Search

This paper addresses the problem of the unsupervised classification of text-centric XML documents. In the context of the INEX mining track 2006, we present methods to exploit the inherent structural information of XML documents in the document clustering process. Using the k-means algorithm, we have experimented with a couple of feature sets, to discover that a promising direction is to use structural information as a preliminary means to detect and put aside structural outliers. The improvement of the semantic-wise quality of clustering is significantly higher through this approach than through a combination of the structural and textual feature sets. The paper also discusses the problem of the evaluation of XML clustering. Currently, in the INEX mining track, XML clustering techniques are evaluated against semantic categories. We believe there is a mismatch between the task (to exploit the document structure) and the evaluation, which disregards structural aspects. An illustration of this fact is that, over all the clustering track submissions, our text-based runs obtained the 1st rank (Wikipedia collection, out of 7) and 2nd rank (IEEE collection, out of 13). [ABSTRACT FROM AUTHOR]

We address the problem of learning to map automatically flat and semi-structured documents onto a mediated target XML schema. We propose a machine learning approach where the mapping between input and target documents is learned from examples. Complex transformations can be learned using only pairs of input and corresponding target documents. From a machine learning point of view, the structure mapping task raises important complexity challenges. Hence we propose an original model which scales well to real world applications. We provide learning and inference procedures with low complexity. The model sequentially builds the target XML document by processing the input document node per node. We demonstrate the efficiency of our model on two structure mapping tasks. Up to our knowledge, there are no other model yet able to solve these tasks. [ABSTRACT FROM AUTHOR]

XML is becoming increasingly popular as a language for representing many types of electronic documents. The consequence of the strict structural document description via XML is that a relatively new task in mining documents based on structural and/or content information has emerged. In this paper we investigate (1) the suitability of new unsupervised machine learning methods for the clustering task of XML documents, and (2) the importance of contextual information for the same task. These tasks are part of an international competition on XML clustering and categorization (INEX 2006). It will be shown that the proposed approaches provide a suitable tool for the clustering of structured data as they yield the best results in the international INEX 2006 competition on clustering of XML data. [ABSTRACT FROM AUTHOR]

We address the problem of structure mapping that arises in xml data exchange or xml document transformation. Our approach relies on xml annotation with semantic labels that describe local tree editions. We propose xml Conditional Random Fields (xcrfs), a framework for building conditional models for labeling xml documents. We equip xcrfs with efficient algorithms for inference and parameter estimation. We provide theoretical arguments and practical experiments that illustrate their expressivity and efficiency. Experiments on the Structure Mapping movie datasets of the inex xml Document Mining Challenge yield very good results. [ABSTRACT FROM AUTHOR]

The natural representation of XML data is to use the underlying tree structure of the data. When analyzing these trees we are ensured that no structural information is lost. These tree structures can be efficiently analyzed due to the existence of frequent pattern mining algorithms that works directly on tree structured data. In this work we describe a classification method for XML data based on frequent attribute trees. From these frequent patterns we select so called emerging patterns, and use these as binary features in a decision tree algorithm. The experimental results show that combining emerging attribute tree patterns with standard classification methods, is a promising combination to tackle the classification of XML documents. [ABSTRACT FROM AUTHOR]

This paper reports the results and experiments performed on the INEX 2006 Document Mining Challenge Corpus with the PCXSS clustering method. The PCXSS method is a progressive clustering method that computes the similarity between a new XML document and existing clusters by considering the structures within documents. We conducted the clustering task on the INEX and Wikipedia data sets. [ABSTRACT FROM AUTHOR]

In this paper we present a method of document representation called Rich Document Representation (RDR) to build XML retrieval engines with high specificity. RDR is a form of document representation that utilizes single words, phrases, logical terms and logical statements for representing documents. The Vector Space model is used to compute index terms weight and similarity between each element and query. This system has participated in INEX 2006 and tested with the Content Only queries of the given collection. The results have been very weak but a failure analysis has revealed that it has been caused by an error in document processing which has produced inconsistent IDs and caused a mismatch between the IDs assigned to document elements such as single terms, phrases and logical terms. However similar experiment on INEX2004 collection yielded very good precision on high specificity task with s3e123 quantization. [ABSTRACT FROM AUTHOR]

