In recent years the amount of information on the web has increased dramatically. As a result, it has become a challenge for the researchers to find effective ways that can help us query and extract meaning from these large repositories. Standard document search engines try to address the problem by presenting the users a ranked list of relevant documents. In most cases, this is not enough as the end-user has to go through the entire document to find out the answer he is looking for. Question answering, which is the retrieving of answers to natural language questions from a document collection, tries to remove the onus on the end-user by providing direct access to relevant information. This thesis is concerned with open-domain complex question answering. Unlike simple questions, complex questions cannot be answered easily as they often require inferencing and synthesizing information from multiple documents. Hence, we considered the task of complex question answering as query-focused multi-document summarization. In this thesis, to improve complex question answering we experimented with both empirical and machine learning approaches. We extracted several features of different types (i.e. lexical, lexical semantic, syntactic and semantic) for each of the sentences in the document collection in order to measure its relevancy to the user query. We have formulated the task of complex question answering using reinforcement framework, which to our best knowledge has not been applied for this task before and has the potential to improve itself by fine-tuning the feature weights from user feedback. We have also used unsupervised machine learning techniques (random walk, manifold ranking) and augmented semantic and syntactic information to improve them. Finally we experimented with question decomposition where instead of trying to find the answer of the complex question directly, we decomposed the complex question into a set of simple questions and synthesized the answers to get our final result. / x, 128 leaves : ill. ; 29 cm

Thesis (MPhil)--Stellenbosch University, 2004 / ENGLISH ABSTRACT: With the introduction of the Internet and the World Wide Web (www), information can be easily accessed and retrieved from the web using information retrieval systems such as web search engines or simply search engines. There are a number of search engines that have been developed to provide access to the resources available on the web and to help users in retrieving relevant information from the web. In particular, they are essential for finding text information on the web for academic purposes. But, how effective and efficient are those search engines in retrieving the most relevant text information from the web? Which of the search engines are more effective and efficient? So, this study was conducted to see how effective and efficient search engines are and to see which search engines are most effective and efficient in retrieving the required text information from the web. It is very important to know the most effective and efficient search engines because such search engines can be used to retrieve a higher number of the most relevant text web pages with minimum time and effort. The study was based on nine major search engines, four search queries and relevancy judgments as relevant/partly-relevanUnon-relevant. Precision and recall were calculated based on the experimental or test results and these were used as basis for the statistical evaluation and comparisons of the retrieval effectiveness of the nine search engines. Duplicated items and broken links were also recorded and examined separately and were used as an additional measure of search engine effectiveness. A response time was also recorded and used as a base for the statistical evaluation and comparisons of the retrieval efficiency of the nine search engines. Additionally, since search engines involve indexing and searching in the information retrieval processes from the web, this study first discusses, from the theoretical point of view, how the indexing and searching processes are performed in an information retrieval environment. It also discusses the influences of indexing and searching processes on the effectiveness and efficiency of information retrieval systems in general and search engines in particular in retrieving the most relevant text information from the web. / AFRIKAANSE OPSOMMING: Met die koms van die Internet en die Wêreldwye Web (www) is inligting maklik bekombaar. Dit kan herwin word deur gebruik te maak van inligtingherwinningsisteme soos soekenjins. Daar is 'n hele aantal sulke soekenjins wat ontwikkel is om toegang te verleen tot die hulpbronne beskikbaar op die web en om gebruikers te help om relevante inligting vanaf die web in te win. Dit is veral noodsaaklik vir die verkryging van teksinligting vir akademiese doeleindes. Maar hoe effektief en doelmatig is die soekenjins in die herwinning van die mees relevante teksinligting vanaf die web? Watter van die soekenjins is die effektiefste? Hierdie studie is onderneem om te kyk watter soekenjins die effektiefste en doelmatigste is in die herwinning van die nodige teksinligting. Dit is belangrik om te weet watter soekenjin die effektiefste is want so 'n enjin kan gebruik word om 'n hoër getal van die mees relevante tekswebblaaie met die minimum van tyd en moeite te herwin. Heirdie studie is baseer op die sewe hoofsoekenjins, vier soektogte, en toepasliksheidsoordele soos relevant /gedeeltelik relevant/ en nie- relevant. Presiesheid en herwinningsvermoë is bereken baseer op die eksperimente en toetsresultate en dit is gebruik as basis vir statistiese evaluasie en vergelyking van die herwinningseffektiwiteit van die nege soekenjins. Gedupliseerde items en gebreekte skakels is ook aangeteken en apart ondersoek en is gebruik as bykomende maatstaf van effektiwiteit. Die reaksietyd is ook aangeteken en is gebruik as basis vir statistiese evaluasie en die vergelyking van die herwinningseffektiwiteit van die nege soekenjins. Aangesien soekenjins betrokke is by indeksering en soekprosesse, bespreek hierdie studie eers uit 'n teoretiese oogpunt, hoe indeksering en soekprosesse uitgevoer word in 'n inligtingherwinningsomgewing. Die invloed van indeksering en soekprosesse op die doeltreffendheid van herwinningsisteme in die algemeen en veral van soekenjins in die herwinning van die mees relevante teksinligting vanaf die web, word ook bespreek.

