Program
of
trackDemos

Demos will be presented during breaks in the exhibition area during the three days of the conference.

List of accepted demos :

• OSTRICH: Versioned Random­-Access Triple Store
  Authors: Ruben Taelman and Ruben Verborgh

  Keywords: Linked Data, RDF, Versioning, OSTRICH, Triple Store

  Abstract:
  The collection of Linked Data is evergrowing and many datasets are frequently being updated. In order to fully exploit the potential of the information that is available in and over historical dataset ver­ sions, we need to be able to store and query Linked Datasets effi­ciently. In this demonstration, we introduce OSTRICH, which is an efficient multi-­version triple store with versioned querying support. We demonstrate the capabilities of OSTRICH using a Web-­based graphical user interface in which a store can be opened or created. Using this interface, the user is able to querying in, between and over different versions, ingest new versions, and retrieve summariz­ing statistics.

• Four-Dimensional Shopping Mall: Sequential Group Willingness Optimization under VR Environments
  Authors: Hong-Han Shuai, Yueh-Hsue Li, Chun-Chieh Feng and Wen-Chih Peng

  Keywords: New Human-Computer Interfaces, Social Networks, E-Commerce

  Abstract:
  Nowadays, people get used to buying a variety of commodities from e-commerce platforms because of the convenience and easy access to Internet. As a result, the sales on e-commerce platforms have grown exponentially. It is promising to customize the online shops for different users under VR environments since users do not need to waste time of going upstairs or downstairs for finding interesting commodities. Therefore, in this demo paper, we propose a novel four-dimensional shopping mall that offers online group shopping services with the customized shop recommendation, where a group of friends in the proposed four-dimensional shopping mall can teleport to the next shop by one click. We formulate a Sequential group wIllingness OptimizatioN (SION) problem, prove SION is NP-hard, and provide an efficient algorithm called zeta-GSS. The experimental results show that the solution quality of the proposed zeta-GSS is close to the optimal solution, while the execution time only requires 3.3%. Finally, we build the prototype of the four-dimensional shopping mall, which can be demonstrated for users to experience the next-generation online shopping.

• Generating Semantic Trajectories Using a Car Signal Ontology
  Authors: Benjamin Klotz, Raphaël Troncy, Daniel Wilms and Christian Bonnet

  Keywords: Demonstration, Semantic trajectories, Automotive ontology, VSS

  Abstract:
  In this paper, we use semantic technologies to deal with the challenge of enriching trajectory data in the automotive industry for offline analysis with focus on non automotive domain experts. Current works do not to combine reusable extensive vocabularies with trajectory models in non-specific use cases. We used a combination of ontologies and created a vehicle signal ontology to provide an environment, where we developed an application that analyzes the variations of signal values and “driving smoothness” percentage as annotation of trajectories. This work can enable semantic web developers to query car data without being domain experts.

• Chisel: Sculpting Tabular and Non-Tabular Data on the Web
  Authors: Johannes Doleschal, Nico Ho?llerich, Wim Martens and Frank Neven

  Keywords: CSV, Schema languages, Semi-structured data

  Abstract:
  Chisel is a tool for flexible manipulation of CSV-like data, motivated by the recent effort of the World Wide Web Consortium (W3C) towards a recommendation for tabular data and metadata on the Web. In brief, Chisel supports an expressive built-in schema language for CSV-like data, that can handle both tabular and non-tabular data. Furthermore, it supports a simple programming language for transforming tabular and non-tabular CSV-like data. In this demo, we showcase the system for specifying and validating schemas, building transformations, and setting up a pipeline for automatic conversion of “wild” CSV-like data into structured tabular data. We present several use cases for Chisel specifically targeted at exemplifying the ease of specifying, modifying, and understanding Sculpt schemas as well as extracting and transforming data.