Our experiments in the 2006 INEX ad'hoc track were based on the use of the proximity of the query terms in the documents to rank them. More precisely we define around each occurrence of a query term an influence function. For an occurrence appearing in the text itself, this influence function is linearly decreasing from 1 to 0 depending on the distance to the occurrence. When a query term happens to appear in a title of a structured document its influence is uniformly 1 from the beginning to the end of the (sub-)section. We use boolean queries and these influence functions are combined according to the tree of a query using fuzzy logic. The score of any part of a document is the summation of the resulting influence function at the root of the query tree on the range of this part. We present and comment the results. [ABSTRACT FROM AUTHOR]

The Graph Neural Network is a relatively new machine learning method capable of encoding data as well as relationships between data elements. This paper applies the Graph Neural Network for the first time to a given learning task at an international competition on the classification of semi-structured documents. Within this setting, the Graph Neural Network is trained to encode and process a relatively large set of XML formatted documents. It will be shown that the performance using the Graph Neural Network approach significantly outperforms the results submitted by the best competitor. [ABSTRACT FROM AUTHOR]

In this paper, we present a framework for classifying XML documents based on structure/content similarity between XML documents. Firstly, an algorithm is proposed for computing the edit distance between an ordered labeled tree and a regular hedge grammar. The new edit distance gives a more precise measure for structural similarity than existing distance metrics in the literature. Secondly, we study schema extraction from XML documents, and an effective solution based on minimum length description (MLD) principle is given. Our schema extraction method allows trade off between schema simplicity and precision based on the user's specification. Thirdly, classification of XML documents is discussed. Representation of XML documents based on the structures and contents is also studied. The efficacy and efficiency of our methodology have been tested using the data sets from XML Mining Challenge. [ABSTRACT FROM AUTHOR]

This article is a report concerning the two years of the XML Mining track at INEX (2005 and 2006). We focus here on the classification and clustering of XML documents. We detail these two tasks and the corpus used for this challenge and then present a summary of the different methods proposed by the participants. We last compare the results obtained during the two years of the track. [ABSTRACT FROM AUTHOR]

XML retrieval is a discipline of information retrieval that focuses on the retrieval of specific document fragments that answer a user information need, while exploiting the structural information available in a document. The contribution of this article is twofold. It presents the effective and scalable retrieval model of the XSee search engine for XML documents, and shows that the model is stable over the different collections used for INEX. Furthermore, we discuss how the effectiveness of the retrieval model can be enhanced using different reranking methods based on the structural characteristics. [ABSTRACT FROM AUTHOR]

This paper describes the setup and results of the Max-Planck-Institut für Informatik's contributions for the INEX 2006 AdHoc Track and Feedback task. The runs were produced with the TopX system, which is a top-k retrieval engine for text and XML data that uses a combination of BM25-based content and structural scores. [ABSTRACT FROM AUTHOR]

It is assumed that by focusing retrieval on a granularity lower than documents XML-IR systems will better satisfy users' information need than traditional IR systems. Participants in INEX's Ad-hoc track develop XML-IR systems based upon this assumption, using an evaluation methodology in the tradition of Cranfield. However, since the inception of INEX, debate has raged on how applicable some of the Ad-hoc tasks are to real users. The purpose of the User-Case Studies track is to explore the application of XML-IR systems from the users' perspective. This paper outlines QUT's involvement in this track. For our involvement we conducted a user experiment using an XML-IR system (GPX) and three interfaces: a standard keyword interface, a natural language interface (NLPX) and a query-by-template interface (Bricks). Following the experiment we interviewed the users about their experience and asked them - in comparison with a traditional XML-IR system - what type of tasks would they use an XML-IR system for, what extra information they would need to interact with an XML-IR system and how would they want to see XML-IR results presented. It is hoped that the outcomes of this study will bring us closer to understanding what users want from XML-IR systems. [ABSTRACT FROM AUTHOR]

