Filters








136 Hits in 2.6 sec

Examining the Coherence of the Top Ranked Tweet Topics

Anjie Fang, Craig Macdonald, Iadh Ounis, Philip Habel
2016 Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval - SIGIR '16  
Topic modelling approaches help scholars to examine the topics discussed in a corpus.  ...  Both of these methods demonstrate a higher performance in producing more interpretable topics than the standard Latent Dirichlet Allocation (LDA) when applied on tweets.  ...  SIGIR ' 16 , 16 July 17-21, 2016, Pisa, Italy c 2016 ACM.  ... 
doi:10.1145/2911451.2914731 dblp:conf/sigir/FangMOH16 fatcat:noghsitbknda3llrfln7ty4grq

Using Word Embedding to Evaluate the Coherence of Topics from Twitter Data

Anjie Fang, Craig Macdonald, Iadh Ounis, Philip Habel
2016 Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval - SIGIR '16  
Scholars often seek to understand topics discussed on Twitter using topic modelling approaches.  ...  We use pairwise comparisons for the three topic modelling approaches, specifically: LDA vs. TLDA, LDA vs. PAM and TLDA vs. PAM.  ...  Turning to Word Embedding approaches, recently scholars have applied Feed-Forward Neural Network (FFNN) for Word Embeddings [1] .  ... 
doi:10.1145/2911451.2914729 dblp:conf/sigir/FangMOH16a fatcat:hhngfqqdpbdnzdl6gaotbjtkiy

CiteSight

Avishay Livne, Vivek Gokuladas, Jaime Teevan, Susan T. Dumais, Eytan Adar
2014 Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval - SIGIR '14  
Google Scholar provides a service that recommends newly published articles to users based on their publication history.  ...  For example, ActiveCite [34] provides recommendations by forwarding searches to Google Scholar and integrating the results.  ... 
doi:10.1145/2600428.2609585 dblp:conf/sigir/LivneGTDA14 fatcat:sbsduglnifc75mn27mhst6tdxq

Evaluation of a simple and effective music information retrieval method

Stephen Downie, Michael Nelson
2000 Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '00  
Scholars and musicians have consulted these printed volumes for over a thousand years (Brook [2] ).  ...  That QLEN plays the most influential role in retrieval performance when compared to CLASS and NLEN is evident in the 0.49 (QLEN) vs. 0.17 (NLEN) vs. 0.03 (CLASS) differences between extremes.  ... 
doi:10.1145/345508.345551 dblp:conf/sigir/DownieN00 fatcat:fj32y3vqgjfm3k4wnrxdz4urzu

Evaluating aggregated search pages

Ke Zhou, Ronan Cummins, Mounia Lalmas, Joemon M. Jose
2012 Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrieval - SIGIR '12  
SIGIR '12, August 12-16, 2012 , Portland, Oregon, USA.  ...  We use the F-measure (VS), Mean Precision (IS), and Kendall-tau correlation (RP).  ... 
doi:10.1145/2348283.2348302 dblp:conf/sigir/ZhouCLJ12 fatcat:b6yv5tsoongh5mnelokcfp3jpe

Multitasking, cognitive coordination and cognitive shifts during web searching

Jia Tina Du
2011 SIGIR Forum  
Cognitive State Shift (this study) vs.  ...  Cognitive State Shift (this study) vs. Information Problem Shift ).  ... 
doi:10.1145/1988852.1988869 fatcat:h2py4yu42naujlpzxzq23hagbu

Query representation for cross-temporal information retrieval

Miles Efron
2013 Proceedings of the 36th international ACM SIGIR conference on Research and development in information retrieval - SIGIR '13  
First, it has no ready probabilistic interpretation that would help us integrate it into CTIR. Also, in CTIR, some operations intuitively merit different costs.  ...  Aside from the obvious difference in focus (CTIR vs. CLIR) and the attendant differences in evidence, our approach differs from He's mathematically and algorithmically.  ... 
doi:10.1145/2484028.2484054 dblp:conf/sigir/Efron13 fatcat:thehxpjijvc3tc2n4x54rfserq