• Enhancing Community Interactions with Data-Driven Chatbots – The DBpedia Chatbot
  Authors: Ram G Athreya and Ricardo Usbeck

  Keywords: chatbot, knowledge graphs, semantic web, dbpedia, information retrieval, dialog systems, interactive intelligence, machine learning, conversational computing

  Abstract:
  In this demo, we introduce the DBpedia Chatbot, a data-driven chatbot over knowledge graphs to improve community interaction. There were three main challenges, namely (1) understanding a user query, (2) fetching relevant information based on the query through the DBpedia knowledge graph or other sources and (3) tailoring the responses based on the standards of each output platform (i.e. Web, Slack, Facebook) as well as (4) developing subsequent user interactions with the \approach.

• THOR: Text-enabled Analytics for Humanitarian Operations
  Authors: Mayank Kejriwal, Daniel Gilley, Pedro Szekely and Jill Crisman

  Keywords: Low-resource languages, humanitarian operations, entity resolution, knowledge graphs, visualization, situational awareness

  Abstract:
  In this demonstration, we present the Text-enabled Humanitarian Operations in Real-time (THOR) framework, which is being prototyped to provide visual and analytical \emph{situational awareness} to humanitarian and disaster relief (HADR) planners. THOR is a collaborative effort between industrial and university research laboratories, designed with an intent to support both military and civilian HADR operations. At its core, THOR is powered by a domain-specific knowledge graph, which is derived from natural language outputs and is amenable to real-time analytics. THOR is designed to operate in low-resource linguistic environments, process heterogeneous data, including news and social media, reason about arbitrary disasters not knowable in advance, and provide advanced graphical interaction capabilities. We will demo the latest prototype of THOR using an interactive case study situation.

• API Learning: Applying Machine Learning to Manage the Rise of API Economy
  Authors: Mehdi Bahrami, Junhee Park, Lei Liu and Wei-Peng Chen

  Keywords: REST API, Machine Learning, Automation

  Abstract:
  Application Programming Interface (API) exposes data and functions of a software application to third-party users. In digital business, API economy is one of the key component for determining the value of provided services. With the rise in number of publicly available APIs, understanding each API endpoint manually is not only labor intensive but also an error prone task for software developers. In addition, APIs are often revised or updated regularly, creating significant overheads for software developers to keep track of all changes. Finally, due to the complexity of understanding the sheer number of APIs, it is difficult for software developers to find the best possible API combinations (i.e. API Mashups). In this demonstration, we introduce API Learning platform which employs machine-learning based technologies to efficiently search APIs, validate APIs, and generate API mashups. These technologies enable a machine to automatically generate machine-readable format of API functionalities from API documentations, understand variety of API descriptions, validate extracted information through automatic API validation, and finally recommend API mashups for a specific purpose. As of now, API Learning platform collected over 14,000 API documentations and generates a machine readable format for REST APIs with an accuracy of 84%. The proposed demo prototype shows how it enables users to quickly find relevant APIs, automatically verify API availability, and get the best possible API mashup recommendations.

• CredEye: A Credibility Lens for Analyzing and Explaining Misinformation
  Authors: Kashyap Popat, Subhabrata Mukherjee, Jannik Strötgen and Gerhard Weikum

  Keywords: Fact Checking, Credibility Analysis, Interpretable Learning

  Abstract:
  Rapid increase of misinformation online has emerged as one of the biggest challenges in this post-truth era. This has given rise to many fact-checking websites that manually assess doubtful claims. However, the speed and scale at which misinformation spreads in online media inherently limits manual verification. Hence, the problem of automatic credibility assessment has attracted great attention. In this work, we present CredEye, a system for automatic credibility assessment. It takes a natural language claim as input from the user and automatically analyzes its credibility by considering relevant articles from the Web. Our system captures the joint interaction between the language style of the articles, their stance towards the claim and the trustworthiness of the sources. In addition, extraction of supporting evidence in the form of enriched snippets makes the verdicts of CredEye transparent and interpretable.