The major goal of information integration is to provide efficient and easy-to-use access to multiple heterogeneous data sources with a single query. At the same time, one of the current trends is to use standard technologies for implementing solutions to complex software problems. In this dissertation, I used XML and XQuery as the standard technologies and have developed an extended projection algorithm to provide a solution to the information integration problem. In order to demonstrate my solution, I implemented a prototype mediation system called Omphalos based on XML related technologies. The dissertation describes the architecture of the system, its metadata, and the process it uses to answer queries. The system uses XQuery expressions (termed metaqueries) to capture complex mappings between global schemas and data source schemas. The system then applies these metaqueries in order to rewrite a user query on a virtual global database (representing the integrated view of the heterogeneous data sources) to a query (termed an outsourced query) on the real data sources. An extended XML document projection algorithm was developed to increase the efficiency of selecting the relevant subset of data from an individual data source to answer the user query. The system applies the projection algorithm to decompose an outsourced query into atomic queries which are each executed on a single data source. I also developed an algorithm to generate integrating queries, which the system uses to compose the answers from the atomic queries into a single answer to the original user query. I present a proof of both the extended XML document projection algorithm and the query integration algorithm. An analysis of the efficiency of the new extended algorithm is also presented. Finally I describe a collaborative schema-matching tool that was implemented to facilitate maintaining metadata.

This project, Database Access through the Web (DAW), implements a database to store academic and general information of graduate students in the Department of Computer Science, CSUSB and provides access to the database from the web. The motivation of the project comes from needs of the Graduate Coordinator, professors and department staff to access through the Internet student information concurrently.

Indiana University-Purdue University Indianapolis (IUPUI) / The literature corresponding to disordered proteins has been on a rise. As the number of publications increase, the time and effort needed to manually identify the relevant publications and protein information to add to centralized repository (called DisProt) is becoming arduous and critical. Existing search facilities on PubMed can retrieve a seemingly large number of publications based on keywords and does not have any support for ranking them based on the probability of the protein names mentioned in a given abstract being added to DisProt. This thesis explores a novel system of using disorder predictors and context based dictionary methods to quickly identify publications on disordered proteins from the PubMed database. NLProt, which is built around Support Vector Machines, is used to identify protein names and PONDR-FIT which is an Artificial Neural Network based meta- predictor is used for identifying protein disorder. The work done in this thesis is of immediate significance in identifying disordered protein names. We have tested the new system on 100 abstracts from DisProt [these abstracts were found to be relevant to disordered proteins and were added to DisProt manually by the annotators.] This system had an accuracy of 87% on this test set. We then took another 100 recently added abstracts from PubMed and ran our algorithm on them. This time it had an accuracy of 68%. We suggested improvements to increase the accuracy and believe that this system can be applied for identifying disordered proteins from literature.