An advanced full-text retrieval and analysis system

J. Smith, S. Weiss, G. Ferguson
1987 Proceedings of the 10th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '87  
The first is the sophisticated, often esoteric, needs of academic professionals, ranging from literary scholars and anthropologists to chemists and computer scientists.  ...  The Command Processor interprets the FLANGE command to produce a sequence of calls to the Analytic Engine.  ... 
doi:10.1145/42005.42027 dblp:conf/sigir/SmithWF87 fatcat:52r432glqval7dgxot7ukreftm

Context & Semantics in News & Web Search

Daan Odijk
2016 SIGIR Forum  
The Table 7 .3: Retrieval effectiveness of the dynamic query modeling (DQM) approach vs the two baselines.  ...  Above, in Section 2.2.1, we detailed how humanities scholars study news archives.  ... 
doi:10.1145/2964797.2964816 fatcat:zmt2clavxrdi5izyo73z6kfg3q

Hierarchical Random Walk Inference in Knowledge Graphs

Qiao Liu, Liuyi Jiang, Minghao Han, Yao Liu, Zhiguang Qin
2016 Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval - SIGIR '16  
The merits of the PRA algorithm are that its inference results are easily interpretable, and it is inherently parallelizable.  ...  PRA), while after incorporating the local inference results into the inference model, the Hits@10 score was improved again by 54.15% (HiRi vs. Baseline).  ... 
doi:10.1145/2911451.2911509 dblp:conf/sigir/LiuJHLQ16 fatcat:7ku6rx5qpje55eeivqx3aictjq

The voting model for people search

Craig Macdonald
2009 SIGIR Forum  
This has recently been resolved by the LETOR dataset, created for the SIGIR series of workshops on Learning to Rank for IR .  ...  However, in general, PL2F exhibited a lower performance than BM25F, (similar to PL2 vs BM25 in Chapter 6).  ... 
doi:10.1145/1670598.1670616 fatcat:loqagreqk5h27n5owrkoxmak4q

Effective and practical neural ranking

Sean MacAvaney
2021 SIGIR Forum  
For instance, the addition no (e.g., fracture vs. no fracture) could change the meaning of the report considerably.  ...  Furthermore, the design of EPIC is interpretable, transparently assigning importance scores to query and document terms.  ... 
doi:10.1145/3476415.3476432 fatcat:fdjy53sggvhgxo5fa5hzpede2i

Aggregated feature video retrieval for MPEG-7 via clustering

Jiamin Ye
2005 SIGIR Forum  
jim, sound, fine, thing, bacterium, commission, plateau, telecharg, busi, castro, leonard, secretan, ken, mercer, citi, fingertip, pope, maine, freight, jame, corpor, steven, work, gentleman, depress, scholar  ...  Sys2 vs. Sys4 Sys2 is the baseline system for Sys4. Both systems use the same aggregated index which is derived from content-based features and ASR output texts.  ... 
doi:10.1145/1067268.1067298 fatcat:b7v4top2pbeu7bjup6uliswpw4

Information extraction from digital social trace data with applications to social media and scholarly communication data

Shubhanshu Mishra
2020 SIGIR Forum  
Figure 6 . 4 : 64 Ratio of user meta-data features vs. sentiment positive. ) Ratio of user-level numerical meta-data Figure 6 . 5 : 65 Meta-data features vs. sentiment classes using recent 200 tweets  ...  However, existing IE systems, if not focused on these aspects, cause a difficulty in interpreting the extracted information.  ... 
doi:10.1145/3451964.3451981 fatcat:36djwlckprhl5hymzhivrbnscu

Health Misinformation in Search and Social Media

Amira Ghenai
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
This work is similar to ours in Chapter 5, as they are both controlled studies in which participants' behavior is interpreted by a set of log files.  ...  Although correlation was found in epidemiological data, some scholars argue that an increase in awareness due the hype of the "vaccination crisis" might have caused the increase in cases.  ... 
doi:10.1145/3077136.3084153 dblp:conf/sigir/Ghenai17 fatcat:na5rjxfgw5e6togqpdkgwmk4se
« Previous Showing results 1 — 15 out of 136 results