Search

The rising popularity of Artificial Intelligence (AI) has brought considerable public interest as well faster and more direct transfer of research ideas into practice. One of the aspects of AI that still trails behind considerably is the role of machines in interpreting, enhancing, modeling, generating, and influencing social behavior. Such behavior is captured as social signals, usually by sensors recording multiple modalities, making it classic multimedia data. Such behavior can also be generated by an AI system when interacting with humans. Using AI techniques in combination with multimedia data can be used to pursue multiple goals, two of which are highlighted here. First, supporting people during social interactions and helping them to fulfil their social needs either actively or passively. Second, improving our understanding of how people collaborate, build relationships, and process self identity. Despite the rise of fields such as Social Signal Processing, a similar panel organised at ACM Multimedia 2014, and an area on social and emotional signals at the ACM MM since 2014, we argue that we have yet to truly fulfil the potential of the combining social signals and multimedia. This panel asks where we have come far enough and what remaining challenges there are in light of recent global events.

This paper describes the fourth edition of Lifelog challenges in ImageCLEF 2020. In this edition, the Lifelog challenges consist of two tasks which are Lifelog Moments Retrieval (LMRT) and Sport Per- formance Lifelog (SPLL). While the Lifelog Moments Retrieval chal- lenge follows the same format of the previous edition, its data is a larger multimodal dataset based on the merger of three previous NT- CIR Lifelog datasets, which contain approximately 191,439 images with corresponding visual concepts and other related metadata. The Sport Performance Lifelog, which is a brand new challenge, is composed of three subtasks that focus on predicting the expected performance of ath- letes who trained for a sport event. In summary, the ImageCLEF Lifelog 2020 receives 50 runs from six teams in total with competitive results. This publication has emanated from research supported in party by research grants from Irish Research Council (IRC) under Grant Number GOIPG/2016/741 and Science Foundation Ireland under grant numbers SFI/12/RC/2289 and SFI/13/RC/2106.

This paper describes the participation of the Organiser Team in the ImageCLEFlifelog 2019 Solve My Life Puzzle (Puzzle) and Lifelog Moment Retrieval (LMRT) tasks. We proposed to use LIFER 2.0, an enhanced version of LIFER, which was an interactive retrieval system for personal lifelog data. We utilised LIFER 2.0 with some additional visual features, obtained by using traditional visual bag-of-words, to solve the Puzzle task, while with the LMRT, we applied LIFER 2.0 only with the provided information. The results on both tasks confirmed that by using faceted filter and context browsing, a user can gain insights from their personal lifelog by employing very simple interactions. These results also serve as baselines for other approaches in the ImageCLEFlifelog 2019 challenge to compare with. publishedVersion

Online discussion forums provide open workspace allowing users to share information, exchange ideas, address problems, and form groups. These forums feature multimodal posts and analyzing them requires a framework that can integrate heterogeneous information extracted from the posts, i.e. text, visual content and the information about user interactions with the online platform and each other. In this paper, we develop a generic framework that can be trained to identify communication behavior and patterns in relation to an entity of interest, be it user, image or text in internet forums. As the case study we use the analysis of violent online political extremism content, which has been a major challenge for domain experts. We demonstrate the generalizability and flexibility of our framework in predicting relational information between multimodal entities by conducting extensive experimentation around four practical use cases.

Item does not contain fulltext

This paper describes the participation of Organizer Team in the ImageCLEFlifelog 2017 Retrieval and Summarization subtasks. In this paper, we propose some baseline approaches, using only the provided information, which require different involvement levels from the users. With these baselines we target at providing references for other approaches that aim to solve the problems of lifelog retrieval and summarization.

In this paper we investigate the potential of social multimedia and open data for automatically identifying regions within the city. We conjecture that the regions may be characterized by specific patterns related to their visual appearance, the manner in which the social media users describe them, and the human mobility patterns. Therefore, we collect a dataset of Foursquare venues, their associated images and users, which we further enrich with a collection of city-specific Flickr images, annotations and users. Additionally, we collect a large number of neighbourhood statistics related to e.g., demographics, housing and services. We then represent visual content of the images using a large set of semantic concepts output by a convolutional neural network and extract latent Dirichlet topics from their annotations. User, text and visual information as well as the neighbourhood statistics are further aggregated at the level of postal code regions, which we use as the basis for detecting larger regions in the city. To identify those regions, we perform clustering based on individual modalities as well as their ensemble. The experimental analysis shows that the automatically detected regions are meaningful and have a potential for better understanding dynamics and complexity of a city.

In this paper, we propose and evaluate various distance-aware weighting strategies to improve reconstruction accuracy of a voxel-based model according to the Truncated Signed Distance Function (TSDF), from the data obtained by low-cost depth sensors. We look at two strategy directions: (a) weight definition strategies prioritizing importance of the sensed data depending on the data accuracy, and (b) model updating strategies defining the level of influence of the new data on the existing 3D model. In particular, we introduce Distance-Aware (DA) and Distance-Aware Slow-Saturation (DASS) updating methods to intelligently integrate the depth data into the synthetic 3D model based on the distance-sensitivity metric of a low-cost depth sensor. By quantitative and qualitative comparison of the resulting synthetic 3D models to the corresponding ground-truth models, we identify the most promising strategies, which lead to an accuracy improvement involving a reduction of the model error by 10 '—' 35%.

Expert finding addresses the problem of retrieving a ranked list of people who are knowledgeable on a given topic. Several models have been proposed to solve this task, but so far these have focused solely on returning the most knowledgeable people as experts on a particular topic. In this paper we argue that in a real-world organizational setting the notion of the “best expert” also depends on the individual user and her needs. We propose a user-oriented approach that balances two factors that influence the user’s choice: time to contact an expert, and the knowledge value gained after. We use the distance between the user and an expert in a social network to estimate contact time, and consider various social graphs, based on organizational hierarchy, geographical location, and collaboration, as well as the combination of these. Using a realistic test set, created from interactions of employees with a university-wide expert search engine, we demonstrate substantial improvements over a state-of-the-art baseline on all retrieval measures.

As retrieval systems become more complex, learning to rank approaches are being developed to automatically tune their parameters. Using online learning to rank approaches, retrieval systems can learn directly from implicit feedback, while they are running. In such an online setting, algorithms need to both explore new solutions to obtain feedback for effective learning, and exploit what has already been learned to produce results that are acceptable to users. We formulate this challenge as an exploration-exploitation dilemma and present the first online learning to rank algorithm that works with implicit feedback and balances exploration and exploitation. We leverage existing learning to rank data sets and recently developed click models to evaluate the proposed algorithm. Our results show that finding a balance between exploration and exploitation can substantially improve online retrieval performance, bringing us one step closer to making online learning to rank work in practice.

We propose a retrieval model for searching microblog posts for a given topic of interest. We develop a language modeling approach tailored to microblogging characteristics, where redundancy-based IR methods cannot be used in a straightforward manner. We enhance this model with two groups of quality indicators: textual and microblog specific. Additionally, we propose a dynamic query expansion model for microblog post retrieval. Experimental results on Twitter data reveal the usefulness of boolean search, and demonstrate the utility of quality indicators and query expansion in microblog search.

Insight-DCU participated in the instance search (INS) and semantic indexing (SIN) tasks in 2014. Two very different approaches were submitted for instance search, one based on features extracted using pre-trained deep convolutional neural networks (CNNs), and another based on local SIFT features, large vocabulary visual bag-of-words aggregation, inverted index-based lookup, and geometric verification on the top-N retrieved results. Two interactive runs and two automatic runs were submitted, the best interactive runs achieved a mAP of 0.135 and the best automatic 0.12. Our semantic indexing runs were based also on using convolutional neural network features, and on Support Vector Machine classifiers with linear and RBF kernels. One run was submitted to the main task, two to the no annotation task, and one to the progress task. Data for the no-annotation task was gathered from Google Images and ImageNet. The main task run has achieved a mAP of 0.086, the best no-annotation runs had a close performance to the main run by achieving a mAP of 0.080, while the progress run had 0.043.

This year the iAd-DCU team participated in three of the assigned TRECVid 2011 tasks; Semantic Indexing (SIN), Interactive Known-Item Search (KIS) and Multimedia Event Detection (MED). For the SIN task we presented three full runs using global features, local features and fusion\ud of global, local features and relationships between concepts respectively. The evaluation results show that local features achieve better performance, with marginal gains found when introducing global features and relationships between concepts. With regard to our KIS submission, similar to our 2010 KIS experiments, we have implemented an iPad interface to a KIS video search tool.\ud The aim of this year’s experimentation was to evaluate different display methodologies for KIS interaction. For this work, we integrate a clustering element for keyframes, which operates over MPEG-7 features using k-means clustering. In addition, we employ concept detection, not simply for search, but as a means of choosing most representative keyframes for ranked items. For our\ud experiments we compare the baseline non-clustering system to a clustering system on a topic by topic basis. Finally, for the first time this year the iAd group at DCU has been involved in the MED Task. Two techniques are compared, employing low-level features directly and using concepts as\ud intermediate representations. Evaluation results show promising initial results when performing event detection using concepts as intermediate representations.

Why do links work? Link-based ranking algorithms are based on the often implicit assumption that linked documents are semantically related to each other, and that link information is therefore useful for retrieval. Although the benefits of link information are well researched, this underlying assumption on why link evidence works remains untested, and the main aim of this paper is to do exactly that. Specifically, we use Wikipedia because it has a dense link structure in combination with a large category structure, which allows for an independent measurement of the semantic relatedness of linked documents. Our main findings are that: 1) global, query-independent link evidence, is not affected by the semantic nature of the links, and 2) for local, query-dependent link evidence, the effectiveness of links increases as their semantic distance decreases. That is, we directly observe that links between semantically related pages are more effective for ad hoc retrieval than links between unrelated ones. These findings confirm and quantify the underlying assumption of existing link-based methods, which sheds further light on our understanding of the nature of link evidence. Such deeper understanding is instrumental for the development of novel link-based methods.

Word clouds are a summarised representation of a document’s text, similar to tag clouds which summarise the tags assigned to documents. Word clouds are similar to language models in the sense that they represent a document by its word distribution. In this paper we investigate the differences between word cloud and language modelling approaches, and specifically whether effective language modelling techniques also improve word clouds. We evaluate the quality of the language model using a system evaluation test bed, and evaluate the quality of the resulting word cloud with a user study. Our experiments show that different language modelling techniques can be applied to improve a standard word cloud that uses a TF weighting scheme in combination with stopword removal. Including bigrams in the word clouds and a parsimonious term weighting scheme are the most effective in both the system evaluation and the user study.

Online news agents provide commenting facilities for their readers to express their opinions or sentiments with regards to news stories. The number of user supplied comments on a news article may be indicative of its importance, interestingness, or impact. We explore the news comments space, and compare the log-normal and the negative binomial distributions for modeling comments from various news agents. These estimated models can be used to normalize raw comment counts and enable comparison across different news sites. We also examine the feasibility of online prediction of the number of comments, based on the volume observed shortly after publication. We report on solid performance for predicting news comment volume in the long run, after short observation. This prediction can be useful for identifying news stories with the potential to "take off," and can be used to support front page optimization for news sites.

We analyse the statistical behavior of query-associated quantities in query-logs, namely, the sum and mean of IDF of query terms, otherwise known as query specificity and query mean specificity. We narrow down the possibilities for modeling their distributions to gamma, log-normal, or log-logistic, depending on query length and on whether the sum or the mean is considered. The results have applications in query performance prediction and artificial query generation.

Users often search for entities instead of documents and in this setting are willing to provide extra input, in addition to a query, such as category information and example entities. We propose a general probabilistic framework for entity search to evaluate and provide insight in the many ways of using these types of input for query modeling. We focus on the use of category information and show the advantage of a category-based representation over a term-based representation, and also demonstrate the effectiveness of category-based expansion using example entities. Our best performing model shows very competitive performance on the INEX-XER entity ranking and list completion tasks.