Topic Detection and Tracking is an event-based information organization task where online news streams are monitored in order to spot new unreported events and link documents with previously detected events. The detection has proven to perform rather poorly with traditional information retrieval approaches. We present an approach that formalizes temporal expressions and augments spatial terms with ontological information and uses this data in the detection. In addition, instead using a single term vector as a document representation, we split the terms into four semantic classes and process and weigh the classes separately. The approach is motivated by experiments.

The harnessing of time-related information from text for the use of information retrieval requires a leap from the surface forms of the expressions to a formalized time-axis. Often the expressions are used to form chronological sequences of events. However, we want to be able to determine the temporal similarity, i.e., the overlap of temporal references of two documents and use this similarity in Topic Detection and Tracking, for example. We present a methodology for extraction of temporal expressions and a scheme of comparing the temporal evidence of the news documents. We also examine the behavior of the temporal expressions and run experiments on English News corpus.

New Event Detection (NED) involves monitoring chronologically-ordered news streams to automatically detect the stories that report on new events. We compare two stories by finding three cosine similarities based on names, topics and the full text. These additional comparisons suggest treating the NED problem as a binary classification problem with the comparison scores serving as features. The classifier models we learned show statistically significant improvement over the baseline vector space model system on all the collections we tested, including the latest TDT5 collection. The presence of automatic speech recognizer (ASR) output of broadcast news in news streams can reduce performance and render our named entity recognition based approaches ineffective. We provide a solution to this problem achieving statistically significant improvements. 1

Abstract: A new event tracking approach is proposed based on the identification of named entity (NE) patterns such as Who, What, Where and When, and their relationship with news domains such as Politics, Economy, Government and Entertainment. This research comprises three parts. The first part uses a set of user studies to identify NE patterns and their relationship with news domains. Second part is to design a prototype system based on NE patterns. The final part evaluates the prototype event tracking system. This paper described the first part which is to evaluate the importance of NE across news domains. We have achieved a better understanding on NE patterns by identifying the distribution of NE across news domains.

Abstract. Today, traders in financial markets are confronted with the problem that information is distributed over diverse sources and that there is too much information available. In our work we develop methods and tools to help traders to overcome this information overload by enabling the integrated view on news from various sources, by filtering relevant news and by providing decision support for traders. Another goal of our work is to propose a formal model of the impact of news on asset prices and thus enable better predictions of stock prices than possible with purely text mining based approaches. 1 Research Problem Traders in financial markets are confronted with the problem that too much information is available from various, heterogenuous sources like newswires, forums, blogs and collaborative tools. In order to make accurate trading decisions, traders have to filter the relevant information efficiently so that they are able to react to new information in a timely manner.

2010 ‘This thesis is the result of the author’s original research. It has been composed by the author and has not been previously submitted for examination which has led to the award of a degree.’ 'The copyright of this thesis belongs to the author under the terms of the United Kingdom Copyright Acts as qualified by University of Strathclyde Regulation 3.50. Due acknowledgement must always be made of the use of any material contained in, or derived from, this thesis.’ Signed: Date: Acknowledgements My PhD is a challenging journey with wonderful experiences. Having two maternity leaves is one of it and I’m very lucky to be surrounded by truly lovely people. Early in the process of completing this project, it became quite clear to me that a researcher cannot complete a PhD thesis alone. Therefore I would like to thank the following persons for their dedication, prayers and support. I would like to express my deep and sincere gratitude to my supervisors, Professor Fabio Crestani and Professor Ian Ruthven. I am grateful to them for their commitment, the freedom they gave me to pursue my ideas, the encouragement they provided when I succeeded, the patience they demonstrated when I failed, the wide range of problems they exposed me to, and the direction they consistently provided. Some people wondered as to how I completed my PhD during my supervisor’s absence. Although Fabio is currently at University of Lugano, Switzerland, but I never felt like struggling alone. Thank you for being a great supervisor. I would also like to thank Dr. Crawford Revie for his useful comments and encouragement at my annual reviews. I would like to thank members of the i-lab group for making my journey a pleasant one.

This thesis contributes to the field of complex event-data analysis novel and formally well-founded methods for similarity searching, both on the level of single events and on the level of sequences of events. As event-based systems may produce highly diverse data sets, the main focus of our considerations is on highest possible flexibility. Also, the approaches shall be intelligible to business analysts and, of course, generate meaningful and intuitive results. Finally, the approaches shall be conceptually independent from concrete Complex Event Processing solutions and instead build upon abstract and generally accepted definitions of events, event types, etc. Our approach on single-event similarity builds upon geometric ideas of similarity, with event attribute values defining the relative positioning of two events in an n-dimensional space. Thereby, the similarity between two events is calculated from weighted attribute-level similarities. The proposed approach on event-sequence similarity outperforms existing approaches by allowing analysts to consider event-level similarities, order, and relative and absolute temporal structures in a highly flexible manner. It builds upon an assignment-based understanding of sequence similarity, where each unit of the pattern sequence is considered either represented by a certain event of the target sequence or missing therein.