Enrique Alfonseca is a research scientist at Google Research Zu- rich where he current- ly manages a team working on language understanding. During the past four years he has been a member of areas of ads quality, search quality and natural language understanding, contributing in areas such as query expansion and relevance estimation for sponsored search, ranking for web search, information extraction, unsupervised semantic parsing, lexical semantics and automatic text summarization. He received a Ph.D. in Computer Science from Universidad Autónoma de Madrid in 2003, and he has over 60 research publications, mainly in the fields of language understanding and information retrieval.
News Understanding for Knowledge Graph Freshness
In this talk I will describe our past and current work on event understanding from news, including two different systems and architectures for learning paraphrases of event patterns that we use for news understanding and headline generation from news collections. From a web-scale corpus of English news, we mine syntactic patterns that a generative model generalizes into event descriptions. At inference time, we query the model with the patterns observed in an unseen news collection, identify the event that better captures the gist of the collection and automatically produce updates for the knowledge graph, and retrieve the most appropriate pattern to generate a headline. The talk will focus on the event understanding and information extraction sides of the system, lessons learned in the past year, and the main challenges we see moving forward.
Peter Mika is a Senior Research Scientist at Yahoo!, based in Barce- lona, Spain. Peter is working on the applica- tions of semantic tech- nology to Web search. He received his MSc and PhD in computer science (summa cum laude) from Vrije Universiteit Amsterdam. He is the author of the book ‘Social Networks and the Semantic Web’ (Springer, 2007). In 2008 he has been selected as one of “AI’s Ten to Watch” by the editorial board of the IEEE Intelligent Systems journal. Peter is a regular speaker at both academic and technology conferences and serves on the advisory board of a number of public and private initiatives. He represents Yahoo! in the leadership of the schema.org collaboration with Google, Bing, and Yandex.
Making the Web Searchable
The key idea of the Semantic Web is to make information on the Web easily consumable by machines. As machines start to understand web pages as sources of data, search on the Web will move well beyond the current paradigm of retrieving pages by keywords. Instead, search engines will start to answer complex queries based on the cumulative knowledge of the Web. In this presentation, we will review the brief history of Semantic Search in academic research and in developments across the search industry. We also look ahead to highlight the research challenges that have surfaced and remain unsolved.
View the slides
Mike Thelwall is head of the Statistical Cybermetrics Research Group at the Univer- sity of Wolverhampton and is a professor of information science. He is also a research associate at the Oxford Internet Institute, Oxford University and a docent in the Department of Information Studies, Åbo Akademi University, Finland. He is an associate editor of the Journal of the Association for Information Science and Technology and sits on three other journal editorial boards. He has published over 230 refereed journal articles, 26 book chapters and two books. His PhD is in Pure Maths from the University of Lancaster (1989) but since then he has researched information from a quantitative social sciences perspective within the information science discipline. His sentiment analysis research has produced the sentiment strength detection software SentiStrength (http://sentistrength.wlv.ac.uk), which is sold commercially as well as being used for academic research. He has also developed free software for gathering and analysing tweets (http://mozdeh.wlv.ac.uk) and for gathering and analysing various types of web data (http://lexiurl.wlv.ac.uk). He has led Wolverhampton’s contribution to four EU projects, eight other international projects and five UK projects as well as providing consultancy to the UK government, the UN and the EU on web-based evaluation.
Sentiment Strength Detection for Social Media Text: Artificial Agents, Answer Ranking and Art Installations
This talk will describe a simple, fast, intuitive and flexible lexical method to detect sentiment strength in short informal text and will illustrate it with a range of research and commercial applications. Implemented in the software SentiStrength and optimised for social web text, the method can process 14,000 tweets per second with human level accuracy in many cases. The talk will describe and demonstrate the method and its evaluations and show how it has been translated from the original English version to many other languages. It is free for researchers to use and its current applications include art installations, answer ranking, opinion mining, stock market prediction, customer feedback analysis, and sentiment-aware artificial agents.
Download the slides