• Using SafeKeeper to Protect Web Passwords
  Authors: Arseny Kurnikov, Klaudia Krawiecka, Andrew Paverd, Mohammad Mannan and N. Asokan

  Keywords: Intel SGX, password databases, web authentication

  Abstract:
  Although passwords are by far the most widely-used user authentication mechanism on the web, their security is threatened by password phishing and password database breaches. SafeKeeper is a system for protecting web passwords against very strong adversaries, including sophisticated phishers and compromised servers. Compared to other approaches, one of the key differentiating aspects of SafeKeeper is that it provides web users with verifiable assurance that their passwords are being protected. In this paper, we demonstrate precisely how SafeKeeper can be used to protect web passwords in real-world systems. We first explain two important deployability aspects: i) how SafeKeeper can be integrated into the popular WordPress platform, and ii) how ordinary web users can use Intel SGX remote attestation to verify that SafeKeeper is running on a particular server. We then describe three demonstrations to illustrate the use of SafeKeeper : i) showing the user experience when visiting a legitimate website; ii) showing the encryption of the password in transit via live packet-capture; and iii) showing how SafeKeeper performs in the presence of phishing.

• I Read but Don’t Agree: Privacy Policy Benchmarking using Machine Learning and the EU GDPR
  Authors: Welderufael B. Tesfay, Peter Hofmann, Toru Nakamura, Shinsaku Kiyomoto and Jetzabel Serna

  Keywords: privacy, privacy policy summarization, data protection regulation

  Abstract:
  With the continuing growth of the Internet landscape, users share large amount of personal, sometimes, privacy sensitive data. Often, users have little or no clear knowledge about what service providers do with the trails of personal data they leave on the Internet. While regulations impose rather strict requirements that service providers should abide by, the defacto approach seems to be communicating data processing practices through privacy policies. However, privacy policies are long and complex for users to read and understand, thus failing their mere objective of informing users about the promised data processing behaviors of service providers. To address this pertinent issue, we propose a machine learning based approach to summarize the rather long privacy policy into short and condensed notes following a risk-based approach and using the European Union Data Protection Regulation (EU GDPR) aspects as assessment criteria. The results are promising and indicate that our tool can summarize lengthy privacy policies in a short period of time, thus supporting users to take informed decisions regarding their information disclosure behaviors.

• MusicLynx: exploring music through artist similarity graphs
  Authors: Alo Allik, Florian Thalmann and Mark Sandler

  Keywords: music discovery and recommendation, music information retrieval, machine learning, graph theory, web programming, Semantic Web, Linked Data, similarity modeling, affective computing

  Abstract:
  MusicLynx is a web application for music discovery that enables users to explore an artist similarity graph constructed by linking together various open public data sources. It provides a multifaceted browsing platform that strives for an alternative, graph-based representation of artist connections to the grid-like conventions of traditional recommendation systems. Bipartite graph filtering of the Linked Data cloud, content-based music information retrieval, machine learning on crowd-sourced information and Semantic Web technologies are combined to analyze existing and create new categories of music artists through which they are connected. The categories can uncover similarities between artists who otherwise may not be immediately associated: for example, they may share ethnic background or nationality, common musical style or be signed to the same record label, come from the same geographic origin, share a fate or an affliction, or have made similar lifestyle choices. They may also prefer similar musical keys, instrumentation, rhythmic attributes, or even moods their music evokes. This demonstration is primarily meant to showcase the graph-based artist discovery interface of MusicLynx: how artists are connected through various categories, how the different graph filtering methods affect the topology and geometry of linked artists graphs, and ways in which users can connect to external services for additional content and information about objects of their interest.

• Walking down a Different Path: Route Recommendation based on Visual and Facility based Diversity
  Authors: Yihong Zhang, Panote Siriaraya, Yuanyuan Wang, Shoko Wakamiya, Yukiko Kawai and Adam Jatowt

  Keywords: Walking Navigation System, Route Recommendation, Visual Diversity

  Abstract:
  For a traveler to enjoy a trip in a city, one important factor is the diversity of sceneries and facilities along the route. Current navigation systems can provide the shortest route between two points, as well as scenic or safe routes. However, diversity is largely ignored in existing works. In this paper, we present a system that provides diversity-based route recommendation. It measures visual-based diversity and facility-based diversity with information extracted from publicly available data such as Google Street View images and FourSquare venues. As we will show, the current prototype system is able to provide diversity-based route recommendation for city areas in San Fransisco and Kyoto.

