Michael Fuller

Using clustering and classification approaches in interactive retrieval

Satisfying non-trivial information needs involves collecting information from multiple resources, and synthesizing an answer that organizes that information. Traditional recall/precision-oriented information retrieval focuses on just one phase of that process: how to efficiently and effectively identify documents likely to be relevant to a specific, focused query. The TREC Interactive Track has as its goal the location of documents that pertain to different instances of a query topic, with no reward for duplicated coverage of topic instances. This task is similar to the task of organizing answer components into a complete answer. Clustering and classification are two mechanisms for organizing documents into groups. In this paper, we present an ongoing series of experiments that test the feasibility and effectiveness of using clustering and classification as an aid to instance retrieval and, ultimately, answer construction. Our results show that users prefer such structured presentations of candidate result set to a list-based approach. Assessment of the structured organizations based on the subjective judgement of the experiment subjects suggests that the structured organization can be more effective; however, assessment based on objective judgements shows mixed results. These results indicate that a full determination of the success of the approach depends on assessing the quality of the final answers generated by users, rather than on performance during the intermediate stages of answer construction.

Structured answers for a large structured document collection

There is a simple method for integrating information retrieval and hypertext. This consists of treating nodes as isolated documents and retrieving them in order of similarity. If the nodes are structured, in particular, if sets of nodes collectively constitute documents, we can do better. This paper shows how the formation of the hypertext, the retrieval of nodes in response to content based queries, and the presentation of the nodes can be achieved in a way that exploits the knowledge encoded as the structure of the documents. The ideas are then exemplified in an SGML based hypertext information retrieval system.

Architecture of a content management server for XML document applications

Describes the data model that is used to implement the SIM content management server (CMS), an SGML/XML-native content server that is designed to support extremely fast data access to and dynamic updating of 100-GByte collections under high loads. This paper describes the requirements for supporting text-intensive applications and for building XML/SGML document management solutions. The SIM CMS employs a data model that is designed to directly support SGML and XML; this model is described, and a comparison with other models based on general-purpose database management systems is made.

Searcher performance in question answering

There are many tasks that require information finding. Some can be largely automated, and others greatly benefit from successful interaction between system and searcher. We are interested in the task of answering questions where some synthesis of information is required-the answer would not generally be given from a single passage of a single document. We investigate whether variation in the way a list of documents is delivered affected searcher performance in the question answering task. We will show that there is a significant difference in performance using a list customized to the task type, compared with a standard web-engine list. This indicates that paying attention to the task and the searcher interaction may provide substantial improvement in task performance.

Querying in a Large Hyperbase

This paper proposes a query language and data organisation for large hyperbase systems. The language allows queries to involve the links as well as the text. The architecture allows the data to be efficiently indexed, as well as supporting multiple users.

Integration of Information Retrieval and Hypertext Via Structure

If we wish to access information from a large collection of unrelated abstracts, then a method that finds all potentially relevant information and presents it in ranked order is likely to be very helpful. If we have a well structured body of small pieces of information, then browsing through the information is likely to be very helpful. That is, there are situations in which information retrieval and in which hypertext are the ideal method of accessing information.

System architectures for structured document data

Semi-structured data, including but not limited to structured documents, has speci c characteristics and is used in ways di erent to tabular data. SGML and XML are widely used to represent information of this type. The demands on systems that manage semi-structured data vary from those on traditional relational systems. This paper reviews the nature and characteristics of semi-structured data, and the functional needs of those applications, including query requirements, document description, manipulation, and document management needs. It examines alternative physical models for semi-structured data, and evaluates and compares alternative system architectures.

Hyperbase Systems: A Structured Architecture.

JOHN L. SCHNASE JOHN J. LEGGETT, DAVID L. HICKS, PETER J. NUERNBERG, AND J. ALFREDO SÁNCHEZ Advanced Technology Group Hypermedia Research Laboratory School of Medicine Library and Biomedical Department of Computer Science Communications Center Texas A&M University Washington University School of Medicine College Station, Texas 77843, USA 660 South Euclid Avenue (Box 8132) St. Louis, Missouri 63110, USA

Case Study: Managing Legislation

We now consider this model of the document lifecycle in relation to a real production system. EnAct is a complex system for managing the complete document lifecycle of legislation.1 Built by RMIT University’s Multimedia Database Systems group, it combines recent research with commercial office solutions experience in government and the private sector. EnAct was commissioned by the government of Tasmania, a state of Australia, primarily for use by the Office of Parliamentary Counsel (OPC), an office of professional legislative drafters who provide drafting services to the legislature. The OPC is a true document factory, with a virtual monopoly on all legislative drafting within Tasmania. They exist exclusively to produce documents and to draft Bills, Statutory Rules, their accompanying documentation, and, increasingly, the final published versions of Acts and Statutory Rules. The major motivation of the project was to develop a system to produce and manage an electronic repository of legislation to track and record legislation as it changes with time, allowing access to the legislation both as it is now and also as it was at any time in the past. Both electronic and paper publication were to be sourced from this central electronic repository.

Electronic Document Description

We have characterized documents as records of messages that have content, structure and metadata. There may be several ways of describing messages using these three constituents. The choice of which to use depends on how the messages themselves are used and accessed. For example, authorship of a document may be part of the content of a document, an attribute of a document, or both. Similarly, it may not be easy to extract the structure of a document, and it may be helpful to do so only if certain parts of a document are needed or if documents with particular structural properties are required.