performance and false recollection rates. In Experiment 1, color boundaries occurred synchronously or asynchronously to semantic boundaries. We found better associative recognition, but also more false recollections, for synchronous than asynchronous boundaries. In Experiment 2, color boundaries occurred synchronous to semantic boundaries or were absent entirely. We found that false recollection rates elicited by semantic boundaries increased when color boundaries were absent. We also tested associative memory performance using a non-semantic, temporal memory task. We found better temporal memory performance for semantic boundaries, as well as a negative correlation between increased false recollection rates and better temporal memory performance for semantic lists, but not for random lists. We discuss implications for false memory theories and segmentation of narrative materials in false memory research.

Query-Based Extractive Text Summarization Using Sense-Oriented Semantic Relatedness Measure

This paper presents a query-based extractive text summarization method by using sense-oriented semantic relatedness measure. We have proposed a Word Sense Disambiguation (WSD) technique to find the exact sense of a word present in the sentence. It helps in extracting query relevance sentences while calculating the sense-oriented sentence semantic relatedness score between the query and input text sentence. The proposed method uses five unique features to make clusters of query-relevant sentences. A redundancy removal technique is also put forward to eliminate redundant sentences. We have evaluated our proposed WSD technique with other existing methods by using Senseval and SemEval datasets. Experimental evaluation and discussion signifies the better performance of proposed WSD method over current systems in terms of F-score. We compare our proposed query-based extractive text summarization method with other methods participated in Document Understanding Conference (DUC) and as well as with current methods. Evaluation and comparison state that the proposed query-based extractive text summarization method outperforms many existing methods. As an unsupervised learning algorithm, we obtained highest ROUGE (Recall-Oriented Understudy for Gisting Evaluation) score for all three DUC 2005, 2006 and 2007 datasets. Our proposed method is also quite comparable with other supervised learning based algorithms. We also observe that our query-based extractive text summarization method can recognize query relevance sentences which meet the query need.

Lexical Selection in Language Production

<p>Two experiments were conducted with first-year university students in an effort to discover more about what happens when a phrase is spoken. A paradigm was constructed with the intention of getting the participants to produce a simple, two-noun phrase at a cue and then 'catch' them out having them say the name of a single picture presented instead. The single picture presented to 'catch' the participants out (instead of the cue) was either the first or second name in the simple two-noun phrase, or a third, unplanned picture. The intention was to compare the relative timings of the different catch pictures in an effort to discover which of two theories of speech production best describes the cognitive processes that underlie such processes. The second experiment was an extension of this idea but also included a semantic relatedness variable, where the catch picture could be semantically related to an item shown during the planning of the simple, two-noun phrase. The results of these experiments were not in line with the hypothesis regarding the relative timings of the catch pictures, but were in line with the hypothesis that it would take longer to name catch pictures that were preceded by semantically related pictures. Implications of such findings are discussed along with possible future modifications to extend the utility of the paradigm used in this study.</p>

Lexical Selection in Language Production

<p>Two experiments were conducted with first-year university students in an effort to discover more about what happens when a phrase is spoken. A paradigm was constructed with the intention of getting the participants to produce a simple, two-noun phrase at a cue and then 'catch' them out having them say the name of a single picture presented instead. The single picture presented to 'catch' the participants out (instead of the cue) was either the first or second name in the simple two-noun phrase, or a third, unplanned picture. The intention was to compare the relative timings of the different catch pictures in an effort to discover which of two theories of speech production best describes the cognitive processes that underlie such processes. The second experiment was an extension of this idea but also included a semantic relatedness variable, where the catch picture could be semantically related to an item shown during the planning of the simple, two-noun phrase. The results of these experiments were not in line with the hypothesis regarding the relative timings of the catch pictures, but were in line with the hypothesis that it would take longer to name catch pictures that were preceded by semantically related pictures. Implications of such findings are discussed along with possible future modifications to extend the utility of the paradigm used in this study.</p>

Semantic relatedness maximisation for word sense disambiguation using a hybrid firefly algorithm