Digital libraries offer convenient access to large volumes of text, but finding the information that is relevant for a given information need is hard. The workshops of the Initiative for the Evaluation of XML retrieval (INEX) provide a forum for testing the effectiveness of retrieval strategies. In this paper, we present the current version of our search engine that was used for INEX 2006: Like at INEX 2005, our search engine exploits structural patterns - in particular, automatic detection of titles - in the retrieval results to find the appropriate results among overlapping elements. This year, we examine how we can change this method to work better with the Wikipedia collection, which is significantly larger than the IEEE collection used in previous years. We show that our optimizations both retain the retrieval quality and reduce retrieval time significantly. [ABSTRACT FROM AUTHOR]

Despite the active research on XML retrieval, it is a great challenge to determine the contexts where the methods can be applied and where the proven results hold. Therefore, having a common taxonomy for the use cases of XML retrieval is useful when presenting the scope of the research. The taxonomy also helps us design more focused user studies that have an increased validity. In the current state, the taxonomy covers most common uses of applying Information Retrieval to XML documents. We are delighted to see how some of the use cases match the tasks of the INEX participants. [ABSTRACT FROM AUTHOR]

We examine the INEX ad hoc search tasks and ask if (or not) it is possible to identify any existing commercial use of the task. In each of the tasks: thorough, focused, relevant in context, and best in context, such uses are found. Commercial use of CO and CAS queries are also found. Finally we present abstract use cases of each ad hoc task. Our finding is that XML-IR, or at least parallels in other semi-structured formats, is in use and has been for many years. [ABSTRACT FROM AUTHOR]

In this paper we describe the planned setup of the INEX 2006 interactive track. As the track has been delayed and data collection has not been completed before the INEX 2006 workshop, the track will continue into 2007. Special emphasis is put on comparing XML element retrieval with passage retrieval, and on investigating differences between multiple dimensions of the search tasks. [ABSTRACT FROM AUTHOR]

This paper describes experiments carried out with the XFIRM system in the INEX 2006 framework. The XFIRM system uses a relevance propagation method to answer CO and CO+S queries. Runs were submitted to the ad-hoc, relevance feedback and multimedia tracks. [ABSTRACT FROM AUTHOR]

Use of XML offers a structured approach for representing information while maintaining separation of form and content. XML information retrieval is different from standard text retrieval in two aspects: the XML structure may be of interest as part of the query; and the information does not have to be text. In this paper, we describe an investigation of approaches to retrieve text and images from a large collection of XML documents, performed in the course of our participation in the INEX 2006 Ad Hoc and Multimedia tracks. We evaluate three information retrieval similarity measures: Pivoted Cosine, Okapi BM25 and Dirichlet. We show that on the INEX 2006 Ad Hoc queries Okapi BM25 is the most effective among the three similarity measures used for retrieving text only, while Dirichlet is more suitable when retrieving heterogeneous (text and image) data. [ABSTRACT FROM AUTHOR]

This paper presents an experimental study that examines the performance of various combination techniques for content-based image retrieval using a fusion of visual and textual search results. The evaluation is comprehensively benchmarked using more than 160,000 samples from INEX-MM2006 images dataset and the corresponding XML documents. For visual search, we have successfully combined Hough transform, Object's color histogram, and Texture (H.O.T). For comparison purposes, we used the provided UvA features. Based on the evaluation, our submissions show that Uva+Text combination performs most effectively, but it is closely followed by our H.O.T- (visual only) feature. Moreover, H.O.T+Text performance is still better than UvA (visual) only. These findings show that the combination of effective text and visual search results can improve the overall performance of CBIR in Wikipedia collections which contain a heterogeneous (i.e. wide) range of genres and topics. [ABSTRACT FROM AUTHOR]

The multimedia track focuses on using the structure of the document to extract, relate, and combine the relevance of different multimedia fragments. This paper presents a brief overview of the track, it's collection tasks and goals. We also report the results and the approaches of the participating groups. [ABSTRACT FROM AUTHOR]

For this year's INEX UC Berkeley focused on the Heterogeneous track, and submitted only two runs for the Adhoc tasks, a "thorough" run and a "BestInContext" run. For these runs we used the TREC2 probabilistic model with blind feedback. The Thorough run was ranked 21st out of the 106 runs submitted for the Thorough task. The majority of our effort for INEX was on the Heterogeneous (Het) track where we submitted three collection ranking runs and two content-only runs. As the only group to actually submit runs for the Het track, we are guaranteed both the highest, and lowest, effectiveness scores for each task. However, because it was deemed pointless to conduct the actual relevance assessments on the submissions of a single system, we do not know the exact values of these results. [ABSTRACT FROM AUTHOR]

While the primary INEX test collection is based on a single DTD, it is realistic to assume that most XML collections consist of documents from different sources. This leads to a heterogeneity of syntax, semantics and document genre. In order to cope with the challenges posed by such a diverse environment, the heterogeneous track was offered at INEX 2006. Within this track, we set up a collection consisting of several different and diverse collections. We defined retrieval tasks and identified a set of topics. These are the foundations for future run submissions, relevance assessments and proper evaluation of the proposed methods dealing with a heterogeneous collection. [ABSTRACT FROM AUTHOR]

XML information retrieval (XML-IR) systems aim to better fulfil users' information needs than traditional IR systems by returning results lower than the document level. In order to use XML-IR systems users must encapsulate their structural and content information needs in a structured query. Historically, these structured queries have been formatted using formal languages such as NEXI. Unfortunately, formal query languages are very complex and too difficult to be used by experienced - let alone casual - users and are too closely bound to the underlying physical structure of the collection. INEX's NLP task investigates the potential of using natural language to specify structured queries. QUT has participated in the NLP task with our system NLPX since its inception. Here, we discuss the changes we've made to NLPX since last year, including our efforts to port NLPX to Wikipedia. Second, we present the results from the 2006 INEX track where NLPX was the best performing participant in the Thorough and Focused tasks. [ABSTRACT FROM AUTHOR]

This article presents the contribution of the LIPN : Laboratoire d'Informatique de Paris Nord (France) to the NLQ2NEXI (Natural Language Queries to NEXI) task (part of the Natural Language Processing (NLP) track) of the Initiative for Evaluation of XML Retrieval (INEX 2006). It discusses the use of shallow parsing methods to analyse natural language queries. [ABSTRACT FROM AUTHOR]

This paper describes the retrieval approaches used by Queen Mary, University of London in the INEX 2006 ad hoc track. In our participation, we mainly investigate element-specific smoothing method within the language modelling framework. We adjust the amount of smoothing required for each XML element depending on its number of topic shifts to provide a focused access to XML elements in the Wikipedia collection. We also investigate whether using non-uniform priors is beneficial for the ad hoc tasks. [ABSTRACT FROM AUTHOR]

We present a Semi-supervised Machine Learning based ranking model which can automatically learn its parameters using a training set of a few labeled and unlabeled examples composed of queries and relevance judgments on a subset of the document elements. Our model improves the performance of a baseline Information Retrieval system by optimizing a ranking loss criterion and combining scores computed from doxels and from their local structural context. We analyze the performance of our supervised and semi-supervised algorithms on CO-Focussed and CO-Thourough tasks using a baseline model which is an adaptation of Okapi to Structured Information Retrieval. [ABSTRACT FROM AUTHOR]

In this paper we report on the retrieval approach taken by the VALORIA laboratory of the University of South-Brittany while participating at INEX 2006 ad-hoc track with the SIRIUS XML IR system. SIRIUS retrieves relevant XML elements by approximate matching both the content and the structure of the XML documents. A weighted editing distance on XML paths is used to approximately match the documents structure while the IDF of the researched terms are used to rank the textual content of the retrieved elements. We briefly describe the approach and the extensions made to the SIRIUS XML IR system to address each of the four subtasks of the INEX 2006 ad-hoc track. Finally we present and analyze the SIRIUS retrieval evaluation results. SIRIUS runs were ranked on the 1st position out of 77 submitted runs for the Best In Context task and obtained several top ten results for both the Focused and All In Context tasks. [ABSTRACT FROM AUTHOR]

