The purpose of this thesis is to compare five different ranking algorithms for query expansion. The algorithms compared are f4, f4mod, porter, wpq, and emim. This is done using a TREC collection, a selection of topics, and relevance judgements. Relative recall is measured before and after the expansion of the query. The study shows that all of the algorithms manage to increase the relative recall, f4 being the one most successful. / Uppsatsnivå: D

The aim of this thesis is to describe passage retrieval (PR), with basis in results from various empirical experiments, and to critically investigate different approaches in PR. The main questions to be answered in the thesis are: (1) What characterizes PR? (2) What approaches have been proposed? (3) How well do the approaches work in experimental information retrieval (IR)? PR is a research topic in information retrieval, which instead of retrieving the fulltext of documents, that can lead to information overload for the user, tries to retrieve the most relevant passages in the documents. This technique was investigated studying a number of central articles in the research field. PR can be divided into three different types of approaches based on the segmentation of the documents. First, you can divide the text considering the semantics and where the topics change. Second, you can divide the text based on the explicit structure of the documents, with help from e.g. a markup language like SGML. And third, you can do a form of PR, where you divide the text in parts containing a fixed number of words. This method is called unmotivated segmentation. The study showed that an unmotivated segmentation resulted in the best retrieval effectiveness even though the results are difficult to compare because of different kinds of evaluation methods and different types of test collections. A combination between full text retrieval and PR also showed improved results. / Uppsatsnivå: D

The purpose of this thesis is to evaluate 7 different search strategies for content-based image retrieval with respect to retrieval effectiveness. The strategies are based on different levels of resolution implemented by the Haar transformation. The result of the study shows an improvement in both mean average precision and mean recall in strategies based on lower levels of resolution. / Uppsatsnivå: D

The aim of this master thesis is to evaluate the performance of the content-based image retrieval system ImBrowse from a semantic point of view. Evaluation of retrieval performance is a problem in content-based image retrieval (CBIR). There are many different methods for measuring the performance of content-based image retrieval systems, but no common way for performing the evaluation. The main focus is on image retrieval regarding the extraction of the visual features in the image, from three semantic levels. The thesis tries to elucidate the semantic gap, which is the problem when the systems extraction of the visual features from the image and the user’s interpretation of that same information do not correspond. The method of this thesis is based on similar methods in evaluation studies of CBIR systems. The thesis is an evaluation of ImBrowse’s feature descriptors for 30 topics at three semantic levels and compared the descriptors performance based on our relevance assessment. For the computation of the results the precision at DCV = 20 is used. The results are presented in tables and a chart. The conclusion from this evaluation is that the retrieval effectiveness from a general point of view did not meet the semantic level of our relevance assessed topics. However, since the thesis do not have another system with the same search functions to compare with it is difficult to draw a comprehensive conclusion of the results. / Uppsatsnivå: D

In query expansion (QE) terms are added to an initial query in order to improve retrieval effectiveness. In this thesis we use QE in the sense that a reformulation of the query is done by deleting the terms in the initial query and instead replacing them with terms from the documents retrieved in the initial run. The aim of this thesis is to, in a experimental full text invironment, study and compare the retrieval result of two different query expansion strategies in relation to each other. The following questions are addressed by the study: How do the two strategies perform in relation to each other regarding recall?What may be causing the result?Are the two strategies retrieving the same relevant documents?Two strategies are designed to simulate a searcher using automatic query expansion (AQE) either with or without relevance feedback. Strategy I is simulating AQE without relevance feedback by taking the top five documents that are retrieved in the initial run and then extracting the top ten most frequently occurring terms in these to create a new query. Correspondingly the Strategy II, is simulating AQE with relevance feedback by taking the top five relevant documents and extracting the top ten terms in these to create a new query. It is concluded that both of the strategies’ retrieval performance was improved for most of the topics. In average Strategy II did achieve 54.63 percent recall compared to Strategy I which did achieve 45.59 percent recall. The two strategies did retrieve different relevant documents for majority of the topics. Hence, it would be reasonable to base a system on both of them. / Uppsatsnivå: D

This thesis is focusing on how global and local search engines retrieve information from the local domain. The three search engines tested are the global search engines Yahoo, Ask.com and the local search engine Web Wombat. The questions we examined were: which search engine has the best retrieval effectiveness? Could there be other reasons than retrieval effectiveness, to choosing a local search engine?For our test we constructed 20 questions related to Australia.We chose to divide the questions into topics like nature, sports and culture. For all questions we evaluated the relevance of the first 20 hits. We used the following measures in our test: Jaccard´s index, precision and average precision. We also looked at factors such as duplicates and error pages, because we consider this to be an important aspect to consider, when looking at the relevance of the first 20 hits. The results of our study showed that Yahoo had the best performance for precision. Web Wombat had faulty precision but results from Jaccard´s index revealed that Web Wombat had many unique documents. Web Wombat had the best average precision on one of our questions. In spite of Web Wombats faulty precision, we think that Web Wombat serve a purpose as an alternative to global search engines. / Uppsatsnivå: D

The purpose of this study is to examine the difference between three query strategies with respect to retrieval effectiveness. The thesis aims at examining how two types of noun phrases containing a modifier to the head word, which is a noun affect the retrieval performance with regard to recall and precision. The noun phrases in the thesis are of two types: 1) noun phrases containing a modifier to the head word (which is a noun) and which are not dictionary phrases (NF) and 2) dictionary phrases. Both types of noun phrases in this thesis contain at least two words. The queries were executed in Query Performance Analyser, QPA, containing the InQuery system and a sub collection of TREC-Uta documents with its topics. The measures used were R@200 and precision at 13 document cut-off values. In total, 20 topics were examined. All the documents and the queries were in English. The result of this study indicates that NFs have a negative effect on the retrieval performance and that dictionary phrases have a rather neutral effect on the retrieval performance. / Uppsatsnivå: D

This Master’s thesis examines different retrieval strategies used in cross-language information retrieval (CLIR). The aim was to investigate if there were any differences between baseline queries and translated queries in retrieval effectiveness; how the retrieval effectiveness was affected by query structuring and if the results differed between different languages. The languages used in this study were Swedish, English and Finnish. 30 topics from the TrecUta collection were translated into Swedish and Finnish. Baseline queries in Swedish and Finnish were made and translated into English using a dictionary and thereby simulating automatic translation. The queries were expanded by adding all the translations from the main entries to the queries. Two kinds of queries – structured and unstructured – were designed. The queries were fed into the InQuery IR system which presented a list of retrieved documents where the relevant ones were marked. The performance of the queries was analysed by Query Performance Analyser (QPA). Average precision at seen relevant documents at DCV 10, average precision at DCV 10 and precision and recall at DCV 200 were used to measure the retrieval effectiveness. Despite the morphological differences between Swedish and Finnish, none or very small differences in retrieval performance were found, except when average precision at DCV 10 was used. The baseline queries performed the best results and the structured queries performed better in both Swedish and Finnish than the unstructured queries. The results are consistent with previous research. / Uppsatsnivå: D

The purpose of this thesis is to evaluate and compare the retrieval effectiveness of two search engines on the Internet: Svensk Medicin and Google Scholar. Svensk Medicin is a medicine-specific search engine and Google Scholar specialises in indexing scholarly material. A set of 20 questions in four categories were used to conduct the searches with the search engines. Genuine medical information needs were selected from a question collection provided by the television program Fråga doktorn. The relevance of the first 10 retrieved documents was judged by using a binary scale and the measure used was precision. An analysis on average precision for all the search questions and an average measure was calculated (macro precision). Results for average precision over all the search questions and a mean average precision is also presented (micro precision and map). These measures shows which search engine performs best for single queries and how the relevant documents spread among the displayed results. The occurrence of duplicates and error pages was noted, because they also show how the search engine performs. Svensk Medicin is performing better than Google Scholar on the single queries and Svensk Medicin also retrieves the highest number of relevant documents. Methodology is discussed close since the experimental design affects the result. / Uppsatsnivå: D

The aim of this Master’s thesis is to examine the retrieval performance of two search engines, Yahoo! News and Sesam, specialized on indexing news articles in Swedish. Thirty search questions were used in this study, divided into three different categories: domestic news, amusement and sport. Precision has been calculated both for the 30 search questions as a total and for the three categories as well. Precision was calculated at every DCV-level and then divided with the number of search questions used, in an attempt to reach an average measure. An analysis was also used were precision was calculated over all the DCV-levels for each question in attempt to see how the search engines perform for a given search question and a given search engine. The relevance of the top 10 retrieved documents was judged using a binary scale. The result of the study shows little difference in precision between the two search engines. Yahoo! News performs slightly better than Sesam in most cases. The results show much variation between the search engines regarding how they perform per given search question. Much of the underlying techniques used by the search engines is not know for the public why the methods used in the study, for example the terms chosen as search terms, choice of highest DCV-level et cetera, and how these different choices made may had an impact on the result in the study is discussed rather than underlying technical functions. / Uppsatsnivå: D