Word sense disambiguation (WSD) refers to determining the right meaning of a vague word using its context. The WSD intermediately consolidates the performance of final tasks to achieve high accuracy. Mainly, a WSD solution improves the accuracy of text summarisation, information retrieval, and machine translation. This study addresses the WSD by assigning a set of senses to a given text, where the maximum semantic relatedness is obtained. This is achieved by proposing a swarm intelligence method, called firefly algorithm (FA) to find the best possible set of senses. Because of the FA is based on a population of solutions, it explores the problem space more than exploiting it. Hence, we hybridise the FA with a one-point search algorithm to improve its exploitation capacity. Practically, this hybridisation aims to maximise the semantic relatedness of an eligible set of senses. In this study, the semantic relatedness is measured by proposing a glosses-overlapping method enriched by the notion of information content. To evaluate the proposed method, we have conducted intensive experiments with comparisons to the related works based on benchmark datasets. The obtained results showed that our method is comparable if not superior to the related works. Thus, the proposed method can be considered as an efficient solver for the WSD task.

Deep Convolutional Neural Network with KNN Regression for Automatic Image Annotation

Automatic image annotation is an active field of research in which a set of annotations are automatically assigned to images based on their content. In literature, some works opted for handcrafted features and manual approaches of linking concepts to images, whereas some others involved convolutional neural networks (CNNs) as black boxes to solve the problem without external interference. In this work, we introduce a hybrid approach that combines the advantages of both CNN and the conventional concept-to-image assignment approaches. J-image segmentation (JSEG) is firstly used to segment the image into a set of homogeneous regions, then a CNN is employed to produce a rich feature descriptor per area, and then, vector of locally aggregated descriptors (VLAD) is applied to the extracted features to generate compact and unified descriptors. Thereafter, the not too deep clustering (N2D clustering) algorithm is performed to define local manifolds constituting the feature space, and finally, the semantic relatedness is calculated for both image–concept and concept–concept using KNN regression to better grasp the meaning of concepts and how they relate. Through a comprehensive experimental evaluation, our method has indicated a superiority over a wide range of recent related works by yielding F1 scores of 58.89% and 80.24% with the datasets Corel 5k and MSRC v2, respectively. Additionally, it demonstrated a relatively high capacity of learning more concepts with higher accuracy, which results in N+ of 212 and 22 with the datasets Corel-5k and MSRC v2, respectively.

How much is a cow like a meow? A novel database of human judgements of audiovisual semantic relatedness

Semantic information about objects, events, and scenes influences how humans perceive, interact with, and navigate the world. Most evidence in support of semantic influence on cognition has been garnered from research conducted with an isolated modality (e.g., vision, audition). However, the influence of semantic information has not yet been extensively studied in multisensory environments potentially because of the difficulty in quantification of semantic relatedness. Past studies have primary relied on either a simplified binary classification of semantic relatedness based on category or on algorithmic values based on text corpora rather than human perceptual experience and judgement. With the aim to accelerate research into multisensory semantics, we created a constrained audiovisual stimulus set and derived similarity ratings between items within three categories (animals, instruments, household items). A set of 140 participants provided similarity judgments between sounds and images. Participants either heard a sound (e.g., a meow) and judged which of two pictures of objects (e.g., a picture of a dog and a duck) it was more similar to, or saw a picture (e.g., a picture of a duck) and selected which of two sounds it was more similar to (e.g., a bark or a meow). Judgements were then used to calculate similarity values of any given cross-modal pair. The derived and reported similarity judgements reflect a range of semantic similarities across three categories and items, and highlight similarities and differences among similarity judgments between modalities. We make the derived similarity values available in a database format to the research community to be used as a measure of semantic relatedness in cognitive psychology experiments, enabling more robust studies of semantics in audiovisual environments.

An Efficient Method for Biomedical Word Sense Disambiguation Based on Web-Kernel Similarity

Searching for the best sense for a polysemous word remains one of the greatest challenges in the representation of biomedical text. To this end, Word Sense Disambiguation (WSD) algorithms mostly rely on an External Source of Knowledge, like a Thesaurus or Ontology, for automatically selecting the proper concept of an ambiguous term in a given Window of Context using semantic similarity and relatedness measures. In this paper, we propose a Web-based Kernel function for measuring the semantic relatedness between concepts to disambiguate an expression versus multiple possible concepts. This measure uses the large volume of documents returned by PubMed Search engine to determine the greater context for a biomedical short text through a new term weighting scheme based on Rough Set Theory (RST). To illustrate the efficiency of our proposed method, we evaluate a WSD algorithm based on this measure on a biomedical dataset (MSH-WSD) that contains 203 ambiguous terms and acronyms. The obtained results demonstrate promising improvements.