A User-Centered Approach to Evaluating Topic Models

Kelly, Diane; Dı́az, Fernando; Belkin, Nicholas J.; Allan, James

doi:10.1007/978-3-540-24752-4_3

Cited by 17 publications

(8 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They proposed a generative probabilistic mixture model for comparative text mining. Kelly et al [50] investigated the techniques used for topic clustering of documents and discovered that most methods performed poorly when evaluated according to the user-defined topic classes. Mei and Zhai [51] used statistical language models to perform a special temporal text mining task, which is studying the patterns of themes in text.…”

Section: Text Classification and Categorization Modelsmentioning

confidence: 99%

“…TDT consists of five tasks, topic 742 H. C. OZMUTLU tracking, link detection, topic detection, first story detection and story segmentation [56]. Topic modeling has been performed significantly in the context of topic detection and tracking [50]. The significant works on TDT can be listed as those of Feng and Allan [56], Larkey et al [57], and Kumaran and Allan [58,59].…”

Section: Text Classification and Categorization Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Markovian analysis for automatic new topic identification in search engine transaction logs

Özmutlu¹

2009

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

Topic analysis of search engine user queries is an important task, since successful exploitation of the topic of queries can result in the design of new information retrieval algorithms for more efficient search engines. Identification of topic changes within a user search session is a key issue in analysis of search engine user queries. This study presents an application of Markov chains in the area of search engine research to automatically identify topic changes in a user session by using statistical characteristics of queries, such as time intervals, query reformulation patterns and the continuation/shift status of the previous query. The findings show that Markov chains provide fairly successful results for automatic new topic identification with a high level of estimation for topic continuations and shifts.

show abstract

Section: Text Classification and Categorization Modelsmentioning

confidence: 99%

Section: Text Classification and Categorization Modelsmentioning

confidence: 99%

Markovian analysis for automatic new topic identification in search engine transaction logs

Özmutlu¹

2009

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

show abstract

“…One problem with all of these approaches is that they are more system-or query-centered, than user-centered, and have not leveraged context. Although some systems have attempted to represent system-defined context through automatic clustering, such clusters do not necessarily reflect how users would cluster their documents [17,22].…”

Section: Introductionmentioning

confidence: 99%

Contextual relevance feedback

Harper

Kelly

2006

Proceedings of the 1st International Conference on Information Interaction in Context - IIiX

Self Cite

View full text Add to dashboard Cite

We present results of a preliminary study of a pile-based information retrieval interface that supports contextual relevance feedback. We designed two interfaces based on the pile metaphor, one which supported contextual relevance feedback and the other which did not, and conducted a within-subjects laboratory evaluation with 24 subjects. Results demonstrate support for the pile-based approach generally, with users indicating that it assisted them in structuring and managing their search results. There were significant differences between usability assessments of each interface and in the interactions in which subjects engaged while they were seeking information. Although we found no significant differences in performance between interfaces, subjects showed a strong preference for the contextual relevance feedback interface over the other, and believed that they did better with that interface.

show abstract

“…In addition, retention (or saving) has been found to be potentially useful as implicit evidence for identifying documents that could be used as seeds for re-weighting (Kelly, Diaz, Belkin, & Allan, 2004). This is quite intuitive.…”

Section: Rlb Personalization Using Search Behaviorsmentioning

confidence: 99%

Personalization in text information retrieval: A survey

Liu

Belkin

2019

Asso for Info Science & Tech

Self Cite

View full text Add to dashboard Cite

Personalization of information retrieval (PIR) is aimed at tailoring a search toward individual users and user groups by taking account of additional information about users besides their queries. In the past two decades or so, PIR has received extensive attention in both academia and industry. This article surveys the literature of personalization in text retrieval, following a framework for aspects or factors that can be used for personalization. The framework consists of additional information about users that can be explicitly obtained by asking users for their preferences, or implicitly inferred from users' search behaviors. Users' characteristics and contextual factors such as tasks, time, location, etc., can be helpful for personalization. This article also addresses various issues including when to personalize, the evaluation of PIR, privacy, usability, etc. Based on the extensive review, challenges are discussed and directions for future effort are suggested.

show abstract

A User-Centered Approach to Evaluating Topic Models

Cited by 17 publications

References 14 publications

Markovian analysis for automatic new topic identification in search engine transaction logs

Markovian analysis for automatic new topic identification in search engine transaction logs

Contextual relevance feedback

Personalization in text information retrieval: A survey

Contact Info

Product

Resources

About