Reported in this dissertation are the results of investigations performed at the New Orleans Center for Mass Spectrometry at the University of New Orleans. The projects that are detailed in the coming pages take on a variety of subjects, but a common thread is that each employs matrix-assisted laser desorption/ionization (MALDI) mass spectrometry to solve a problem. Fundamental aspects of MALDI in-plume ionization are implicated in the introduction of a newly developed electron-transfer secondary ionization matrix. The remaining projects are related to the ever expanding field of proteomics. Mass spectrometry was used to investigate the site specificity of a newly developed bilin-lyase enzyme, a new approach was developed to distinguish between A-ring and D-ring attachment of bilins, and F. grandis protein expression pattern was investigated in several tissues. All obtained results were acquired using a MALDI TOF/TOF mass spectrometer. The sensitivity, mass accuracy, mass resolution and the ability to perform collision induced decomposition (CID) experiments were all valuable features that served to raise the quality of data, and thereby improved the detail of inferences to be drawn for the different projects.

The purpose of this study was to investigate the use of electronic databases by postgraduate students in the Faculty of Science and Agriculture at the University of KwaZulu-Natal, Pietermaritzburg, with a view to proposing a set of recommendations based on the findings that would contribute to a more effective use of these databases. The study focussed on the use students made of electronic databases provided by the University of KwaZulu-Natal, Pietermaritzburg Library and any other electronic databases which provided information and were needed for their studies. The study population consisted of 500 registered postgraduate students. The survey instrument used to elicit data was a self-administered questionnaire that was distributed to a sample population of 100 postgraduate students. The questionnaire sought to establish if postgraduate students used electronic databases, which electronic databases postgraduate students used and the reasons for their choice, to identify problems postgraduate students encountered, how students became aware of the databases and how often they used the available databases. A total of 65 postgraduate students responded, a response rate of 65%. The results were analysed in terms of frequency responses and they are graphically displayed in the form of tables and graphs. The study found that postgraduate students did use the electronic databases, but a few of the databases were not used. A number of problems were experienced when using the databases. Students became aware of the availability of electronic databases from a variety of sources such as friends, library orientation programmes and academic staff. Search engines were identified as a resource that was very popular with almost all the students. There was a need for training on use of the databases and a need for improving access for all campus and off-campus users. Recommendations for action and further research, based on the conclusions of the study, are made. / Thesis (M.I.S.)-University of KwaZulu-Natal, Pietermaritzburg, 2005.

The proliferation ofinformation available in electronic format has been perpetuated by rapid technological advances. Users have a choice between print and electronic resources. At the same time, the use of these resources is determined by what is available to the users and what the users prefer. Problems that users encounter when accessing these resources determine the use of print and electronic resources. The current study evaluated the use of, and preference for, print and electronic resources by students at the University of Malawi College of Medicine. The study population comprised 179 undergraduate students. A self-administered questionnaire was used to establish the students' use of, and preference for, print and electronic resources, reasons for their preferences and the problems they encountered in accessing print and electronic resources. An interview schedule was used to elicit background information from the College Librarian on issues of budgeting, technological infrastructure, licensing and copyright agreements, archiving and library staff and training. Questionnaire results were quantitatively analysed and presented in terms of frequency tables and graphs. Interview results were analysed qualitatively. The study findings show that the students used both print and electronic resources. However, print resources were more heavily used than electronic ones. The students also preferred print resources to electronic. A lack of sufficient computers and low levels of computer and information literacy contributed to the underutilisation of electronic resources. Recommendations in terms of technological infrastructure, user training, staff training, user support and archiving were made based on the findings of the study. / Thesis (M.I.S.)-University of KwaZulu- Natal, Pietermaritzburg, 2005.

In South Africa, government (Mbeki 1996:37) and educators (SAIDE Report 1998:9) have expressed concern over the provision of microcomputers for learners. Their provision to schools would allow their effective use across the curriculum and enhance education. This study investigated the role of microcomputers as information retrieval tools in the 1997 schools' Water Audit projects in the greater Pietermaritzburg area. The study considered, firstly, a discussion of environmental education with emphasis on water conservation, and secondly the Water Audit projects. An overview of issues relating to microcomputer systems as information retrieval tools in education was presented. A descriptive survey method was employed for the study, with questionnaires as the data collection technique. Thirty out of 40 teachers/school project co-ordinators were sampled, with a 24 (80%) response. Twenty percent of 550 pupils who participated in the projects also responded. Data was presented by the use of tables. The study revealed an unequal availability of microcomputers among the schools in the departments of education, as they existed prior to 1994, that participated in the projects. Findings also revealed that only a minority of pupils used the microcomputer system for information retrieval during the projects. The use of hard copy sources was significantly high within the departments. A high number of pupils who used the microcomputer found it useful. A higher number that used hard copy sources found them useful. Training of pupils to acquire computer and information skills was inadequate. Major problems encountered during the projects included congestion, lack of computer skills on the part of both teachers and pupils and inadequate training of participants. The study found infrastructural backlogs and logistical problems as hindrances to the delivery of education in South Africa using microcomputers. Non-involvement of media teachers negatively impacted on the results of the projects. It was recommended that schools without microcomputers liaise with donor agencies to acquire microcomputers, while government initiates policies to address the issue of equity. Educators and media teachers should be given concerted training in computer and information skills, as training underpins the use of microcomputers in an information age school. / Thesis (MIS)-University of Natal, Pietermaritzburg, 1999.

