No abstract available.

Article Share on Natural language and the information layer Author: Karen Spärck Jones University of Cambridge, Cambridge, UK University of Cambridge, Cambridge, UKView Profile Authors Info & Claims SIGIR '07: Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrievalJuly 2007 Pages 3–6https://doi.org/10.1145/1277741.1277744Online:23 July 2007Publication History 0citation689DownloadsMetricsTotal Citations0Total Downloads689Last 12 Months6Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

Three times a year I get my copy of a wholly respectable mainstream linguistics journal. Its scholarly articles are rich in examples from varied languages, and alongside these detailed analyses it advances theoretical claims and counterclaims. Its many reviews point to much more of the same. But this journal content is of interest here for another reason than these scholarly ones. First, references to computing are conspicuous by their absence. Just occasionally, grammar types or semantic models appear that have computational connections, for example in a shared view of feature sets; and there are references to computational corpus analysis, though more often in reviews than in major articles. Very very occasionally, there are articles that are more manifestly computational ones, for instance in applying unification to syntactic structures. But in general, the notion that computation in a serious sense, not just as some highly abstract grounding or, maybe, politically correct meta-reference, has something important to say to linguistics never figures. Does this matter? Specifically, whether it matters to the linguists or not, does it matter to us (computational types)? In a world of proliferating specialist journals and, increasingly, conference proceedings, why worry about the lack of computational reference in mainstream linguistics journals? Being computational is not the only way of being legitimate in any field, but there are plenty of publishing venues for the computational. So it need not be a criticism of a linguistics journal like the one I get that it does not have more on computational linguistics, or natural language processing, or natural language engineering, or human language technology. We may find the lack of computational reference a surprise, but even if there are a lot of non-computational linguists out there (as there indubitably are), why should this be a problem? The linguists may argue they have things we ought to learn from them, and it’s up to us to bridge the gap. We may argue just the reverse. But in my view there is a deeper problem. This has to do with the fact that as far as I can tell, the journal I have referred to, and other linguistics journals, are dominated by some Chomskyan paradigm. But does it matter that the linguistic zeitgeist is Chomskyan? Further, does it matter precisely which Chomskyan paradigm: older or newer, broader or more specific, pure or modified?

This article offers a personal perspective on the development of language and information processing over the last half century, focusing on the use of statistical methods. Introduced, with computers, in the 1950s, these have not always been highly regarded, but were revived in the 1990s. They have proved effective in more ways than might have been expected, and encourage new thinking about what language and information processing involve.

This paper examines the use of generic summaries for indexing in information retrieval. Our main observations are that: (1) With or without pseudo-relevance feedback, a summary index may be as effective as the corresponding fulltext index forprecision-oriented search of highly relevant documents. %43 But a reasonably sophisticated summarizer, using a compression ratio of 10-30%, is desirable for this purpose. (2) In pseudo-relevance feedback, using a summary index at initial search and a fulltext index at final search is possibly effective for precision-oriented search, regardless of relevance levels. This strategy is significantly more effective than the one using the summary index only and probably more effective than using summaries as mere term selection filters. %the use of summaries as mere term selection filters. %The summary quality is probably not a critical factor for this strategy, For this strategy, the summary quality is probably not a critical factor, and a compression ratio of 5-10% appears best.

No abstract available.

The effects of out-of-vocabulary (OOV) items in spoken document retrieval (SDR) are investigated. Several sets of transcriptions were created for the TREC-8 SDR task using a speech recognition system varying the vocabulary sizes and OOV rates, and the relative retrieval performance measured. The effects of OOV terms on a simple baseline IR system and on more sophisticated retrieval systems are described. The use of a parallel corpus for query and document expansion is found to be especially beneficial, and with this data set, good retrieval performance can be achieved even for fairly high OOV rates.

This paper describes the spoken document retrieval system that we have been developing and assesses its performance using automatic transcriptions of about 50 hours of broadcast news data. The recognition engine is based on the HTK broadcast news transcription system and the retrieval engine is based on the techniques developed at City University. The retrieval performance over a wide range of speech transcription error rates is presented and a number of recognition error metrics that more accurately reflect the impact of transcription errors on retrieval accuracy are defined and computed. The results demonstrate the importance of high accuracy automatic transcription. The final system is currently being evaluated on the 1998 TREC-7 spoken document retrieval task.

No abstract available.

This paper reviews the Video Mail Retrieval (VMR) project at Cambridge University and ORL. The VMR project began in September 1993 with the aim of developing methods for retrieving video documents by scanning the audio soundtrack for keywords. The project has shown, both experimentally and through the construction of a working prototype, that speech recognition can be combined with information retrieval methods to locate multimedia documents by content. The final version of the VMR system uses pre-computed phone lattices to allow extremely rapid word spotting and audio indexing, and statistical information retrieval (IR) methods to mitigate the effects of spotting errors. The net result is a retrieval system that is open-vocabulary and speaker-independent, and which can search audio orders of magnitude faster than real time.

The goal of the video mail retrieval (VMR) project is to integrate state-of-the-art document retrieval methods with speech recognition to yield a robust and efficient retrieval system. The work presented extends VMR towards an open-vocabulary, talker-independent system for retrieving spontaneously-spoken audio and video messages. We present results showing successful retrieval using a standard large-vocabulary (LV) recogniser, despite the lack of a matched language model and vocabulary. We further show that integrating a LV recogniser with conventional word spotting (WS) gives more robust retrieval performance than either method alone. This paper gives details of the message archive used, the speech recognition methodologies, the information retrieval methods, and experimental results.

This paper presents recentwork on amultimedia retrieval project at Cambridge University and Olivetti Research Limited (ORL). We present novel techniques that allow extremely rapid audio indexing, at rates approaching several thousand times real time. Unlike other methods, these techniques do not depend on a fixed vocabulary recognition system or on keywords that must be known well in advance. Using statistical methods developed for text, these indexing techniques allow rapid and efficient retrieval and browsing of audio and video documents. This paper presents the project background, the indexing and retrieval techniques, and a video mail retrieval application incorporating content-based audio indexing, retrieval, and browsing.

This paper presents domain-independent methods of spoken document retrieval.Both a continuous-speech large vocabulary recognition system, and a phone-lattice word spotter,

The goal of the video mail retrieval project is to integrate state-of-the-art document retrieval methods with high accuracy word spotting to yield a robust and efficient retrieval system. This paper describes a preliminary study to determine the extent to which retrieval precision is affected by word spotting performance. It includes a description of the database design, the word spotting algorithm, and the information retrieval method used. Results are presented which show audio retrieval performance very close to that of text.

The goal of the Video Mail Retrieval (VMR) project is to integTate state-of-the-art information retrieval (IR) methods with high-accuracy word spotting to yield a robust and efficient multimedia retrieval system. This paper concerns open-talker and arbitrary-keyword retrieval based on talker-independent subword models. Because talker-independent subword models can not be expected to work as well as the talker-dependent whole-keyword models used in previous VMR experiments, speaker adaptation is investigated as a means of improving perfor- mance (especially for talkers with non-British accents). Both standard FOM word spotting measures and actual retrieval results are computed. The results show that the FOM is not necessarily a good indicator of retrieval performance, and that talker adaptation can substantially improve both spotting and retrieval results.

This paper presents current work on a video retrieval project at Cambridge University and Olivetti Research Limited (ORL). We show that statistical methods developed for text retrieval are also effective for retrieving and browsing multimedia documents. These methods allow rapid retrieval of news broadcasts by information content determined from teletext subtitles. Information retrieval results for experiments performed on a large archive of news broadcasts are presented. This is made possible by the ORL Medusa system, which allows practical recording, storage, and playback of tens of gigabytes of multimedia data. This work is a step towards practical retrieval of multimedia documents, where the information content is determined from speech recognition performed on the audio soundtrack. We describe the project background, the ORL Medusa multimedia system, and retrieval application, as well as the news broadcast corpus and methods of browsing the retrieved news stories.

This paper describes a current research project investigating belief revision in intelligent systems by modelling the librarian in interaction with a literature-seeking user. The work is designed both to test a theory of agent behaviour based on belief revision proposed by Galliers, and to evaluate a model of the librarian developed by Belkin, Brooks and Daniels, through computational implementation. Agent communication is seen as motivated by and motivating belief changes, where belief revision is determined by coherence, combining endorsement, connectivity and conservatism. The librarian is viewed as a distributed expert system with many individual specialised functions operating in particular belief domains. The paper describes our first implementation of the belief revision mechanism and of a very primitive librarian, designed to test the basic viability of our ideas and to allow us to explore different forms of the distributed system architecture.

This opinion article presents a view of the scope for artificial intelligence in information retrieval. It considers four potential roles for AI in IR, evaluating AI from a realistic point of view and within a wide information management context. The conclusion is that AI has limited potential, not just because AI is itself insufficiently developed, but because many information management tasks are properly shallow information processing ones. There is nevertheless an important place for specific applications of AI or AI-derived technology when particular constraints can be placed on the information management tasks involved. © 1991 John Wiley & Sons, Inc.

This paper is in two parts, following the suggestion that I first comment on my own past experience in information retrieval, and then present my views on the present and future.

No abstract available.

A linguistically motivated approach to indexing, that is the provision of descriptive terms for texts of any kind, is presented and illustrated. The approach is designed to achieve good, i.e. accurate and flexible, indexing by identifying index term sources in the meaning representations built by a powerful general purpose analyser, and providing a range of text expressions constituting semantic and syntactic variants for each term concept. Indexing is seen as a legitimate form of shallow text processing, but one requiring serious semantically based language processing, particularly to obtain well-founded complex terms, which is the main objective of the project described. The type of indexing strategy described is further seen as having utility in a range of applications environments.

Abstract This paper examines statistical techniques for exploiting relevance information to weight search terms. These techniques are presented as a natural extension of weighting methods using information about the distribution of index terms in documents in general. A series of relevance weighting functions is derived and is justified by theoretical considerations. In particular, it is shown that specific weighted search methods are implied by a general probabilistic theory of retrieval. Different applications of relevance weighting are illustrated by experimental results for test collections.

Abstract Indexing exhaustivity, which may be broadly defined as the number of terms assigned to a document, is thought to be of some importance in retrieval, and it has been suggested that there may be an optimal level of exhaustivity for a particular collection. Experiments with two distinct collections, using three levels of indexing exhaustivity for both documents and requests, show that substantially the same performance is obtained for very different levels of document indexing, if suitable choices are made of request level.

Computer programs for automatic classification are a desideratum in many fields. Work on suitable procedures for handling large bodies of object/property descriptions has been in progress at the Cambridge Language Research Unit for some years: this paper describes the current series of general-purpose programs which have been developed there, in which classes or “clumps” of objects are obtained, using a similarity matrix, by a simple iterative scan of the universe of objects, distributing them in such a way that an appropriate cohesion function is minimized. This actual clump-finding process is embedded in an overall package in which the information given by a classification is manipulated in a variety of ways. The current applications of the programs, especially for information retrieval, are described.

L'Automatisation des Recherches Documentaires: un modèle générale, Le SYNTOL R. C. Cros, J. C. Gardin and F. Levy , 1964 : 260. (Paris : Gauthier-Villars , 30F.)