Tweet Timeline Generation with Determinantal Point Processes

The task of tweet timeline generation (TTG) aims at selecting a small set of representative tweets to generate a meaningful timeline and providing enough coverage for a given topical query. This paper presents an approach based on determinantal point processes (DPPs) by jointly modeling the topical relevance of each selected tweet and overall selectional diversity. Aiming at better treatment for balancing relevance and diversity, we introduce two novel strategies, namely spectral rescaling and topical prior. Extensive experiments on the public TREC 2014 dataset demonstrate that our proposed DPP model along with the two strategies can achieve fairly competitive results against the state-of-the-art TTG systems.

[1]  Tetsuya Sakai,et al.  TREC 2013 Temporal Summarization , 2013, TREC.

[2]  Chen Lin,et al.  Generating event storylines from microblogs , 2012, CIKM.

[3]  Ben Taskar,et al.  k-DPPs: Fixed-Size Determinantal Point Processes , 2011, ICML.

[4]  Ben Taskar,et al.  Near-Optimal MAP Inference for Determinantal Point Processes , 2012, NIPS.

[5]  Kristen Grauman,et al.  Diverse Sequential Subset Selection for Supervised Video Summarization , 2014, NIPS.

[6]  Rui Yan,et al.  Timeline generation with social attention , 2013, SIGIR.

[7]  Claire Cardie,et al.  Timeline generation: tracking individuals on twitter , 2013, WWW.

[8]  Krithi Ramamritham,et al.  Real Time Discovery of Dense Clusters in Highly Dynamic Graphs: Identifying Real World Events in Highly Dynamic Environments , 2012, Proc. VLDB Endow..

[9]  Ben Taskar,et al.  Expectation-Maximization for Learning Determinantal Point Processes , 2014, NIPS.

[10]  Dragomir R. Radev,et al.  DivRank: the interplay of prestige and diversity in information networks , 2010, KDD.

[11]  Yan Zhang,et al.  Timeline Generation through Evolutionary Trans-Temporal Summarization , 2011, EMNLP.

[12]  Ben Taskar,et al.  Learning Determinantal Point Processes , 2011, UAI.

[13]  J. Vondrák,et al.  Submodular Function Maximization via the Multilinear Relaxation and Contention Resolution Schemes , 2014 .

[14]  Wubai Zhou,et al.  Generating textual storyline to improve situation awareness in disaster management , 2014, Proceedings of the 2014 IEEE 15th International Conference on Information Reuse and Integration (IEEE IRI 2014).

[15]  Elad Yom-Tov,et al.  Updating Users about Time Critical Events , 2013, ECIR.

[16]  Chandra Chekuri,et al.  Submodular function maximization via the multilinear relaxation and contention resolution schemes , 2011, STOC '11.

[17]  Walid Magdy,et al.  QCRI at TREC 2014: Applying the KISS principle for the TTG task in the Microblog Track , 2014, TREC.

[18]  Chao Lv,et al.  PKUICST at TREC 2014 Microblog Track: Feature Extraction for Effective Microblog Search and Adaptive Clustering Algorithms for TTG , 2014, TREC.

[19]  Dimitrios Gunopulos,et al.  On burstiness-aware search for document sequences , 2009, KDD.

[20]  Ben Taskar,et al.  Discovering Diverse and Salient Threads in Document Collections , 2012, EMNLP.

[21]  Laks V. S. Lakshmanan,et al.  Incremental cluster evolution tracking from highly dynamic network data , 2014, 2014 IEEE 30th International Conference on Data Engineering.

[22]  Douglas W. Oard,et al.  HLTCOE at TREC 2014: Microblog and Clinical Decision Support , 2014, TREC.

[23]  Jasper Snoek,et al.  A Determinantal Point Process Latent Variable Model for Inhibition in Neural Spiking Data , 2013, NIPS.

[24]  Jimmy J. Lin,et al.  Overview of the TREC-2014 Microblog Track , 2014, TREC.

[25]  ChengXiang Zhai,et al.  Learn from web search logs to organize search results , 2007, SIGIR.

[26]  Tao Li,et al.  Generating Pictorial Storylines Via Minimum-Weight Connected Dominating Set Approximation in Multi-View Graphs , 2012, AAAI.