• Will This Video Go Viral? Explaining and Predicting the Popularity of Youtube Videos
  Authors: Quyu Kong, Marian-Andrei Rizoiu, Siqi Wu and Lexing Xie

  Keywords: Popularity Modeling, Youtube Videos, Interactive Visualization, Video Comparison, Channel Comparison, Identifying Future Popular Videos, Promotion Simulation

  Abstract:
  What makes content go viral? Which videos become popular and why others don’t? Such questions have elicited significant attention from both researchers and industry, particularly in the context of online media. A range of models have been recently proposed to explain and predict popularity; however, there is a short supply of practical tools, accessible for regular users, that leverage these theoretical results. HIPie – an interactive visualization system – is created to fill this gap, by enabling users to reason about the virality and the popularity of online videos. It retrieves the metadata and the past popularity series of Youtube videos, it employs Hawkes Intensity Process, a state-of-the-art online popularity model for explaining and predicting video popularity, and it presents videos comparatively in a series of interactive plots. This system will help both content consumers and content producers in a range of data-driven inquiries, such as to comparatively analyze videos and channels, to explain and predict future popularity, to identify viral videos, and to estimate response to online promotion.

• Real-Time Emulation of a Marshall JCM 800 Guitar Tube Amplifier, Audio FX Pedals, in a Virtual Pedalboard
  Authors: Michel Buffa and Jerôme Lebrun

  Keywords: WebAudio, Tube Guitar Amp Simulation, Audio Effects, Plugin Architecture, Web Standards

  Abstract:
  The ANR project WASABI [12] will last 42 months and consists in developing a 2 million songs database with interactive WebAudio enhanced client applications. Client applications target composers, music schools, sound engineering schools, musicologists, music streaming services and journalists. In this paper, we present a virtual pedalboard (a set of chainable audio effects on the form of “pedals”), and a guitar tube amplifier simulation for guitarists, that will be associated with songs from the WASABI database. Music schools and music engineering schools are interested in such tools that can be run in a Web page, without the need to install any further software. Take a classic rock song: isolate the guitar solo, study it, then mute it and play guitar real-time along the other tracks using an online guitar amplifier that reproduces the real guitar amp model used in the song, with its signature sound, proper dynamic and frequency response. Add some audio effects such as a reverberation, a delay, a flanger, etc. in order to reproduce Pink Floyd’s guitar sound or Eddie Van Halen famous “brown sound”. Learn interactively, guitar in hands, how to fine tune a compressor effect, or how to shape the sound of a tube guitar amp, how to get a “modern metal” or a “Jimi Hendrix” sound, using only your Web browser.

• NDlib: a python library to model and analyze diffusion processes over complex networks
  Authors: Giulio Rossetti, Letizia Milli and Salvatore Rinzivillo

  Keywords: network dynamics, diffusion, analytical software

  Abstract:
  Nowadays the analysis of dynamics of and on networks represents a hot topic in the Social Network Analysis playground. To support students, teachers, developers and researchers we introduced a novel framework, named NDlib, an environment designed to describe diffusion simulations. NDlib is designed to be a multi-level ecosystem that can be fruitfully used by different user segments. Upon NDlib, we designed a simulation server that allows remote execution of experiments as well as an online visualization tool that abstracts its programmatic interface and makes available the simulation platform to non-technicians.