Our submission to the INEX 2006 Ad-hoc retrieval track is described. We study how to utilize the Wikipedia structure (XML documents with hyperlinks) by combining XML and Web retrieval. In particular, we experiment with different combinations of language models and the HITS algorithm. An important feature of our techniques is a filtering phase that identifies the relevant part of the corpus, prior to the processing of the actual XML elements. We analyze the effect of the above techniques based on the results of our runs in INEX 2006. [ABSTRACT FROM AUTHOR]

This paper describes the retrieval approach proposed by the SIG/EVI group of the IRIT research centre at INEX'2006. This XML approach is based on direct contribution of the components constituting an information need. This paper focuses on the impact of changes between INEX'2005 and INEX'2006 notably the corpus change. This paper describes the search engine configurations and evolutions resulting from training on previous INEX testbeds and used to participate to INEX'2006. It presents also the results of the official experiments carried out at INEX'2006 and additional results. [ABSTRACT FROM AUTHOR]

This paper describes the participation of the Information Retrieval and Interaction group of Robert Gordon University in the INEX 2006 ad hoc track. We focused on two questions: "What potential evidence do human assessors use to identify relevant XML elements?" and "How can this evidence be used by computers for the same task?". Our main strategy was to investigate evidence taken not only from the content, but also from the shallow features of how texts were displayed. We employed the vector space model and the language model combining estimates based on element full-text and the compact representation of the element. We analyzed a range of non-content priors to boost retrieval effectiveness. [ABSTRACT FROM AUTHOR]

In this paper we present an extension of information retrieval based on Kullback-Leibler divergence (with backoff smoothing) to support structured queries on structured documents. The proposed method applies to several common retrieval tasks characterized by an implication relationship among texts, including fielded topics and XML documents. We discuss how to choose the method parameters to make the computation of the ranking function efficient. We finally report some experimental results obtained using a loose approximation of the model based on a discriminative selection strategy. [ABSTRACT FROM AUTHOR]

This paper exposes the results of our participation in INEX'06. Two runs were submitted to the Ad Hoc Thorough track obtained with Garnata, our Information Retrieval system for structured documents. We have implemented two different models based on Influence Diagrams, the SID and CID models. The result of this first participation has been very poor. In the paper, we describe the models, the system, and analyse the possible reason of such a bad performance. [ABSTRACT FROM AUTHOR]

We present in this paper our XML information retrieval experiments at INEX 2006 (ad-hoc track). We have developed a Structured Information Retrieval System based on an IR model considering Web documents as "reading paths", instead of a set of atomic and flat pages. Our algorithm is based on information propagation along the reading paths. We present some preliminary results at INEX 2006 (ad-hoc track). [ABSTRACT FROM AUTHOR]

Retrieval queries that combine structural constraints with keyword search are placing new challenges on retrieval systems. This paper presents TReX—a new retrieval system for XML. TReX can efficiently return either all the answers to a given query or only the top-k answers. In this paper, we discuss our participation in the annual Initiative for the Evaluation of XML Retrieval (INEX) workshop in the ad-hoc track. Our main contribution is to investigate the use of summaries and the flexibility they provide when dealing with structural constraints. We describe algorithms for retrieval using summaries. Experimental results are presented showing that TReX answers queries efficiently and effectively. [ABSTRACT FROM AUTHOR]

In previous INEX years we presented an XML component ranking algorithm that was based on separation of nested XML elements to different indices. This worked fine for the IEEE collection which has a small number of potential component types that can be returned as query results. However, such an assumption doesn't scale to this year Wikipedia collection where there is a large set of potential component types that can be returned. We show a new version of the Component ranking algorithm that does not assume any knowledge on the set of component types. We then show some preliminary work we did to exploit the connectivity of the Wikipedia collection to improve ranking. [ABSTRACT FROM AUTHOR]

