Les systèmes intégrant des traitements venant du traitement automatique des langues reposent souvent sur des lexiques et des grammaires, parfois indirectement sur des corpus. A cause de la quantité et de la complexité des informations qu‘elles contiennent, ces ressources linguistiques deviennent facilement une source d‘incohérence. Dans cette thèse, nous explorons les moyens d‘améliorer la gestion des nombreuses ressources linguistiques d‘un moteur de recherche industriel en dix-neuf langues qui fait appel à une analyse textuelle élaborée. Nous proposons une méthode pour formaliser l‘architecture linguistique des traitements linguistiques et des ressources utilisées par ceux-ci. Cette formalisation explicite la façon dont les connaissances contenues dans les ressources sont exploitées. Grâce à elle, nous pouvons construire des outils de gestion qui respectent l‘architecture du système. L‘environnement ainsi mis en place se concentre sur la mise à jour et l‘acquisition des ressources linguistiques, leur exploitation étant figée par des contraintes industrielles.

<p>In modern cars a lot of effort is put on reducing the background noise level. Despite these efforts it is often difficult for persons in the rear seat(s) to hear the persons in the front seat. This is partly due to the background noise, but also geometry and acoustics properties of the passenger compartment. </p><p>The aim of this thesis was to implement a speech enhancement system to increase the audibility between the driver and the rear passenger(s). The speech enhancement system should not affect the directivity of the speech or increase the background noise level. </p><p>A speech enhancement system has been implemented on a DSP in a test car. A microphone was placed in front of the driver to collect his/her speech. The microphone signal was bandpass filtered to remove the main part of the background noise and to avoid aliasing. The signal was delayed before it was sent out in the rear loudspeaker. The delay made the speech from the driver reaching the rear passenger before the sound the rear loudspeakers. This delay was enough to get the right directivity of the sound, i.e. making speech sounding as if it came from the driver instead of the rear loudspeakers. </p><p>In the thesis other methods to reduce background noise and get directivity of the sound were evaluated, but not implemented in the test car. The evaluations of the system showed that the audibility was increased. At the same time the background noise level was not noticeable increased. The work has been performed at A2 Acoustics AB in Linköping, during spring 2003.</p>

<p>Bakgrund: Finns det verkligen en investeringsstrategi som ger en garanterad överavkastning och som skulle kunna innebära att efterfrågan på de tjänster som professionella kapitalförvaltare tillhandahåller försvinner?</p><p>Syfte: Syftet med vår uppsats är att undersöka huruvida det går att generera överavkastning på Stockholmsbörsen genom att investera i en portfölj enbart innehållande aktier med låga P/Etal. Frågan är alltså om det existerar det någon så kallad P/E-talseffekt?</p><p>Genomförande: P/E-talseffekten testades genom att vi för var och en av våra 28 perioder fr.o.m. 1991 t.o.m.2004 skapade två portföljer där den ena innehöll aktier med de femton lägsta P/E-talen på Stockholmsbörsen och den andra innehöll aktier med höga P/E-tal. Den låga P/E-talsportföljen riskjusterade avkastning jämfördes sedan med den höga P/Etalsportföljens. En jämförelse gjordes även med den riskjusterade avkastningen för AFGX och SIXRX.</p><p>Slutsats: Vi har efter att noggrant analyserat våra resultat kunnat fastställa med 95 procent sannolikhet att det sett till hela vår undersökningsperiod, fr.o.m. 1991 t.o.m. 2004, inte går att påvisa en P/E-talseffekt på Stockholmsbörsen. Ser vi däremot enbart till perioderna efter ITbubblan, hösten 2000 till hösten 2003, kan vi med hjälp av våra resultat konstatera att det under denna period var möjligt att erhålla en överavkastning genom att investera i aktier med</p><p>låga P/E-tal.</p> / <p>Background: Is there an investment strategy that yields a guaranteed abnormal return and that could imply that the demand for the services provided by professional capital managers would disappear?</p><p>Purpose: The purpose with this thesis is to examine whether it is possible to generate an abnormal return at the Stockholm Stock Exchange by investing in a portfolio that only contains stocks with low P/E ratios. The question is consequently if there exists a so called P/E effect.</p><p>Implementation: We tested the P/E effect by creating two portfolios for each of our 28 periods from the beginning of 1991 until the end of 2004. The first portfolio included stocks with the fifteen lowest P/E ratios on the Stockholm Stock Exchange and the other portfolio included stocks with high P/E ratios. The risk adjusted return of the low P/E ratio portfolio was then compared to that of the high P/E ratio portfolio. A comparison was also made with the risk adjusted return from AFGX and SIXRX.</p><p>Conclusion: We can, after a thorough analysis of our results, with 95 per cent probability say that a P/E effect didn’t existed on the Stockholm Stock Exchange if we look at our entire research period from the beginning of 1991 until the end of 2004. We can however say that it was possible to generate an abnormal return by investing in stocks with low P/E ratios, if we look solely at the periods after the IT bubble, fall 2000 until fall 2003.</p>

No description available.

