J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: sentence selection and evaluation metrics. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval, pages 121--128, NY, USA, 1999. ACM Press.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Text Summarization) project on multi document summarization. To select important content, we used techniques that proved effective in single document summarization such as sentence position [1] term frequency [13] topic signature [11,7] and term clustering. To remove redundancy, we used MMR [3]. To improve cohesion and coherence, we used stigma word filters [4] and time stamps. Although most of the individual techniques are not new, assembling them and applying them to multi document summarization is new. Also, including lead sentences to ensure coherence is new, and turned out to be ....

....than 5 words, then take the next one. For example (where x.y stands for document number . sentence number) 4.3, 6.6, 2.5, 5.2. 4.1, 4.3, 6.1, 6.6, 2.1, 2.5, 5.1, 5.2. 3. 4 Filter for Length Select the required number of sentence pairs using a simplified version of CMU s MMR algorithm [3]: 4.a include first pair 4.b using a simplified version of MMR, find the sentence pair most different from the included ones, and include it too. In the DUC 2001 implementation, NeATS did not consider the sentence pair, just the sentence. This caused some degradation. 4.c repeat step 4.b ....

Goldstein, J., M. Kantrowitz, V. Mittal, and J. Carbonell. 1999. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. Proceedings of the 22 nd International ACM Conference on Research and Development in Information Retrieval (SIGIR-99), Berkeley, CA, 121--128.

.... be achieved by developing annotated corpora which can provide reference data to support intrinsic evaluations, whether in the form of a representation of the input source document, as in the Q A annotated mini corpus, or in the form of reference summaries linked to their sources [13] 5] 22] [8], etc. 4 Presentation Strategies Unfortunately, very little work has been done in this area. 24] evaluated the summarization component of their Broadcast News Navigator, which performs an analysis of news video along with the closedcaptioned text. The summaries used lists of topic terms and ....

Goldstein, J., Kantrowitz, M., Mittal, V., and Carbonell, J. 1999. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. Proceedings of the 22nd International Conference on Research and Development in Information Retrieval (SIGIR'99), pp. 121-128.

.... or to determine whether it is relevant to a query [11, 30] Can a reader correctly answer a reading comprehension test using the summary [26] Can a reader assign the correct keywords to a summary [36] Compare agreement between sentences selected by humans and sentences selected by computer [35, 13], or compare agreement in the ranks of sentences that a system generates [9] The first evaluation is not actually an evaluation, though it is useful for giving a sense of a summarizer s capabilities. The next set of evaluations all require that a human be part of a evaluation loop. This approach ....

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of SIGIR, pages 121--128, 1999.

....to the set of sentences that is known to be a good summary. To the extent that an approach chooses the right sentences, that approach is good; when it veers wildly from the ideal set, the approach is inappropriate to the task. Our approach is similar in spirit to other sentence based evaluations [20, 8, 7], but is modified significantly to take into account the time based nature of our summaries. We formalize the temporal summarization problem as follows. A news topic is made up of a set of events and is discussed in a sequence of news stories. Most sentences of the news stories discuss one or ....

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of SIGIR, pages 121--128, 1999.

....indicative summaries can help a user to decide the relevance of the document and user focused summaries can present the content that is closely related to the initial user query. In order to meet these requirements, a few approaches tried to apply a query expansion technique to text summarization[2, 3]. These approaches use the query expansion over a set of sentences in a document, while the user query is used for an initial query. In these approaches, however, it is observed that the summaries may be inappropriate because the feedback query is biased during the query expansion. In this paper, ....

Jade Goldstein, Mark Kantrowitz, Vibhu Mittal, and Jaime Carbonell. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Proceedings of ACM-SIGIR'99. 1999, pp.121--128.

....queries, in whichtherewas no limit on the number of sentences extracted. We compared these no limit summaries (with a sentence average of 41) to the fixed 10 sentence summaries (see Table 1) as well as characteristics of single document summaries for the newswire genre from our previous work [8]. For multi document10 sentence summaries, the assessors used on average 1.3 first 169 1. 2 25 I heard a terrible crash. and) thought at first that we had collided with an elk, Jeanette Haug, 23, told Norway s NTB news agency. 2. 31ASTA, Norway (Reuters) Norwegian rescue workers will ....

Jade Goldstein, Mark Kantrowitz, Vibhu O. Mittal, and Jaime Carbonell. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Proceedings of SIGIR-99,Berkeley, CA, August 1999.

....This approach has several drawbacks, including the inability to generate e ective summaries shorter than a sentence. This is problematic when short headline style summaries with only a few words are desired because (1) sentences with summary content are actually usually longer than average [2], and (2) information in the document is often scattered across multiple sentences; extractive summarization cannot combine concepts in di erent text spans of the source document without using the whole spans. We describe an alternative approach to summarization, not based on sentence extraction, ....