The purpose of this study was to examine the use of library resources by doctoral students of the University of KwaZulu-Natal (UKZN), College of Humanities, Pietermaritzburg campus. The study investigated the extent of the use of the library resources by doctoral students in the College of Humanities, the type of library resources used and the factors that influenced such use. The study also looked at the competency of doctoral students to use them. The study population consisted of registered doctoral students in the College of Humanities as at April 2012 when the data was collected. Out of a total of 118 questionnaires distributed, 99 doctoral students responded, which yielded a response rate of 83.9%. The method that the research took was triangulation where both qualitative and quantitative data were collected. The researcher used one-on-one interviews with subject librarians at the UKZNP Library and a questionnaire as data collection instruments. The quantitative data was analyzed using SPSS version 21, while the qualitative data was analyzed using thematic content analysis. The questionnaire was pre-tested before it was used. The findings of the study revealed that a majority of the doctoral students did use the UKZNP Library resources for their research studies. As regards the importance of the library resources, out of 98 respondents who responded to the question, 77 (78.6%) of the respondents indicated that their use of library resources was because of its importance to their research study. Regarding the dependency on library resources, out of 98 respondents who responded to the question, 60 (61.2%) respondents said their extent of use of the library‟s resources was because they were very dependent on the resources of the library for their research study. In terms of usage of the library‟s resources, out of the 99 respondents who responded to the question, 50 (53.8%) said they use the library‟s resources more for their research study. When investigating the types of library resources used by doctoral students in the College of Humanities, 85 (86.7%) respondents used electronic databases, while 77 (78.6%) used print materials (books and journals); 68 (69.4%) used the inter-library loan service and 60 (61.2%) used theses and dissertations amongst other resources. With regards to the influencing factors for such use of library resources by doctoral students in the College of Humanities, the study found that 93 (97.9%) influencing factors for the respondents was to borrow materials, like printed books and journal articles, for their research studies, while for 52 (54.7%) respondents, a factor was to use inter-library loan services for their research studies and for 38 (40.0%) it was to search the databases for information for their research studies. Furthermore, the study also wanted to determine how competent the doctoral students in the College of Humanities were in using the different library resources of the UKZNP Library. Out of 99 respondents, 79 (79.8%) considered themselves to be competent users of the different library resources with particular reference to electronic resources, while 16.2 (16.2%) considered themselves not to be competent and four (4.0%) did not respond to the question. Having seen the extent of use of the library resources by doctoral students in the College of Humanities, the type of resources the doctoral students used, the factors that influenced such use and the competency of use, the study also discovered that there were challenges that these doctoral students encountered while using the library‟s resources. Some of these challenges were: difficulties in locating resources on shelves as a result of the misplaced and disorganized arrangement of material; out of date printed resource material and insufficient time for borrowing books on reserve. With electronic databases, some of the challenges were that important journals and books were not available electronically, remote access was difficult because of passwords and logins, and some electronic databases only offer abstracts rather than full text articles. In addition, electronic databases were often very slow to conclude searches. A final challenge encountered by the doctoral students was that subject librarians were always very busy attending to students and sometimes not available to help. Recommendations based on the findings of the study were also presented in chapter five and six of this study. Suggestions on how the library can better serve the doctoral students of the UKZNP campus were also given in the concluding chapter. / Thesis (M.I.S)-University of KwaZulu-Natal, Pietermaritzburg, 2013.