• DARQL: Deep Analysis of SPARQL Queries
  Authors: Angela Bonifati, Wim Martens and Thomas Timm

  Keywords: RDF, SPARQL, Query Analysis

  Abstract:
  In this demonstration, we showcase DARQL, the first tool for deep, large-scale analysis of SPARQL queries. We have harvested a large corpus of query logs with different lineage and sizes, from DBPedia to BioPortal and Wikidata, whose total number of queries amounts to 180M. We ran a wide range of analyses on the corpus, spanning from simple tasks (keyword counts, triple counts, operator distributions), moderately deep tasks (projection test, query classification), and deep analysis (shape analysis, well-designedness, weakly well-designedness, hypertreewidth, and fractional edge cover). The key goal of our demonstration is to let the users dive into the SPARQL query logs of our corpus and let them discover the inherent characteristics of the queries. The entire corpus of SPARQL queries is stored in a DBMS. The tool has a GUI that allows users to ask sophisticated analytical queries on the SPARQL logs. These analytical queries can both be directly written in SQL or composed by a visual query builder tool. The results of the analytical queries are represented both textually (as SPARQL queries) and visually. The DBMS performs the searches within the corpus quite efficiently. To the best of our knowledge, this is the first demonstration of this kind on such a large corpus and with such a number of varied tests.

• SmartPub: A Platform for Long-Tail Entity Extraction from Scientific Publications
  Authors: Sepideh Mesbah, Alessandro Bozzon, Christoph Lofi and Geert-Jan Houben

  Keywords: Information Extraction, Document Metadata, Named Entity Recognition, Long-Tail Entity Types, Training Data Generation

  Abstract:
  This demo presents SmartPub, a novel web-based platform that supports the exploration and visualization of shallow meta-data (e.g., author list, keywords) and deep meta-data (e.g., long tail named entities which are rare, and often relevant only in specific knowledge domain) of scientific publications. The platform collects documents from different sources (e.g. DBLP and Arxiv), and extracts the domainspecific named entities from the text of the publications using a novel Named Entity Recognizer (NER) which we can train with minimal human supervision even for rare entity types. The platform further enables the interaction with the Crowd for filtering purposes or training data generation, and provides extended visualization and exploration capabilities. SmartPub will be demonstrated using sample collection of scientific publications focusing on the computer science domain and will address the entity types Dataset(i.e. dataset presented or used in a publication), and Methods(i.e. algorithms used to create/enrich/analyse a data set).

• Social Smart Meter: Identifying Energy Consumption Behavior in User-Generated Content
  Authors: Andrea Mauri, Achilleas Psyllidis and Alessandro Bozzon

  Keywords: social media, energy consumption, machine learning, social media content analysis

  Abstract:
  Having a thorough understanding of energy consumption behavior is an important element of sustainability studies. Traditional sources of information about energy consumption, such as smart meter devices and surveys, can be costly to deploy, may lack contextual information or have infrequent updates. In this paper, we examine the possibility of extracting energy consumption-related information from user-generated content. More specifically, we develop a pipeline that helps identify energy-related content in Twitter posts and classify it into four categories (dwelling, food, leisure, and mobility), according to the type of activity performed. We further demonstrate a web-based application – Social Smart Meter – that implements the proposed pipeline and enables different stakeholders gain an insight into daily energy consumption behavior, and showcase it in case studies involving several world cities.

• MetaExp: Interactive Exploration and Explanation of Large Knowledge Graphs
  Authors: Freya Behrens, Sebastian Bischoff, Pius Landenburger, Julius Rückin, Laurenz Seidel, Fabian Stolp, Michael Vaichenker, Adrian Ziegler, Davide Mottin, Fatemeh Aghaei, Emmanuel Müller, Martin Preusse, Nikola Müller and Michael Hunger

  Keywords: graph exploration, data exploration, knowledge graphs, explanation

  Abstract:
  We present MetaExp, a system that assists the user during the exploration of large knowledge graphs, given two sets of initial nodes. At its core, MetaExp presents a small set of meta-paths to the user, which are sequences of relationships among nodes. Such meta-paths do not overwhelm the user with complex structures, yet they preserve semantically-rich relationships in a graph. MetaExp engages the user in an interactive procedure, which involves simple meta-paths evaluations to infer a user-specific similarity measure. This similarity measure incorporates the domain knowledge and the preferences of the user, overcoming the fundamental limitations of previous methods based on local node neighborhoods or fixed similarity scores. Our system provides a user-friendly interface for searching initial nodes and guides the user towards progressive refinements of the meta-paths. The system is demonstrated on three datasets, one ontology, a movie database, and a biological network.