<p>I samband med de ökade kraven på ett vädersäkrat nät har funderingar kring att placera ut fler frånskiljare/brytare i luftledningsnätet för att reducera medelavbrottstiden för anslutna kunder (SAIDI) kommit fram. I Finland använder man till stor del fjärrmanövrerade apparater för att reducera SAIDI. Rapportens inledande avsnitt innefattar en förstudie om ursprunget till avbrottstid och SAIDI-statistik i Vattenfall Eldistributions nät. En jämförelse med Eldistributions nät och klimatet i Finland utförs och av detta fås olika förutsättningar för nätet i de olika länderna. Som en generell slutsats av väderjämförelsen mellan Sverige och Finland fastställs att Eldistribution i Sverige har sämre förutsättningar vädermässigt när det gäller luftledningsnät. Resultaten från beräkningarna på sex exempelnät visar att den maximala reduceringen av SAIDI uppgår till 36 % och den minimala med 17 % i exempelnäten. Om alla nät skulle gå att bygga om enligt liknande principer skulle det innebära en reducering av Eldistributions mellanspännings-SAIDI med cirka 12 % ifall de 200 sämsta linjerna ur SAIDI-synpunkt kompletterades med fjärrfrånskiljare i motsvarande grad som exempelnäten. Jämförelser av investeringskostnader mellan fjärrfrånskiljaralternativ och kablifieringsalternativ visar att det är mer kostnadseffektivt att bygga om med fjärrfrånskiljare och därför är detta lämpligt om önskan är att så billigt som möjligt reducera SAIDI. Som en allmän rekommendation bör reservmatningar i nätet vara fjärrmanövrerade och dessutom bör linjer ha minst en fjärrfrånskiljare utlokaliserad. Då delas linjen upp i fler fjärrsektioneringsområden vilket förbättrar kundavbrottstiden för linjen. Ifall en linje inte har reservmatningmöjligheter bör det utredas ifall det är möjligt att med acceptabelt stora resurser bygga ihop nätet med ett annat nät för att möjliggöra en reservmatningsväg. Om det finns fler platser på linjen som är lämpliga för komplettering med fjärrfrånskiljare bör dessa utnyttjas för att så prisvärt som möjligt få ner SAIDI på den aktuella linjen. Ett allmänt råd angående antalet fjärrfrånskiljare per linje är svårt att ge men generellt sett är den första fjärrfrånskiljaren mycket mer kostnadseffektiv än de efterföljande. Stolpbrytare (reclosers) bör placeras ut där goda förutsättningar finns för att det skall bli lönsamt. Exempel på sådana platser kan vara vid övergång från kabel till luftledningsnät och innan en sträcka där fel ofta inträffar.</p> / <p>An alternative approach in Vattenfall Eldistribution’s network has evolved from the increased demands on reducing network downtime. Eldistribution are using remote controlled disconnectors / reclosers in Finland to reduce the system average interruption duration index (SAIDI). The initial section of the report includes a study on the source of downtime and SAIDI-statistics in Eldistribution’s network. A comparison with network and the climate in Finland is carried out which indicate different conditions in the two countries. The weather comparison shows that it’s more difficult to use non-insulated overhead-lines in Sweden because of the climate conditions. The result shows a maximum reduction of SAIDI by 36% and a minimum of 17% based on calculations on six example-lines. If all the lines would be possible to build under similar principles it would result in a reduction of Eldistributions SAIDI by approximately 12 % if the 200 worst lines was supplemented with remote controlled disconnectors with a corresponding amount of reduced SAIDI-minutes. Comparisons with cable alternatives show that it’s more cost-effective to invest in remoted controlled disconnectors if the target is a quick and cheap reduction of SAIDI. As a general recommendation the backup powering of a network should be remote controlled and the lines should have at least one outsourced remote controlled disconnector. This will split the line in multiple remote controlled areas, which will reduce disruption time for customers connected to the line. If a line doesn’t have any backup powering it should be investigated whether it’s possible to connect the line to another line to ensure backup-feeding if this is possible to do at acceptable costs. If there are more locations on the line suitable for supplementation of remote controlled disconnectors, those places should be used to get as much reduced SAIDI as possible at a cost-effective level. It's difficult to give a general advice concerning the number of remote controlled disconnectors but the first disconnector is generally more cost-effective than the subsequent ones. Reclosers should be deployed at appropriate locations to be equally costefficient as remoted controlled disconnectors. Examples of such places can be when switching from cable to non-insulated line and before a section where errors often occur.</p>

<p>During the 17th century the production and distribution of printed works increased in Stockholm and Uppsala. However, all works that the publisher wanted to print, did not reach the reader. This thesis aimed to show when, where and how printed books and dissertations did not reach the reader. To achieve this aim, aspects of several well known models showing systems of book production and distribution were combined and adapted to suit Swedish conditions in the 17th century. This combined model focused on processes of selection and was presented and used as a theoretical approach. Processes of selection in this paper are processes in which often a conscious choice was made of which books people produced and distributed. By using literature that had already been written about the chosen subject the paper examines what this material tells us about processes of selection during the selected period. Six processes of selection have been identified and discussed, three belonging to the production of works and three belonging to the distribution of works. During the production publishers’ motives and conditions for publishing and how this affected what they chose and did not choose to publish were discussed. Publishers with economical motives might publish fewer books which contained new ideas as reprinting popular books probably resulted in a higher profit. If a publisher wanted to use the publishing of a book as a way to further his career that probably could lead to fewer books containing offensive material being published. The state of the printing material and the economy of the printer and publisher were things that could affect which books that were printed. Books that came to Stockholm and Uppsala from other countries could be lost while being transported. The increase and decrease of the book collection at the university library in Uppsala affected the reader’s range of books to choose from. This is a two years master’s thesis in library and information science.</p>

During the 17th century the production and distribution of printed works increased in Stockholm and Uppsala. However, all works that the publisher wanted to print, did not reach the reader. This thesis aimed to show when, where and how printed books and dissertations did not reach the reader. To achieve this aim, aspects of several well known models showing systems of book production and distribution were combined and adapted to suit Swedish conditions in the 17th century. This combined model focused on processes of selection and was presented and used as a theoretical approach. Processes of selection in this paper are processes in which often a conscious choice was made of which books people produced and distributed. By using literature that had already been written about the chosen subject the paper examines what this material tells us about processes of selection during the selected period. Six processes of selection have been identified and discussed, three belonging to the production of works and three belonging to the distribution of works. During the production publishers’ motives and conditions for publishing and how this affected what they chose and did not choose to publish were discussed. Publishers with economical motives might publish fewer books which contained new ideas as reprinting popular books probably resulted in a higher profit. If a publisher wanted to use the publishing of a book as a way to further his career that probably could lead to fewer books containing offensive material being published. The state of the printing material and the economy of the printer and publisher were things that could affect which books that were printed. Books that came to Stockholm and Uppsala from other countries could be lost while being transported. The increase and decrease of the book collection at the university library in Uppsala affected the reader’s range of books to choose from. This is a two years master’s thesis in library and information science.

No description available.

This essay examines masculinity in the Swedish military paper Värnpliktsnytt during 1971-1991. The purpose is to study the masculine norms and the hegemonic masculinity that Värnpliktsnytt conveys during these years. Värnpliktsnytt was handed out for free to all soldiers during 1971-2010. During the 1970’s the women were only allowed on non-combat posts and in the 1980’s they more and more were included and accepted as soldiers and staff in the military. In 1989 women were allowed to apply for active and non-combat duty in all military sectors. This affected the masculine ideal of Värnpliktsnytt, and the desirable view of man started to change. Before the 1980’s strong men in the ranger divisions constituted the hegemonic masculinity’s peak, but during this decade the regular conscripted soldiers with more individual thought and softer feelings rose in the hegemonic hierarchy. Also with the common debate on gender equality men in the military was encouraged to take a bigger part in the home life and to act more responsible towards the upbringing of the children. Apart from the duty to family, the military duty is also central to the masculinity in Värnpliktsnytt during the whole examined period. Men are expected to fulfill their military duty, although it is accepted to complain about wage andworking conditions. The overall results points in the direction that the ideal man becomes more complex.