Goldstein, J., Kantrowitz, M., Mittal, V. O., and Carbonell, J. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Proceedings of the 22nd ACM SIGIR Conference (SIGIR-99) (1999).

.... unseen trigrams) using the Good Turing estimate [9] For the final two steps we used the publicly available CMU Cambridge Language Modelling Toolkit [4] 6 Evaluation Summarization research has grappled for years with the issue of how to perform a rigorous evaluation of a summarization system [8, 10, 12]. We have not solved that problem here, but nonetheless present a series of quantitative and qualitative assessments of the functionality of the various components of ocelot. 6.1 Measuring word overlap We begin by examining the behavior of the simplest of the proposed gisting algorithms. To this ....

Goldstein, J., Kantrowitz, M., Mittal, V. O., and Carbonell, J. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Proceedings of the 22nd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR-99) (Berkeley, CA., 1999), pp. 121--128.

....by how well they perform on certain (extrinsic tasks [23] indicating document relevance to a topic, indicating a category for the document or intrinsic tasks whether they contain answers to specific questions. Summaries can also be evaluated by whether they extract the relevant portions of text [13,28] (intrinsic) which is the focus of my research. 2.2 Multi Document In the past few years, multi document summarization has become a subject of great interest due to the rapid increase in textual information, which has made it virtually impossible for users to browse or read many individual ....

....for multiple languages, the above items are the principal foci of my proposed work. 4.1. 1 Passage Unit Choice Using the sentence as a unit may be suitable for single document summarization as demonstrated by my evaluation results as well as the results of TIPSTER and other evaluations [13,17,23,28,41], however, it is not clear what is the most appropriate unit for multi document summarization. Incorrect referents can quickly mislead readers when sentences with pronouns are strung together from multiple documents. Thus I will investigate whether multiple sentences that include the referent need ....

[Article contains additional citation context not shown here]

J. Goldstein, M. Kantrowitz, V. O. Mittal, and J. Carbonell. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Proceedings of SIGIR-99, Berkeley, CA, August 1999.

No context found.

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: sentence selection and evaluation metrics. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval, pages 121--128, NY, USA, 1999. ACM Press.

No context found.

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of the 22nd ACM SIGIR, pages 121--128, 1999.

No context found.

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of the ## Annual International ACM SIGIR, pages 121--128, 1999.

No context found.

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of the 22 Annual International ACM SIGIR, pages 121--128, 1999.

No context found.

J. Goldstein, M. Kantrowitz, V. O. Mittal, and J. G. Carbonell. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. In Research and Development in Information Retrieval, pages 121--128, 1999.

No context found.

Jade Goldstein, Mark Kantrowitz, Vibhu Mittal, and Jamie Carbonell. Summarizing text documents: sentence selection and evaluation metrics. In Proceedings of the 22nd ACM SIGIR Conference on Research and Development in Information Retrieval, Berkeley, CA, 1999.

No context found.

Jade Goldstein, Mark Kantrowitz, Vibhu O. Mittal, and Jaime G. Carbonell. Summarizing text documents: Sentence selection and evaluation metrics. In Research and Development in Information Retrieval, pages 121--128, Berkeley, California, 1999.

No context found.

J. Goldstein, M. Kantrowitz, V. Mittal, and J. Carbonell. Summarizing text documents: Sentence Selection and Evaluation Metrics. Proceedings of SIGIR 1999.

No context found.

Goldstein, J., Kantrowitz, M., Mittal, V., and Carbonell, J. Summarizing text documents: Sentence selection and evaluation metrics. In Proceedings of the 22nd Annual International ACM SIGIR conference on Research and development in information retrieval (Berkeley, California, 8 1999), pp. 121--128.

No context found.

Goldstein, J., M. Kantrowitz, V. Mittal, and J. Carbonell. 1999. Summarizing Text Documents: Sentence Selection and Evaluation Metrics. Proceedings of the 22 nd International ACM Conference on Research and Development in Information Retrieval (SIGIR-99), Berkeley, CA, 121--128.

No context found.

Goldstein, J., Kantrowitz, M., Mittal, V. and Carbonell, J.: Summarizing Text Documents: Sentence Selection and Evaluation Metrics, ACM SIGIR '99 Proceegings, pp. 121-128 (1999).

No context found.

Jade Goldstein, Mark Kantrowitz, Vibhu Mittal, and Jaime Carbonell, Summarizing Text Documents: Sentence Selection and Evaluation Metrics, In Proceedings of ACM-SIGIR'99, Berkeley, CA, August 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC