Sentence Compression with Supervised Machine Learning

The subject of the thesis is automatic sentence compression with machine learning, so that the compressed sentences remain both grammatical and retain their essential meaning. There are multiple possible uses for the compression of natural language sentences. In this thesis the focus is generation of television program subtitles, which often are compressed version of the original script of the program. The main part of the thesis consists of machine learning experiments for automatic sentence compression using different approaches to the problem. The machine learning methods used for this work are linear-chain conditional random fields and support vector machines. Also we take a look which automatic text analysis methods provide useful features for the task. The data used for machine learning is supplied by Lingsoft Inc. and consists of subtitles in both compressed an uncompressed form. The models are compared to a baseline system and comparisons are made both automatically and also using human evaluation, because of the potentially subjective nature of the output. The best result is achieved using a CRF - sequence classification using a rich feature set. All text analysis methods help classification and most useful method is morphological analysis. Tutkielman aihe on suomenkielisten lauseiden automaattinen tiivistäminen koneellisesti, niin että lyhennetyt lauseet säilyttävät olennaisen informaationsa ja pysyvät kieliopillisina. Luonnollisen kielen lauseiden tiivistämiselle on monta käyttötarkoitusta, mutta tässä tutkielmassa aihetta lähestytään television ohjelmien tekstittämisen kautta, johon käytännössä kuuluu alkuperäisen tekstin lyhentäminen televisioruudulle paremmin sopivaksi. Tutkielmassa kokeillaan erilaisia koneoppimismenetelmiä tekstin automaatiseen lyhentämiseen ja tarkastellaan miten hyvin erilaiset luonnollisen kielen analyysimenetelmät tuottavat informaatiota, joka auttaa näitä menetelmiä lyhentämään lauseita. Lisäksi tarkastellaan minkälainen lähestymistapa tuottaa parhaan lopputuloksen. Käytetyt koneoppimismenetelmät ovat tukivektorikone ja lineaarisen sekvenssin mallinen CRF. Koneoppimisen tukena käytetään tekstityksiä niiden eri käsittelyvaiheissa, jotka on saatu Lingsoft OY:ltä. Luotuja malleja vertaillaan Lopulta mallien lopputuloksia evaluoidaan automaattisesti ja koska teksti lopputuksena on jossain määrin subjektiivinen myös ihmisarviointiin perustuen. Vertailukohtana toimii kirjallisuudesta poimittu menetelmä. Tutkielman tuloksena paras lopputulos saadaan aikaan käyttäen CRF sekvenssi-luokittelijaa laajalla piirrejoukolla. Kaikki kokeillut teksin analyysimenetelmät auttavat luokittelussa, joista tärkeimmän panoksen antaa morfologinen analyysi.

Proximity Based Authentication in IoT

IoT consists of essentially thousands of tiny sensor nodes interconnected to the internet, each one of which executes the programmed functions under memory and power limita- tions. The sensor nodes are distributed mainly for gathering data in various situations. IoT envisions the future technologies such as e-health, smart city, auto-mobiles automa- tion, construction sites automation, and smart home. Secure communication of data under memory and energy constraints is major challenge in IoT. Authentication is the first and important phase of secure communication. This study presents a protocol to authenticate resource constraint devices in physical proximity by solely using the shared wireless communication interfaces. This model of authentication only relies on the abundance of ambient radio signals to authenticate in less than a second. To evaluate the designed protocol, SkyMotes are emulated in a network environment simulated by Contiki/COOJA. Results presented during this study proves that this approach is immune against passive and active attacks. An adversary located as near as two meters can be identified in less than a second with minimal expense of energy. Since, only radio device is used as required hardware for the authentication, this technique is scalable and interoperable to heterogeneous nature of IoT.

Oletko iloinen kuin peipponen? – Persoonallisuusmittari NEO-PI-3:n suomennoksessa, ruotsinnoksessa ja ranskannoksessa käytetyt käännösstrategiat

