Automatic sarcasm detection from text is an important classification task that can help identify the actual sentiment in user-generated data, such as reviews or tweets. Despite its usefulness, remains a challenging task, due to lack of any vocal intonation facial gestures textual data. To date, most approaches addressing problem have relied on hand-crafted affect features, pre-trained models non-contextual word embeddings, Word2vec. However, these inherit limitations render them inadequate...

As online social networking emerges, there has been increased interest to utilize the underlying network structure as well available information on peers improve needs of a user. In this paper, we focus improving performance collection from neighborhood user in dynamic network. We introduce sampling-based algorithms efficiently explore user's respecting its and quickly approximate quantities interest. analyze variants basic sampling scheme exploring correlations across our samples. Models...

Small changes on the structure of a graph can have dramatic effect its connectivity. While in traditional theory, focus is well-defined properties connectivity, such as biconnectivity, context social , connectivity typically manifested by ability to carry processes . In this paper, we consider problem adding small set nonexisting edges ( shortcuts ) with main objective minimizing characteristic path length This property determines average distance between pairs vertices and essentially...

Small changes in the network topology can have dramatic effects on its capacity to disseminate information. In this paper, we consider problem of adding a small number ghost edges order minimize average shortest-path distance between nodes, towards smaller-world network. We formalize suggesting and propose novel method for quickly evaluating importance sparse graphs. Through experiments real synthetic data sets, demonstrate that our approach performs very well, varying range conditions, it...

Trajectory prediction aims to estimate an entity’s future path using its current position and historical movement data, benefiting fields like autonomous navigation, robotics, human analytics. Deep learning approaches have become key in this area, utilizing large-scale trajectory datasets model patterns, but face challenges managing complex spatial dependencies adapting dynamic environments. To address these challenges, we introduce TrajLearn , a novel for that leverages generative modeling...

Affective tasks such as sentiment analysis, emotion classification, and sarcasm detection have been popular in recent years due to an abundance of user-generated data, accurate computational linguistic models, a broad range relevant applications various domains. At the same time, many studies highlighted importance text preprocessing, integral step any natural language processing prediction model downstream task. While preprocessing affective systems is well-studied, word vector-based models...

One popular thread of research in computational sarcasm detection involves modeling as a contrast between positive and negative sentiment polarities or exploring more fine-grained categories emotions such happiness, sadness, surprise, so on. Most current models, however, treat these affective features independently, without regard for the sequential information encoded among states. In order to explore role transitions states, we formulate task sequence classification problem by leveraging...

Affective tasks, including sentiment analysis, emotion classification, and sarcasm detection have drawn a lot of attention in recent years due to broad range useful applications various domains. The main goal affect tasks is recognize <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">states</i> such as mood, sentiment, emotions from textual data (e.g., news articles or product reviews). Despite the importance utilizing...

The capacity to collect and analyze the actions of individuals in online social systems at minute-by-minute time granularity offers new perspectives on collective human behavior research. Macroscopic analysis massive datasets raises interesting observations patterns processes. But working a large scale has its own limitations, since it typically doesn't allow for interpretations microscopic level. We examine how different types individual affect decisions friends network. begin with problem...

Mining large-scale trajectory data streams (of moving objects) has been of ever increasing research interest due to an abundance modern tracking devices and its large number critical applications. In this paper, we are interested in mining group patterns objects. Group pattern describes a special type task that requires efficiently discover trajectories objects found close proximity each other for period time. particular, focus on pedestrians coming from motion video analysis interactive...

The bulk synchronous parallel (BSP) is a celebrated synchronization model for general-purpose computing that has successfully been employed distributed training of machine learning models. A prevalent shortcoming the BSP it requires workers to wait straggler at every iteration. To ameliorate this classic BSP, we propose ELASTICBSP aims relax its strict requirement. proposed offers more flexibility and adaptability during phase, without sacrificing on accuracy trained model. We also an...

Scene Classification has been addressed with numerous techniques in computer vision literature. However, the increasing number of scene classes datasets field, it become difficult to achieve high accuracy context robotics. In this paper, we implement an approach which combines traditional deep learning natural language processing methods generate a word embedding based algorithm. We use key idea that (objects scene) image should be representative label meaning group objects could assist...

As online social networking emerges, there has been increased interest to utilize the underlying structure as well available information improve search. In this paper, we focus on improving performance of collection from neighborhood a user in dynamic network. To end, introduce sampling based algorithms quickly approximate quantities vicinity user's graph. We then and analyze variants basic scheme exploring correlations across our samples. Models centralized distributed networks are...

Understanding and predicting the emotional trajectory in multi-party multi-turn conversations is of great significance. Such information can be used, for example, to generate empathetic response human-machine interaction or inform models pre-emptive toxicity detection. In this work, we introduce novel problem Predicting Emotions Conversations (PEC) next turn (n+1), given combinations textual and/or emotion input up n. We systematically approach by modeling three dimensions inherently...

Speech emotion recognition (SER) is the task of automatically recognizing emotions expressed in spoken language. Current approaches focus on analyzing isolated speech segments to identify a speaker's emotional state. Meanwhile, recent text-based methods have effectively shifted towards conversation (ERC) that considers conversational context. Motivated by this shift, here we propose SERC-GCN, method for (SERC) predicts state incorporating context, speaker interactions, and temporal...

Large-scale network mining and analysis is key to revealing the underlying dynamics of networks, not easily observable before. Lately, there a fast-growing interest in learning low-dimensional continuous representations networks that can be utilized perform highly accurate scalable graph tasks. A family these methods based on performing random walks learn its structural features providing sequence as input deep architecture embedding. While well, they only operate static networks. However,...