I samband med de ökade kraven på ett vädersäkrat nät har funderingar kring att placera ut fler frånskiljare/brytare i luftledningsnätet för att reducera medelavbrottstiden för anslutna kunder (SAIDI) kommit fram. I Finland använder man till stor del fjärrmanövrerade apparater för att reducera SAIDI. Rapportens inledande avsnitt innefattar en förstudie om ursprunget till avbrottstid och SAIDI-statistik i Vattenfall Eldistributions nät. En jämförelse med Eldistributions nät och klimatet i Finland utförs och av detta fås olika förutsättningar för nätet i de olika länderna. Som en generell slutsats av väderjämförelsen mellan Sverige och Finland fastställs att Eldistribution i Sverige har sämre förutsättningar vädermässigt när det gäller luftledningsnät. Resultaten från beräkningarna på sex exempelnät visar att den maximala reduceringen av SAIDI uppgår till 36 % och den minimala med 17 % i exempelnäten. Om alla nät skulle gå att bygga om enligt liknande principer skulle det innebära en reducering av Eldistributions mellanspännings-SAIDI med cirka 12 % ifall de 200 sämsta linjerna ur SAIDI-synpunkt kompletterades med fjärrfrånskiljare i motsvarande grad som exempelnäten. Jämförelser av investeringskostnader mellan fjärrfrånskiljaralternativ och kablifieringsalternativ visar att det är mer kostnadseffektivt att bygga om med fjärrfrånskiljare och därför är detta lämpligt om önskan är att så billigt som möjligt reducera SAIDI. Som en allmän rekommendation bör reservmatningar i nätet vara fjärrmanövrerade och dessutom bör linjer ha minst en fjärrfrånskiljare utlokaliserad. Då delas linjen upp i fler fjärrsektioneringsområden vilket förbättrar kundavbrottstiden för linjen. Ifall en linje inte har reservmatningmöjligheter bör det utredas ifall det är möjligt att med acceptabelt stora resurser bygga ihop nätet med ett annat nät för att möjliggöra en reservmatningsväg. Om det finns fler platser på linjen som är lämpliga för komplettering med fjärrfrånskiljare bör dessa utnyttjas för att så prisvärt som möjligt få ner SAIDI på den aktuella linjen. Ett allmänt råd angående antalet fjärrfrånskiljare per linje är svårt att ge men generellt sett är den första fjärrfrånskiljaren mycket mer kostnadseffektiv än de efterföljande. Stolpbrytare (reclosers) bör placeras ut där goda förutsättningar finns för att det skall bli lönsamt. Exempel på sådana platser kan vara vid övergång från kabel till luftledningsnät och innan en sträcka där fel ofta inträffar. / An alternative approach in Vattenfall Eldistribution’s network has evolved from the increased demands on reducing network downtime. Eldistribution are using remote controlled disconnectors / reclosers in Finland to reduce the system average interruption duration index (SAIDI). The initial section of the report includes a study on the source of downtime and SAIDI-statistics in Eldistribution’s network. A comparison with network and the climate in Finland is carried out which indicate different conditions in the two countries. The weather comparison shows that it’s more difficult to use non-insulated overhead-lines in Sweden because of the climate conditions. The result shows a maximum reduction of SAIDI by 36% and a minimum of 17% based on calculations on six example-lines. If all the lines would be possible to build under similar principles it would result in a reduction of Eldistributions SAIDI by approximately 12 % if the 200 worst lines was supplemented with remote controlled disconnectors with a corresponding amount of reduced SAIDI-minutes. Comparisons with cable alternatives show that it’s more cost-effective to invest in remoted controlled disconnectors if the target is a quick and cheap reduction of SAIDI. As a general recommendation the backup powering of a network should be remote controlled and the lines should have at least one outsourced remote controlled disconnector. This will split the line in multiple remote controlled areas, which will reduce disruption time for customers connected to the line. If a line doesn’t have any backup powering it should be investigated whether it’s possible to connect the line to another line to ensure backup-feeding if this is possible to do at acceptable costs. If there are more locations on the line suitable for supplementation of remote controlled disconnectors, those places should be used to get as much reduced SAIDI as possible at a cost-effective level. It's difficult to give a general advice concerning the number of remote controlled disconnectors but the first disconnector is generally more cost-effective than the subsequent ones. Reclosers should be deployed at appropriate locations to be equally costefficient as remoted controlled disconnectors. Examples of such places can be when switching from cable to non-insulated line and before a section where errors often occur.