• Smart-MD: Neural Paragraph Retrieval of Medical Topics
  Authors: Rudolf Schneider, Sebastian Arnold, Tom Oberhauser, Tobias Klatt, Thomas Steffek and Alexander Löser

  Keywords: Neural Information Classification, Paragraph Retrieval, Health care information systems

  Abstract:
  We demonstrate Smart-MD, an information retrieval system for medical professionals. The system supports topical queries in the form [disease topic], such as [lyme treatments]. In contrast to document- oriented retrieval systems, Smart-MD retrieves relevant paragraphs and reduces the reading load of a doctor drastically. We recog- nize diseases and topical aspects with a novel paragraph retrieval method based on bidirectional LSTM neural networks. We demon- strate Smart-MD on a dataset that contains 3,469 diseases from the English language part of Wikipedia and 6,876 distinct medical aspects extracted from Wikipedia headlines.

• SenHint: A Joint Framework for Aspect-level Sentiment Analysis by Deep Neural Networks and Linguistic Hints
  Authors: Yanyan Wang, Qun Chen, Xin Liu, Murtadha Ahmed and Zhanhuai Li

  Keywords: Deep neural networks, Linguistic hints, Aspect-level sentiment analysis

  Abstract:
  The state-of-the-art techniques for aspect-level sentiment analysis focus on feature modeling using a variety of deep neural networks (DNN). Unfortunately, their practical performance may fall short of expectations due to semantic complexity of natural languages. Motivated by the observation that linguistic hints (e.g. explicit sentiment words and shift words) can be strong indicators of sentiment, we present a joint framework, SenHint, which integrates deep neural networks and linguistic hints into a coherent reasoning model based on Markov Logic Network (MLN). In SenHint, linguistic hints are used in two ways: (1) to identify easy instances, whose sentiment can be automatically determined by machine with high accuracy; (2) to capture implicit relations between aspect polarities. We also empirically evaluate the performance of SenHint on both English and Chinese benchmark datasets. Our experimental results show that SenHint can effectively improve accuracy compared with the state-of-the-art alternatives.

• A journey from the Physical Web to the Physical Semantic Web
  Authors: Michele Ruta, Floriano Scioscia, Giuseppe Loseto, Filippo Gramegna, Saverio Ieva, Agnese Pinto and Eugenio Di Sciascio

  Keywords: Semantic Web of Things, Physical Web, Resource discovery, Non-standard reasoning

  Abstract:
  The Physical Semantic Web (PSW) is a novel paradigm built upon the Google Physical Web (PW) approach and devoted to improve the quality of interactions in the Web of Things. Beacons expose semantic annotations instead of basic identifiers, i.e., machine-understandable descriptions of physical resources lay the foundations for novel ontology-based object advertisement and discovery. They in turn enable advanced user-to-thing and autonomous thing-to-thing interactions. The demo shows the evolution from the PW to the PSW in a discovery scenario set in a winery, where bottles are equipped with Bluetooth Low Energy beacons and a customer can discover them using her smartphone. It demonstrates benefits of PSW over basic PW, including: rich semantic-based object annotation; dynamic annotations exploiting on-board sensors; enhanced discovery and ranking of nearby objects through semantic matchmaking; availability of interactions even without working Internet infrastructure, by means of point-to-point data exchanges.