Yhdysvaltapainotteisen psykologisen tutkimuksen takia psykologiset testit on useimmiten laadittu englanniksi ja amerikkalaiseen kohdekulttuuriin, ja tästä syystä ne pitää kääntää ja adaptoida eurooppalaiseen kohdekulttuuriimme. Tämän pro gradu -tutkielman tarkoituksena oli selvittää, mitä käännösstrategioita Costan ja McCraen (1992) NEO-PI-3-persoonallisuusmittarin suomen-, ruotsin- ja ranskankielisissä käännöksissä on käytetty. Kvalitatiivisessa tarkastelussa eri kieliversioissa esiintyneitä käännösratkaisuja vertailtiin Jan Pedersenin (2007; 2005) käännösstrategialuokitukseen perustuvan luokittelun avulla, jossa strategiat on jaettu lähde- ja kohdetekstiorientoituneisiin ratkaisuihin. Oletin tutkimushypoteesissani, että NEO-PI-3-persoonallisuusmittarin käännöksissä on käytetty enemmän kotouttavia eli kohdetekstiorientoituneita kuin vieraannuttavia eli lähdetekstiorientoituneita strategioita. Oletuksen vastaisesti tutkimustulokset osoittivat, että yleisimmin käytetty strategia oli lähdetekstiuskollinen suora käännös. Suomen- ja ruotsinkielisissä versioissa vieraannuttavia käännösratkaisuja esiintyi myös yhteenlaskettuna kotouttavia ratkaisuja useammin. Hypoteesia tuki kuitenkin tutkimustulos, jonka mukaan persoonallisuusmittarin ranskannoksessa kotouttavia strategioita oli käytetty vieraannuttavia ahkerammin. Yhteenvetona voi sanoa, että käytettyjen strategioiden kirjo oli laaja ja vaihteli tarkasteluun valitsemieni teemojen sisällä sekä kieliversioittain. Monenlaisilla käännösratkaisuilla voidaan saavuttaa käännös, joka on ekvivalentti ja toimiva uudessa kohdekulttuuriympäristössään. Kääntäjät kohdekielen ja -kulttuurin tuntijoina voivat tuoda psykologisten testimetodien kehittelyyn arvokasta asiantuntijuutta tarvittavan psykologisen tietämyksen lisäksi.

Heat-transfer improvements in an axial-flux permanent-magnet synchronous machine

Axial-flux machines tend to have cooling difficulties since it is difficult to arrange continuous heat path between the stator stack and the frame. One important reason for this is that no shrink fitting of the stator is possible in an axial-flux machine. Using of liquid-cooled end shields does not alone solve this issue. Cooling of the rotor and the end windings may also be difficult at least in case of two-stator-single-rotor construction where air circulation in the rotor and in the end-winding areas may be difficult to arrange. If the rotor has significant losses air circulation via the rotor and behind the stator yokes should be arranged which, again, weakens the stator cooling. In this paper we study a novel way of using copper bars as extra heat transfer paths between the stator teeth and liquid cooling pools in the end shields. After this the end windings still suffer of low thermal conductivity and means for improving this by high-heat-conductance material was also studied. The design principle of each cooling system is presented in details. Thermal models based on Computational Fluid Dynamics (CFD) are used to analyse the temperature distribution in the machine. Measurement results are provided from different versions of the machine. The results show that significant improvements in the cooling can be gained by these steps.

Evaluation of Machine Learning Classifiers for Mobile Network Intrusion Detection Systems

Mobile malwares are increasing with the growing number of Mobile users. Mobile malwares can perform several operations which lead to cybersecurity threats such as, stealing financial or personal information, installing malicious applications, sending premium SMS, creating backdoors, keylogging and crypto-ransomware attacks. Knowing the fact that there are many illegitimate Applications available on the App stores, most of the mobile users remain careless about the security of their Mobile devices and become the potential victim of these threats. Previous studies have shown that not every antivirus is capable of detecting all the threats; due to the fact that Mobile malwares use advance techniques to avoid detection. A Network-based IDS at the operator side will bring an extra layer of security to the subscribers and can detect many advanced threats by analyzing their traffic patterns. Machine Learning(ML) will provide the ability to these systems to detect unknown threats for which signatures are not yet known. This research is focused on the evaluation of Machine Learning classifiers in Network-based Intrusion detection systems for Mobile Networks. In this study, different techniques of Network-based intrusion detection with their advantages, disadvantages and state of the art in Hybrid solutions are discussed. Finally, a ML based NIDS is proposed which will work as a subsystem, to Network-based IDS deployed by Mobile Operators, that can help in detecting unknown threats and reducing false positives. In this research, several ML classifiers were implemented and evaluated. This study is focused on Android-based malwares, as Android is the most popular OS among users, hence most targeted by cyber criminals. Supervised ML algorithms based classifiers were built using the dataset which contained the labeled instances of relevant features. These features were extracted from the traffic generated by samples of several malware families and benign applications. These classifiers were able to detect malicious traffic patterns with the TPR upto 99.6% during Cross-validation test. Also, several experiments were conducted to detect unknown malware traffic and to detect false positives. These classifiers were able to detect unknown threats with the Accuracy of 97.5%. These classifiers could be integrated with current NIDS', which use signatures, statistical or knowledge-based techniques to detect malicious traffic. Technique to integrate the output from ML classifier with traditional NIDS is discussed and proposed for future work.

