95,385 Hits in 4.1 sec

Evaluating the Effectiveness of Keyword Search

William Webber
2010 IEEE Data Engineering Bulletin  
We compare practices with the longer-standing full-text evaluation methodologies in information retrieval.  ...  As a result, the empirical evaluation of a keyword retrieval system's effectiveness is essential.  ...  The most fruitful of the similarity metrics combined both a structural and a full-text component [9] .  ... 
dblp:journals/debu/Webber10 fatcat:xva7qp4szzcqpm7fprixanbcji

Emotion Embedding Spaces for Matching Music to Stories

Minz Won, Justin Salamon, Nicholas J. Bryan, Gautham Mysore, Xavier Serra
2021 Zenodo  
We formalize this task as a cross-modal text-to-music retrieval problem.  ...  ., Word2Vec and metric learning) to bridge this gap.  ...  Metric learning is also used for bridging the modality gap between text and audio, such as keyword spotting [42] , text-based audio retrieval [43, 44] , and tag-based music retrieval [35, 38] in both  ... 
doi:10.5281/zenodo.5624482 fatcat:uqlm3s5korb5rm2ybkbvr42qpi

A Brief Survey of Question Answering Systems

Michael Caballero
2021 International Journal of Artificial Intelligence & Applications  
Lastly, some common datasets in this field are introduced and different evaluation metrics are discussed.  ...  After discussing QA history, this paper summarizes the different approaches to the architecture of QA systems -- whether they are closed or open-domain and whether they are text-based, knowledge-based,  ...  The next evaluation metric, the Mean Reciprocal Rank (MRR), is one of the best for measuring the performance of QA systems.  ... 
doi:10.5121/ijaia.2021.12501 fatcat:jhydawao5vgaxoghxrp6v37f2q

Combination of Textual and Structural Context for Retrieving Multimedia Elements [chapter]

Sana Fakhfakh, Mohamed Tmar, Walid Mahdi
2014 Advances in Intelligent Systems and Computing  
for multimedia retrieval.  ...  Using a geometric metric, we can represent structural information in XML document with a vector for each element. Experimental evaluation is carried out using the INEX 2007, ImageCLEF 2010 and 2014.  ...  We evaluate our system into three databases extracted from three collections : INEX 2007 (Initiative for the Evaluation of XML Retrieval) Ad Hoc task, ImageCLEF 2010 Wikipedia image retrieval task and  ... 
doi:10.1007/978-3-319-07650-8_9 fatcat:ylsitg7rfvh2tjktjb5ca64fke

Analytical Classification of Multimedia Index Structures by Using a Partitioning Method-Based Framework [article]

Mohammadreza keyvanpour Department of Computer Engineering, Alzahra University, Tehran, Iran, Department of Computer Engineering, Islamic Azad University, Qazvin Branch, Qazvin, Iran)
2011 arXiv   pre-print
Due to variety and plenty of proposed index structures, we suggest a systematic framework based on partitioning method used in these structures to classify multimedia index structures, and then we evaluated  ...  Lots of research on this field has led to proposition of different kinds of index structures to support fast and efficient similarity search to retrieve multimedia data from these databases.  ...  In past years researchers have tried extending retrieval techniques used in text retrieval to the area of multimedia retrieval.  ... 
arXiv:1105.1948v1 fatcat:hofkrcz3o5dp7efaua5242dpoq

Table Caption Generation in Scholarly Documents Leveraging Pre-trained Language Models [article]

Junjie H. Xu, Kohei Shinden, Makoto P. Kato
2021 arXiv   pre-print
Our experimental results showed that T5 is the better generation model for this task, as it outperformed GPT-2 in BLEU and METEOR implying that the generated text are clearer and more precise.  ...  of caption generation methods for scholarly documents with different strategies to retrieve relevant sentences from the paper body.  ...  These three metrics are widely used for evaluating models that generate text. Table 2 shows the evaluation results in terms of ROUGE, BLEU and METEOR.  ... 
arXiv:2108.08111v1 fatcat:nq2pttn5ozgstafqdhaiurvgny

Does Structure Matter? Leveraging Data-to-Text Generation for Answering Complex Information Needs [article]

Hanane Djeddal, Thomas Gerald, Laure Soulier, Karen Pinel-Sauvagnat, Lynda Tamine
2021 arXiv   pre-print
We evaluate both the generated answer and its corresponding structure and show the effectiveness of planning-based models in comparison to a text-to-text model.  ...  The experimental evaluation is performed using the TREC Complex Answer Retrieval (CAR) dataset.  ...  Acknowledgement We would like to thank projects ANR JCJC SESAMS (ANR-18-CE23-0001) and ANR COST (ANR-18-CE23-0016) for supporting this work.  ... 
arXiv:2112.04344v1 fatcat:cxscuzygtrd7hkssb4hdibo7wm

Hybrid XML Retrieval: Combining Information Retrieval and a Native XML Database

Jovan Pehcevski, James A. Thom, Anne-Marie Vercoustre
2005 Information retrieval (Boston)  
and different choices of evaluation metrics) -- the XML retrieval systems exhibit varying behaviour and the best performance can be reached for different values of the retrieval parameters.  ...  This paper investigates the impact of three approaches to XML retrieval: using Zettair, a full-text information retrieval system; using eXist, a native XML database; and using a hybrid system that takes  ...  Acknowledgements We thank Wolfgang Meier for his assistance with using eXist, and Nick Lester, Falk Scholer and other members of the Search Engine Group at RMIT for their support with using Zettair.  ... 
doi:10.1007/s10791-005-0748-1 fatcat:x4rwxiu3rndnrlozun2q6rcptm

