Employing web search query click logs for multi-domain spoken language understanding

Hakkani‐Tür, Dilek; Tür, Gökhan; Heck, Larry; Çelikyılmaz, Aslı; Fidler, Ashley; Hillard, Dustin; Iyer, Rishabh; Parthasarathy, S.

doi:10.1109/asru.2011.6163968

Cited by 12 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our work follows the line of work by [5,8,9]. [5] shows exploiting web query click logs using a semi-supervised method outperforms the fully supervised approach using limited annotated data on domain classification.…”

Section: Related Workmentioning

confidence: 94%

Understanding computer-directed utterances in multi-user dialog systems

Wang¹,

Hakkani‐Tür

Tür

2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

This work aims to understand user requests when multiple users are interacting with each other and a spoken dialog system. More specifically, we explore the use of multi-human conversational context to improve domain detection in a human-computer interaction system. We investigate the different effects of human-directed context and computer-directed context, and compare the impact of using different context window sizes. Furthermore, we employ topic segmentation to chunk conversations for determining context boundaries. The experimental results show that the use of conversational context helps reduce domain detection error rate, especially in some specific domains. And though computer directed context is more reliable, the results show that the combination of both computer and human addressed utterances within a reasonable window size performs the best.

show abstract

Section: Related Workmentioning

confidence: 94%

Understanding computer-directed utterances in multi-user dialog systems

Wang¹,

Hakkani‐Tür

Tür

2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Much of the later work on supervised learning focused on CRFs, for example (Sarikaya et al, 2016), or neural networks (Deoras and Sarikaya, 2013;Yao et al, 2013;Liu et al, 2015;Celikyilmaz and HakkaniTur, 2015). Unsupervised (or weakly-supervised) methods also were used for NLU tasks, primarily leveraging search query click logs (Hakkani-Tur et al, 2011a,b, 2013 and knowledge graphs ; hybrid methods, for example as described in (Kim et al, 2015a;Chen et al, 2016), also exist. Our approach Table 4: Percentange of queries with timex and location slots in each of our target domains.…”

Section: Related Workmentioning

confidence: 99%

Bag of Experts Architectures for Model Reuse in Conversational Language Understanding

Jha¹,

Marin²,

Shivaprasad³

et al. 2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

View full text Add to dashboard Cite

Slot tagging, the task of detecting entities in input user utterances, is a key component of natural language understanding systems for personal digital assistants. Since each new domain requires a different set of slots, the annotation costs for labeling data for training slot tagging models increases rapidly as the number of domains grow. To tackle this, we describe Bag of Experts (BoE) architectures for model reuse for both LSTM and CRF based models. Extensive experimentation over a dataset of 10 domains drawn from data relevant to our commercial personal digital assistant shows that our BoE models outperform the baseline models with a statistically significant average margin of 5.06% in absolute F1-score when training with 2000 instances per domain, and achieve an even higher improvement of 12.16% when only 25% of the training data is used.

show abstract

“…Initially the natural language query is given to the syntax based transformation to convert the natural languages to the normal query terms and then the efficiency was calculated. The calculation shows that it improves significantly for efficient domain detection preferable in web based utterances [14]. Ryen W. White, Mikhail Bilenko and Silviu Cucerzan proposed the web search interaction mechanism by conducting the study.…”

Section: Task Clusteringmentioning

confidence: 99%

Predicting implicit search behaviors usinglog analysis

LeemaPriyadharshini¹,

Florence²,

Prema³

et al. 2018

IJET

View full text Add to dashboard Cite

Search engines provide ranked information based on the query given by the user. Understanding user search behavior is an important task for satisfaction of the users with the needed information. Understanding user search behaviors and recommending more information or more sites to the user is an emerging task. The work is based on the queries given by the user, the amount of time the user spending on the particular page, the number of clicks done by the user particular URL. These details will be available in the dataset of web search log. The web search log is nothing but the log which contains the user searching activities and other details like machine ID, browser ID, timestamp, query given by the user, URL accessed etc., four things considered as the important: 1) Extraction of tasks from the sequence of queries given by the user 2) suggesting some similar query to the user 3) ranking URLs based on the implicit user behaviors 4) increasing web page utilities based on the implicit behaviors. For increasing the web page utility and ranking the URLs predicting implicit user behavior is a needed task. For each of these four things designing and implementation of some algorithms and techniques are needed to increase the efficiency and effectiveness.

show abstract

Employing web search query click logs for multi-domain spoken language understanding

Cited by 12 publications

References 21 publications

Understanding computer-directed utterances in multi-user dialog systems

Understanding computer-directed utterances in multi-user dialog systems

Bag of Experts Architectures for Model Reuse in Conversational Language Understanding

Predicting implicit search behaviors usinglog analysis

Contact Info

Product

Resources

About