Prospects of Bitcoin diffusion

Currency is something people deal with every day in their lives. The contemporary society is very much revolving around currencies. Even though technological development has been rapid, the principle of currency has stayed relatively unchanged for a long time. Bitcoin is a digital currency that introduced an alternative to other digital currencies, and to the traditional physical currencies. Bitcoin is peer-to-peer, open source, and it erases the need of a third party in transactions. Bitcoin has since inception gained certain fame, but it has not established itself as a common currency in the world. The purpose of this study was to analyse what kind of potential does Bitcoin have to become a widely accepted currency in day-to-day transactions. The main research question was divided into three sub questions: • What kind of a process is the diffusion of new innovations? • What kinds of factors speak for the wider adoption of Bitcoin? • What kinds of factors speak against the wider adoption of Bitcoin? The purpose of the study was approached by having diffusion of innovations as the theoretical framework. The four elements in diffusion of innovations are, innovation, communication, time, and social system. The theoretical framework is applied to Bitcoin, and the research questions answered by analysing Bitcoin’s potential diffusion prospects. The body of research data consisted of media texts and statistics. In this study, content analysis was the research method. The main findings of the study are that Bitcoin has clear strengths, but it faces a large amount of uncertainty. Bitcoin’s strong areas are the transactions. They are fast, easy, and cheap. From the innovation diffusion perspective Bitcoin is still relatively unknown, and the general public’s attitudes towards it are sceptical. The research findings purport that Bitcoin has potential demand especially when the financial system of a region is dysfunctional, or when there is a financial crisis. Bitcoin is not very trusted, and the majority of people do not see a reason to start using Bitcoin in the future. A large number of people associate it with illegal activities. In general people are largely unaware of what Bitcoin is or what are the strengths and weaknesses. Bitcoin is an innovative alternative currency. However, unless people see a major need for Bitcoin due to a financial crisis, or dysfunctionality in the financial system, Bitcoin will not become much more widespread as it is today. Bitcoin’s underlying technology can be harnessed to multiple uses. Developments in that field in the future are something that future researchers could look into.

Design and implementation of machine vision system for the mobile assembly robot

The review of intelligent machines shows that the demand for new ways of helping people in perception of the real world is becoming higher and higher every year. This thesis provides information about design and implementation of machine vision for mobile assembly robot. The work has been done as a part of LUT project in Laboratory of Intelligent Machines. The aim of this work is to create a working vision system. The qualitative and quantitative research were done to complete this task. In the first part, the author presents the theoretical background of such things as digital camera work principles, wireless transmission basics, creation of live stream, methods used for pattern recognition. Formulas, dependencies and previous research related to the topic are shown. In the second part, the equipment used for the project is described. There is information about the brands, models, capabilities and also requirements needed for implementation. Although, the author gives a description of LabVIEW software, its add-ons and OpenCV which are used in the project. Furthermore, one can find results in further section of considered thesis. They mainly represented by screenshots from cameras, working station and photos of the system. The key result of this thesis is vision system created for the needs of mobile assembly robot. Therefore, it is possible to see graphically what was done on examples. Future research in this field includes optimization of the pattern recognition algorithm. This will give less response time for recognizing objects. Presented by author system can be used also for further activities which include artificial intelligence usage.

Interoperability for IoT Health Systems Using HL7 Standard

Internet of Things or IoT is revolutionizing the world we are living in, similarly the way Internet and the web did few decades ago. It is changing how we interact with the things surrounding us. Electronic health and remote patient monitoring are the ways of utilizing these technological improvements towards the healthcare. There are many applications of IoT in eHealth such as, it will open the gate to provide healthcare to the remote areas of the world, where healthcare through traditional hospital systems cannot be provided. To connect these new eHealth IoT systems with the existing healthcare information systems, we can use the existing interoperability standards commonly used in healthcare information systems. In this thesis we implemented an eHealth IoT system based on Health Level 7 interoperability standard for continuous data transmission. There is not much previous work done in implementing the HL7 for continuous sensor data transmission. Some of the previous work was limited to sensors which are not continuous in nature and some of it is only theatrical architecture. This thesis aims to prove that it is possible to implement an eHealth IoT system by using sensors which require continues data transmission, such as respiratory sensors, and to connect it with the existing eHealth information system semantically by using HL7 interoperability standard. This system will be beneficial in implementing eHealth IoT systems for those patients, who requires continuous healthcare personal monitoring. This includes elderly people and patients, whose health need to be monitored constantly. To implement the architecture, HL7 v2.5 is selected due to its ease of implementation and low size. We selected some open source technologies because of their open licenses and large developer community. We will also review the most efficient technology available in every layer of eHealth IoT system and will propose an efficient system.