The INEX 2006 evaluation was based on the Wikipedia collection in XML format. It consisted of several tasks that required different approaches to element selection. In this paper we describe the approach that we adopted in an attempt to satisfy the requirements of all the tasks, Thorough, Focused, Relevant in Context, and Best in Context. We have used the same underlying system to approach all tasks. The retrieval strategy is based on the construction of a collection sub-tree, consisting of all nodes that contain one or more of the search terms. Nodes containing search terms were then assigned a score using the GPX ranking scheme which incorporates TF-IDF or BM25 variants, but extends them. Scores are recursively propagated to ancestors in the document XML tree, and finally all scoring XML elements are ranked. We present results that demonstrate that the approach is versatile and produces consistently good performance. We also provide empirical analysis of the GPX ranking scheme and compare its performance against a baseline TF-IDF and a BM25 scoring scheme. [ABSTRACT FROM AUTHOR]

The Wikipedia XML documents are considered an interesting challenge to any XML retrieval system that is capable of indexing and retrieving XML without prior knowledge of the structure. Although the structure of the Wikipedia XML documents is highly irregular and thus unpredictable, EXTIRP manages to handle all the well-formed XML documents without problems. Whether the high flexibility of EXTIRP also implies high performance concerning the quality of IR has so far been a question without definite answers. The initial results do not confirm any positive answers, but instead, they tempt us to define some requirements for the XML documents that EXTIRP is expected to index. The most interesting question stemming from our results is about the line between high-quality XML markup which aids accurate IR and noisy "XML spam" that misleads flexible XML search engines. [ABSTRACT FROM AUTHOR]

As part of the INEX 2006 Adhoc Track, we conducted a range of experiments with filtering and clustering XML element retrieval results. Our basic retrieval engine retrieves arbitrary elements from the collection (corresponding to the Thorough Task). These runs are filtered to remove textual overlap between elements (corresponding to the Focused Task). The resulting runs can be clustered per article (corresponding to the All in Context Task). Finally, we select the "best" element for each article (corresponding to the Best in Context Task). Our main findings are the following. First, a complete element index outperforms a restricted index based on section-structure, albeit the differences are small. Second, grouping non-overlapping elements per article does not lead to performance degradation, but may improve scores. Third, all restrictions of the "pure" element runs (by removing overlap, by grouping elements per article, or by selecting a single element per article) lead to some but only moderate loss of precision. [ABSTRACT FROM AUTHOR]

In this paper, we describe the Centre for Interactive Systems Research's participation in the INEX 2006 adhoc track. Rather than using a field-weighted BM25 model in INEX 2005, we revert back to using the traditional BM25 weighting function. Our main research aims in this year are to investigate the effects of document filtering (by considering only the top-ranked documents), element filtering by length cut-off and the effect of using phrases. The initial results show the latter two methods did not do well, while the first one did well on FOCUSED TASK and RELEVANT IN CONTEXT TASK. Finally, we propose a novel method for BEST IN CONTEXT TASK, and present our initial results. [ABSTRACT FROM AUTHOR]

This paper describes our methodology for the dynamic retrieval of XML elements, an overview of its implementation in a structured environment, and the challenges introduced by applying it to the INEX Wikipedia [4] collection, which can more aptly be described as semi-structured. Our system is based on the vector space model [9] and its basic functions are performed using the Smart experimental retrieval system [8]. A major change in the system this year is the incorporation of a method for the dynamic computation of query term weights [6] to be correlated with the dynamically generated and weighted element vectors. Dynamic element retrieval requires only a single indexing of the document collection at the level of the basic indexing node (in this case, the paragraph). It returns a rank-ordered list of elements equivalent to that produced by the same query against an all-element index of the collection. (A detailed description of this method appears in [1].) As we move from a well structured collection, such as the INEX IEEE documents, to Wikipedia, changes in the structure of the articles must be accommodated. [ABSTRACT FROM AUTHOR]