Open Domain Question Answering with A Unified Knowledge Interface [article]

Kaixin Ma, Hao Cheng, Xiaodong Liu, Eric Nyberg, Jianfeng Gao
2022 arXiv   pre-print
In this work, we bridge this gap and use the data-to-text method as a means for encoding structured knowledge for ODQA.  ...  Specifically, we propose a verbalizer-retriever-reader framework for ODQA over data and text where verbalized tables from Wikipedia and graphs from Wikidata are used as augmented knowledge sources.  ...  Acknowledgements We would like to thank Ruohong Zhang for helpful discussions and anonymous reviewers for their valuable suggestions on this paper.  ... 
arXiv:2110.08417v2 fatcat:qizlkm5ogfgkvk65zyvmm2dg7a

Revisiting Cross Modal Retrieval [article]

Shah Nawaz, Muhammad Kamran Janjua, Alessandro Calefati, Ignazio Gallo
2018 arXiv   pre-print
In our knowledge, this work is the first of its kind in terms of employing a single network and fused image-text embedding for cross-modal retrieval.  ...  However, in our work image-text encoding can achieve comparable results in terms of cross-modal retrieval without having to use a separate network for each modality.  ...  Evaluation Metric For the evaluation of cross modal retrieval, we use same metric as described in [51] i.e. R@K (read as Recall at K).  ... 
arXiv:1807.07364v1 fatcat:l4pp5cq6f5e77gnwd3tcbggnne

Investigating the document structure as a source of evidence for multimedia fragment retrieval

Mouna Torjmen-Khemakhem, Karen Pinel-Sauvagnat, Mohand Boughanem
2013 Information Processing & Management  
Multimedia objects can be retrieved using their context that can be for instance the text surrounding them in documents. This text may be either near or far from the searched objects.  ...  Results showed that structural evidences are of high interest to tune the importance of textual context for multimedia retrieval.  ...  For both steps, we will study the impact of text proximity for relevance evaluation thanks to the underlying structural information.  ... 
doi:10.1016/j.ipm.2013.06.001 fatcat:ac4vtvzcorbu7mi33kwuv2wkla

Variational Recurrent Sequence-to-Sequence Retrieval for Stepwise Illustration [chapter]

Vishwash Batra, Aparajita Haldar, Yulan He, Hakan Ferhatosmanoglu, George Vogiatzis, Tanaya Guha
2020 Lecture Notes in Computer Science  
We evaluate the model for the application of stepwise illustration of recipes, where a sequence of relevant images are retrieved to best match the steps described in the text.  ...  This synthetic image embedding point associated with every text embedding point can then be employed for either image generation or image retrieval as desired.  ...  -We propose a new variational recurrent seq2seq (VRSS) retrieval model for seq2seq retrieval, which employs temporally-dependent latent variables to capture the sequential semantic structure of text-image  ... 
doi:10.1007/978-3-030-45439-5_4 fatcat:bjd23a7kfnednokmphtjpg6ttm

Do Cross Modal Systems Leverage Semantic Relationships?

Shah Nawaz, M. Kamran Janjua, Ignazio Gallo, Arif Mahmood, Alessandro Calefati, Faisal Shafait
2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
We also propose a novel cross modal retrieval system using a single stream network for bidirectional retrieval.  ...  Current cross modal retrieval systems are evaluated using R@K measure which does not leverage semantic relationships rather strictly follows the manually marked image text query pairs.  ...  Re-evaluating R@K Measure Consider a cross modal retrieval scenario with a pairwise loss function L c (x i ,x ± i )=d(x i ,x ± i ); where d(x i ,x ± i ) is a distance metric.  ... 
doi:10.1109/iccvw.2019.00551 dblp:conf/iccvw/NawazJGMCS19 fatcat:2dtr55xyxnerbiybkaqfxmwy64

Do Cross Modal Systems Leverage Semantic Relationships? [article]

Shah Nawaz, Muhammad Kamran Janjua, Ignazio Gallo, Arif Mahmood, Alessandro Calefati, Faisal Shafait
2019 arXiv   pre-print
We also propose a novel cross-modal retrieval system using a single stream network for bidirectional retrieval.  ...  Current cross-modal retrieval systems are evaluated using R@K measure which does not leverage semantic relationships rather strictly follows the manually marked image text query pairs.  ...  Re-evaluating R@K Measure Consider a cross modal retrieval scenario with a pairwise loss function L c (x i , x ± i ) = d(x i , x ± i ); where d(x i , x ± i ) is a distance metric.  ... 
arXiv:1909.01976v1 fatcat:vllxjbtswnfo5lhpdu7s2guhea

Ad-hoc object retrieval in the web of data

Jeffrey Pound, Peter Mika, Hugo Zaragoza
2010 Proceedings of the 19th international conference on World wide web - WWW '10  
We connect this task to the traditional ad-hoc document retrieval and discuss appropriate evaluation metrics.  ...  Here we propose a formal model of one specific semantic search task: ad-hoc object retrieval.  ...  This is a result of our simple baseline metric used for retrieval. Our results, however, do contain enough diversity in relevance to conduct an evaluation for distinguishing ranking algorithms.  ... 
doi:10.1145/1772690.1772769 dblp:conf/www/PoundMZ10 fatcat:jzt6omdgxffdtexuxnscxtxxoi
« Previous Showing results 1 — 15 out of 95,385 results