• Finding Tours for a Set of Interests
  Authors: Mohamed Abdel Maksoud, Gaurav Pandey and Shuaiqiang Wang

  Keywords: Information Retrieval, Ranking, Search Result Organization

  Abstract:
  This paper addresses a novel tour discovery problem in the domain of travel search. We create a ranking of tours for a set of travel interests, where a tour is a group of city documents and a travel interest is a query. While generating and ranking tours, it is aimed that each interest (from the interest set) is satisfied by at least one city in a tour and the distance traveled to cover the tour is not too large. Firstly, we generate tours for the interest set, by utilizing the available ranking of cities for the individual interests and the distances between the cities. Then, in absence of existing methods directly related to our problem, we devise our novel techniques to calculate ranking scores for the tours and present a comparison of these techniques in our results. We demonstrate our web application Travición, that utilizes the best tour scoring technique.

• BarcelonaNow: Empowering Citizens with Interactive Dashboards for Urban Data Exploration
  Authors: Mirko Marras, Matteo Manca, Ludovico Boratto, Gianni Fenu and David Laniado

  Keywords: Data Exploration, Data Visualization, Urban Dashboard

  Abstract:
  The advent of massively interconnected technologies over cities raises equally massive challenges regarding the end-user interfaces that enable citizens to make sense of urban data for improving their daily life and for participation and decision making. The existing dashboards include only pre-defined and limited use cases which can only address the most common needs of citizens, but do not allow for personalization. As consequence, the great effort of cities to make data widely available has still scarce capacity to get an impact on the public good. In this paper, we propose an open source dashboard with a set of tools and services which enables citizens to easily create and explore interactive visualizations of city-related data. Moreover the user can personalize the dashboard based on their individual, local, task-specific goals and interests, and share the resulting dashboard to promote co-creation. With it, citizens can build up a data-driven public awareness, supporting an open, transparent, and collaborative city, where they are actively involved in local activities and issues.

• GeoSensor: On-line, Scalable Change and Event Detection over Big Data
  Authors: Giorgos Argyriou, George Papadakis, Argyros Argyridis, Nikiforos Pittaras, George Giannakopoulos, Sergio Albani, Michele Lazzarini, Emanuele Angiuli, Anca Popescu and Manolis Koubarakis

  Keywords: Big Data, Image processing, Event Detection, Semantic Web

  Abstract:
  We present GeoSensor, a novel system that enriches Change Detection over Earth Observation products (i.e., satellite images) with Event Detection over news items and social media content. GeoSensor faces the major challenges of Big Data: Volume (a single satellite image typically occupies few GBs), Variety (its data sources include two different types of satellite images and various types of user-generated content) and Veracity, as the accuracy of the end result is crucial for the usefulness of our system. To overcome these three Vs, while offering an on-line functionality, GeoSensor comrpises a complex architecture that is based on the open-source platform “Big Data Infrastructure”. Through our demonstration, we highlight both the effectiveness and the efficiency of GeoSensor’s functionalities.

• Etymo: A New Discovery Engine for AI Research
  Authors: Weijian Zhang, Jonathan Deakin, Nicholas Higham and Shuaiqiang Wang

  Keywords: web search, content analysis, similarity-based network, graph centrality, data visualisation

  Abstract:
  We present Etymo (https://etymo.io), a discovery engine to facilitate artificial intelligence (AI) research and development.It aims to help readers navigate a large number of AI-related papers published every week by using a novel form of search that finds relevant papers and displays related papers in a graphical interface. Etymo constructs and maintains an adaptive similarity-based network of research papers as an all-purpose knowledge graph for ranking, recommendation, and visualisation. The network is constantly evolving and can learn from user feedback to adjust itself. We will demonstrate Etymo’s search and feed engines in a web browser on a laptop (wireless internet access is needed). A screencast is available at: https://youtu.be/T4FDPk_TmN0

• VIZ-Wiki: Generating Visual Summaries to Factoid Threads in Community Question Answering Services
  Authors: Tanya Chowdhury, Aashay Mittal and Tanmoy Chakraborty

