Two systems for spoken document classification are implemented by combining an automatic speech recognizer with the two classification algorithms naive Bayes and logistic regression. The focus is on how to handle the inherent uncertainty in the output of the speech recognizer. Feature extraction is performed by computing expected word counts from speech recognition lattices, and subsequently removing words that are found to carry little or noisy information about the topic label, as determined by the information gain metric. The systems are evaluated by performing cross-validation on broadcast news stories, and the classification accuracy is measured with different configurations and on recognition output with different word error rates. The results show that a relatively high classification accuracy can be obtained with word error rates around 50%, and that the benefit of extracting features from lattices instead of 1-best transcripts increases with increasing word error rates.

Communication technologies are continuosly evolving since were invented the telegraph. Nowadays, one of the most interesting and relatively youngest field of telecommunications are the optical communications. The search of the shortest event ever created to measure very fast/short events is now possible because of the Light Amplification by Stimulated Emission of Radiation, longer known as laser.So well, we need to find the shortest event but we need to know how to measure it, since it is supposed that there is not a shorter event. So we need the pulse to measure itself, and its sound as we need to do a kind of autocorrelation.In order to measure a pulse, we need to get its intensity and phase in either the time or frequency domains. And, before the Frequency-Resolved Optical-Gating appearance, it was done with autocorrelation (time-domain) and spectrum (frequency-domain).

Road traffic has seen a constant augmentation these last decades. The noise it generates has grown more or less linearly with the traffic and has created a major environmental problem. It affects the human body in negative ways by engendering sleep disorders, stress and even cardiovascular diseases. The Norwegian government came with regulations to reduce the number of people exposed to high noise levels and many people have now had their habitation supplemented with noise-reducing measures. There is however a lack of researches made on these noise control measures in the indoor noise; especially on how the residents experience these modifications. A telephone-based survey is done in this research where 76 households have been selected on the parcel from Gardemoen to Biri on the new E6 in Norway. Between these habitations, the average outdoors and indoor noise levels are, respectively, 61dB and 34dB prior the installation of any noise reducing measures. The results, treated statistically using the software SPSS, are showing the nuisance degree experienced by the habitants for different noise levels both outdoors and indoors. Additionally, the subjective improvement of the noise situation felt by the habitants after the measures shows a rather different picture outdoor and indoor, as the amelioration is generally bigger inside the habitation.Indeed, while nearly 90% of the respondents felt annoyed to extremely annoyed outdoors only 50% showed the same nuisance levels after the noise control measures were installed. For the indoors situation 40% of the participants of the survey felt annoyed to very annoyed prior the measures while only 5% of them felt the same degree of annoyance after. Furthermore, while 42% were sleep disturbed and 50% experienced stress and 66% felt a reduction of their well-being only 13%, 20% and 18% felt the same health issues after the noise-measures were installed. Around half of the interviewees declared they were satisfied with both Sweco and Statens Vegvesen which were key firms for the planning and building of the measures along the chosen parcel. Finally, taking all this into account, it appears that 40% of the 76 selected residents are satisfied with the noise-reducing measures, 10% are unsatisfied and the rest is neutrally satisfied.

Dette arbeidet har blitt utført med et mål om å utvikle et system for automatisk gjenkjenning av dialekter basert på akustisk modellering. Systemet ble utviklet ved bruk av Hidden Markov ToolKit (HTK), et samling av ferdige språkmodelleringsverktøy utviklet av Cambridge University Engineering Department i 1996. Trening og evaluering ble gjort for to dialektformer innenfor de tre språkene spansk, engelsk og mandarin.

Short Message Service (SMS) has become a very commonly used service. It does not only work as a substitute for voice telephony, but is also used for automated services. Some of these service are related to security issues like SMS banking, or one time passwords, even though SMS messages can be spoofed or eavesdropped.We propose a design where we add security to SMS by making an easily configurable module that utilizes a fast cryptographic scheme called Elliptic Curve Signcryption. To prove our concept, we implement an SMS client for Android smart phones that utilizes our security module and serves end-to-end data protection of SMS messages with the same security level as Top Secret content.

Today, major IT-companies, such as Microsoft, Amazon and Google, are offering online storage services to their customers. This is a favourable solution -- as opposed to regular storage -- in terms of low costs, reliability, scalability and capacity. However, important security features such as data privacy and integrity are often absent.To address these issues, a cryptographic architecture is proposed that ensures the confidentiality and integrity of the data stored by users, independent of the trust of the provider. This includes secure sharing of private data among users of the same service.The underlying cryptographic architecture is based on existing open source systems and cryptographic primitives. The architecture was implemented as a reusable general library in Java. An Android client was created, and several performance tests were conducted. The proof of concept system shows that it is possible to implement the proposed scheme, and that the cryptographic operations does not significantly affect the user experience on an Android device. Possible weaknesses of the scheme are identified.We present a scheme for secure storage and sharing of files on an untrusted server, and argue for its validity. To support streaming functionality, the scheme could be extended with hash trees to validate small parts of a file at the time.

Performance analysis and comparison of different techniques used for anomaly detection in packet switched backbone networks.

This report focuses on the quality aspects of media delivery over the Internet. Weinvestigate the constructs of Knowledge Plane, Monitor Plane and Action Planeas controlling functions for the Internet. Our goal is to implement functionality formonitoring services in a home network, allowing the router to reason and take actionsto obtain an optimal traffic situation based on user preferences. The actions takento alter ongoing traffic are implemented in a modular router framework called Click.We will use this router to affect the media stream TCP connections into behavingin accordance with the networks optimal state. New features are implemented tocomplement the functionality found in Click, giving us the tools needed to obtainthe wanted results.Our focus is on adaptive video streaming in general and Silverlight SmoothStreaming in particular. Using custom Silverlight client code, we implemented asolution which allows the applications to report usage statistics to the home gateway.This information will be used by the home gateway to obtain an overview of traffic inthe network. Presenting this information to the user, we retrieve the user preferencesfor the given video streams. The router then dynamically reconfigures itself, andstarts altering TCP packets to obtain an optimal flow of traffic in the home network.Our system has been implemented on a Linux PC where it runs in its currentform. All the different areas of the solution, ranging from the clients, router, Knowl-edge Plane and traffic manipulation elements are put together. They form a workingsystem for QoE/QoS optimization which we have tested and demonstrated. In ad-dition to testing the concept on our own streaming services, the reporting featurefor Silverlight clients has also been implemented in a private build of TV2 Sumo.This is the Internet service of the largest commercial television station in Norway.Further testing with the TV2 Sumo client has given promising results. The systemis working as it is, although we would like to see more complex action reasoning toimprove convergence time for achieving the correct bit rate.