We used a flexible XML retrieval system for evaluating structured document retrieval and multimedia retrieval tasks in the context of the INEX 2006 benchmarks. We investigated the differences between article and element retrieval for Wikipedia data as well as the influence of an elements context on its ranking. We found that article retrieval performed well on many tasks and that pinpointing the relevant passages inside an article may hurt more than it helps. We found that for finding images in isolation the associated text is a very good descriptor in the Wikipedia collection, but we were not very succesful at identifying relevant multimedia fragments consisting of a combination of text and images. [ABSTRACT FROM AUTHOR]

In this year's INEX participation, CSIRO took part in the Ad-hoc Track, contributing to three of the four given tasks, namely the Thorough Task, the Focused Task and the Best in Context Task. We relied on our own text-and-metadata retrieval system PADRE for indexing the data and processing the queries. Since PADRE is designed to retrieve documents rather than XML elements we preprocessed the collection to enable retrieval of sub-elements. From the set of queries we identified all the element-types required to be retrieved. We then added new pseudo documents corresponding to all the retrievable elements from the originals. Finally, this expanded collection was indexed with PADRE. When processing queries, query elements were extracted from the INEX topics and pre-processed to generate queries in PADRE syntax. Results were post-processed to achieve the requirements of each particular task, such as elimination of super- and sub-elements of elements already retrieved. Results obtained from this simple-minded approach were not particularly competitive but serve as a good basis for identification of necessary future enhancements. [ABSTRACT FROM AUTHOR]

We developed a passage retrieval system for XML documents using the vector space model. To be more flexible for the query, we also developed a method of unification of multiple retrieved elements and a fragment indexing system. Our system is composed of an inverted file and an XML Path Language (XPath) path list. The validity of the method was tested as part of the ad hoc track in the Initiative for the Evaluation of XML Retrieval (INEX) 2006. [ABSTRACT FROM AUTHOR]

This article presents the general Wikipedia XML Collection developped for Structured Information Retrieval and Structured Machine Learning. This collection has been built from the Wikipedia Enclyclopedia. We detail particularly here which parts of this collection have been used during INEX 2006 for the Ad-hoc track and for the XML Mining track. Note that other tracks of INEX - multimedia track for example - have also been based on this collection. [ABSTRACT FROM AUTHOR]

This paper describes the official measures of retrieval effectiveness employed at the ad hoc track of INEX 2006. [ABSTRACT FROM AUTHOR]

The paper investigates the measures of retrieval effectiveness employed in the ad-hoc track of INEX 2006. In particular, it looks at how the evaluation of the Focused task is affected when different methodology is employed in generating the so-called ideal recall-base, which forms the ground-truth of the evaluation. The results show that the choice of methodology can impact on the obtained performance scores and the relative ranking of systems in relation to each other, especially when the effectiveness scores are uniformly low across all systems. Most XCG measures show very similar levels of sensitivity to changes in the ideal recall-base. [ABSTRACT FROM AUTHOR]

Since 2002, INEX has been working towards the goal of establishing an infrastructure, in the form of a large XML test collection and appropriate scoring methods, for the evaluation of content-oriented XML retrieval systems. This paper provides an overview of the work carried out as part of INEX 2006. [ABSTRACT FROM AUTHOR]

The web is currently the information searching and browsing environment of choice for scholars and lay users alike. The goal of most cultural heritage applications is to interest a large audience, and therefore web interfaces are being developed even when part of their functionality is not offered to the general public. We present a web-based interface for managing, browsing and searching a repository of historic documents. The documents pertain to a region which has been an important regional power in medieval times and their originals are under the custody of the Portuguese national archives. The challenges of the project came from its requisites in three aspects: rigorous archival description, the incorporation of document analysis and a flexible search interface. The system is an instance of a multimedia database framework providing both browse and retrieval functionalities to end users and configuration and content management services to the collection administrators. [ABSTRACT FROM AUTHOR]