  Keywords: Summarization, Community question answering, Factoid questions

  Abstract:
  In this demo, we present VIZ-Wiki, a browser extension which generates an overview of summarizable threads in Question Answering forums. It reduces a user’s effort to go through lengthy text-based, sarcastic and highly critiqued answers. Our tool can be used to collect community opinion from popular discussion sites like Quora, Yahoo! Answers, Reddit etc. as well as topic-centric ones such as Askubuntu, Stackoverflow. We rely on textual information of these forums to extract insightful summaries for a reader. VIZ-Wiki provides users a pie-graph view marking popular choices when such a question link is raised. A button further guides them to detailed statistics and relevant list of answers. VIZ-Wiki deals with answers contradicted by other users, prioritizes highly-recommended ones and avoids sarcasm. We test our model on the factoid questions on a dataset of Yahoo! Answers and obtain a macro precision of 0.6 on displayed answers and a macro recall of 0.69, beating the baseline significantly. To the best of our knowledge, VIZ-Wiki is the first attempt to generate automated list of answers for questions in community question answering services. In the spirit of reproducibility, we have released the code and a demonstration video public at http://goo.gl/cyx3EF and http://goo.gl/4p855a respectively.

• Assessing the News Landscape: A Multi-Module Toolkit for Evaluating the Credibility of News
  Authors: Benjamin D. Horne, William Dron, Sara Khedr and Sibel Adali

  Keywords: content analysis, information credibility, computational journalism, machine learning, news analysis, human-centered computing

  Abstract:
  Today, journalist, information analyst, and everyday news consumers are tasked with discerning and fact-checking the news. This task has became complex due to the ever-growing number of news sources and the mixed tactics of maliciously false sources. To mitigate these problems, we introduce the The News Landscape (NELA) Toolkit: an open source toolkit for the systematic exploration of the news landscape. NELA allows users to check the credibility of news articles using content-based prediction techniques, as well as, filter and sort through article predictions based on the user’s own needs. In addition, NELA allows users to visualize the media landscape at different time slices using a variety of features computed at the source level. NELA is built with a modular, pipeline design, to allow researchers to add new tools to the toolkit with ease. Our demo is an early transition of automated news credibility research to assist human fact-checking efforts and increase the understanding of the news ecosystem as a whole.

• VideoKen: Automatic Video Summarization and Course Curation to Support Learnin
  Authors: Debabrata Mahapatra, Ragunathan Mariappan, Vaibhav Rajan and Kuldeep Yadav

  Keywords: Videoken, Video Summarization, Course Curation, Table of Contents, Phrase Cloud, Kenlist

  Abstract:
  The number of high quality online videos is increasing rapidly. Online courses as well as universities do not fully leverage the content due to several open challenges in video search, indexing, summarization and customization requirements for specific courses, instructors or learners. We present a new web-based social learning platform called Videoken. Using novel video summarization algorithms, Videoken automatically creates Table of Contents for videos. This allows a textbook-like facility of non-linear search and navigation through the video, enables extraction of semantically coherent clips from within a video and improves video search through better semantic indexing. The platform also allows new ways of course creation and sharing of learning modules; and can be both integrated with existing Learning Management Systems and used independently.

• EMOFIEL: Mapping Emotions of Relationships in a Story
  Authors: Harshita Jhavar and Paramita Mirza

  Keywords: emotion analysis, fictional character relationships, fictional narratives

  Abstract:
  We present EMOFIEL: a system that identifies characters and scenes in a story from a fictional narrative summary, generates appropriate scene descriptions, identifies the emotion flow between a given directed pair of story characters for each interaction, and organizes them along the story timeline to populate a knowledge base about stories. These emotions are identified using two emotion modelling approaches: categorical and dimensional emotion models. The generated plots clearly show that in a particular scene, two characters can share multiple emotions together with different intensity. Furthermore, the directionality of the emotion can be captured as well. EMOFIEL provides a web-based GUI that allows users to query the constructed knowledge base to explore the emotion mapping of a given character pair throughout a given story, and to explore scenes for which a certain emotion peaks.