A5083881309- Topic Modeling
- Misinformation and Its Impacts
- Information Retrieval and Search Behavior
- Speech and dialogue systems
- Spam and Phishing Detection
- Mobile Crowdsensing and Crowdsourcing
- Expert finding and Q&A systems
- Natural Language Processing Techniques
- Advanced Text Analysis Techniques
- Web Data Mining and Analysis
- Complex Network Analysis Techniques
- Personal Information Management and User Behavior
- Hate Speech and Cyberbullying Detection
- Sentiment Analysis and Opinion Mining
- Recommender Systems and Techniques
- AI in Service Interactions
- Data Quality and Management
- Ethics and Social Impacts of AI
- Adversarial Robustness in Machine Learning
- Semantic Web and Ontologies
- Music and Audio Processing
- Video Analysis and Summarization
- Explainable Artificial Intelligence (XAI)
- Speech Recognition and Synthesis
- Wikis in Education and Collaboration
RMIT University
2015-2024
Delft University of Technology
2024
UNSW Sydney
2024
The New Press
2024
MIT University
2015-2021
National University of Distance Education
2010-2014
Yangon University Of Distance Education
2011-2013
Universidad Complutense de Madrid
2008
In this work, we explore the types of triggers that spark trends on Twitter, introducing a typology with following 4 types: news , ongoing events memes and commemoratives . While previous research has analyzed trending topics over long term, look at earliest tweets produce trend, aim categorizing early on. This allows us to provide filtered subset end users. We experiment set straightforward language‐independent features based social spread categorize them using typology. Our method provides...
We conducted a laboratory-based observational study where pairs of people performed search tasks communicating verbally. Examination the discourse allowed commonly used interactions to be identified for Spoken Conversational Search (SCS). compared existing models behaviour. find that SCS is more complex and interactive than traditional search. This work enhances our understanding different behaviours proposes research opportunities an audio-only system. Future will focus on creating...
This report documents the program and outcomes of Dagstuhl Seminar 23031 "Frontiers Information Access Experimentation for Research Education", which brought together 38 participants from 12 countries. The seminar addressed technology-enhanced information access (information retrieval, recommender systems, natural language processing) specifically focused on developing more responsible experimental practices leading to valid results, both research as well scientific education. featured a...
We deal with shrinking the stream of tweets for scheduled events in real-time, following two steps: (i) sub-event detection, which determines if something new has occurred, and (ii) tweet selection, picks a to describe each sub-event. By comparing summaries three languages live reports by journalists, we show that simple text analysis methods do not involve external knowledge lead cover 84% sub-events on average, 100% key types (such as goals soccer).
We present preliminary findings from a study of mixed initiative conversational behaviour for informational search in an acoustic setting. The aim the observational is to reveal insights into how users would conduct searches over voice where screen absent but are able converse interactively with system. conducted alaboratory-based 13 pairs participants each completing three tasks different cognitive complexity levels. communication between was analyzed interaction patterns used process. This...
Like other disease outbreaks, the COVID-19 pandemic has led to rapid generation and dissemination of misinformation fake news. We investigated whether subscribers a fact checking newsletter (n = 1397) were willing share possible misinformation, predictors sharing are same as for general samples. also willingness have vaccine found that although acceptance was high on average, it decreased function lower belief in science higher conspiracy mentality. 24% participants had shared this predicted...
Twitter summarizes the great deal of messages posted by users in form trending topics that reflect top conversations being discussed at a given moment. These tend to be connected current affairs. Different happenings can give rise emergence these topics. For instance, sports event broadcasted on TV, or viral meme introduced community users. Detecting type origin facilitate information filtering, enhance real-time data processing, and improve user experience. In this paper, we introduce...
Creating and deploying customized applications is crucial for operational success enriching user experiences in the rapidly evolving modern business world. A prominent facet of integration chatbots or voice assistants. The rapid evolution Large Language Models (LLMs) has provided a powerful tool to build conversational applications. We present Walert, LLM-based agent able answer frequently asked questions about computer science degrees programs at RMIT University. Our demo aims showcase how...
This paper examines the ethical question, 'What is a good search engine?' Since engines are gatekeepers of global online information, it vital they do their job ethically well. While Internet now several decades old, topic remains under-explored from interdisciplinary perspectives. presents novel role-based approach involving four models types engine behavior: Customer Servant, Librarian, Journalist, and Teacher. It explores these with reference to research field information retrieval, by...
Reputation management experts have to monitor--among others--Twitter constantly and decide, at any given time, what is being said about the entity of interest (a company, organization, personality...). Solving this reputation monitoring problem automatically as a topic detection task both essential--manual processing data either costly or prohibitive--and challenging--topics for are usually fine-grained suffer from sparsity. We focus on solution that (i) learns pairwise tweet similarity...
The purpose of the SIGIR 2019 workshop on Fairness, Accountability, Confidentiality, Transparency, and Safety (FACTS-IR) was to explore challenges in responsible information retrieval system development deployment. To this end, aimed crowd-source from larger community draft an actionable research agenda five key dimensions retrieval: fairness, accountability, confidentiality, transparency, safety. Such can guide others that are interested pursuing FACTS-IR research, as well inform potential...
research-article Open Access Share on Human-AI Cooperation to Tackle Misinformation and Polarization Authors: Damiano Spina RMIT University, Melbourne, Australia AustraliaView Profile , Mark Sanderson Daniel Angus Queensland University of Technology, Brisbane, Gianluca Demartini The Queensland, Dana Mckay Lauren L. Saling Ryen W. White Microsoft Research, Redmond, WA, USA USAView Authors Info & Claims Communications the ACMVolume 66Issue 7July 2023pp...
Many evaluation metrics have been defined to evaluate the effectiveness ad-hoc retrieval and search result diversification systems. However, it is often unclear which metric should be used analyze performance of systems given a specific task. Axiomatic analysis an informative mechanism understand fundamentals their suitability for particular scenarios. In this paper, we define constraint-based axiomatic framework study existing in The informed definition Rank-Biased Utility (RBU) --...
Misinformation is an ever increasing problem that difficult to solve for the research community and has a negative impact on society at large. Very recently, been addressed with crowdsourcing-based approach scale up labeling efforts: assess truthfulness of statement, instead relying few experts, crowd (non-expert) judges exploited. We follow same study whether crowdsourcing effective reliable method statements during pandemic. specifically target related COVID-19 health emergency, still...