122,851 Hits in 9.5 sec

Topic Set Size Design with the Evaluation Measures for Short Text Conversation [chapter]

Tetsuya Sakai, Lifeng Shang, Zhengdong Lu, Hang Li
2015 Lecture Notes in Computer Science  
In this study, we apply the topic set size design technique of Sakai to decide on the number of test topics, using variance estimates of the above evaluation measures.  ...  for each of our evaluation measures.  ...  ∑ n j=1 (x ij −x i• ) 2 m(n − 1) . (1) Evaluation Measures for Short Text Conversation The official evaluation measures of the STC task are graded-relevance IR evaluation measures for navigational intents  ... 
doi:10.1007/978-3-319-28940-3_25 fatcat:mgvbw72u4na5laoudvv2fqj5ca

Does Size Matter? When Small is Good Enough

Anna Lisa Gentile, Amparo Elizabeth Cano Basave, Aba-Sah Dadzie, Vitaveska Lanfranchi, Neil Ireson
2011 Workshop on Making Sense of Microposts  
The results obtained show that the accuracy of topic classification for micropost-size texts is a suitable approximation of classification performed on longer texts.  ...  Our hypothesis is that based on a specific task (in this case, topic classification), results obtained using longer texts may be approximated by short texts, of micropost size, i.e., maximum length 140  ...  The task chosen for the evaluation is text classification on non-predefined topics.  ... 
dblp:conf/msm/GentileBDLI11 fatcat:rx3fnqhj6vdp7i7c3ay2t52ujy

Conversational Structure Aware and Context Sensitive Topic Model for Online Discussions [article]

Yingcheng Sun and Kenneth Loparo and Richard Kolacinski
2020 arXiv   pre-print
Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.  ...  Topic modelling is an efficient way of better understanding large text datasets at scale.  ...  This coherence measure retrieves co-occurrence counts for the given words using a context window with the window size 5.  ... 
arXiv:2002.02353v1 fatcat:jgk7ty6zv5bwfmdyrcv2r2b2pm

On Estimating Variances for Topic Set Size Design

Tetsuya Sakai, Lifeng Shang
2016 NTCIR Conference on Evaluation of Information Access Technologies  
Topic set size design is a suite of statistical techniques for determining the appropriate number of topics when constructing a new test collection.  ...  Recently, we ran an IR task at NTCIR-12 where the number of topics was actually determined using topic set size design with an initial pilot data set based on only five similar runs; a test collection  ...  Recently, we ran an IR task at NTCIR-12 (namely, the Chinese subtask of the new Short Text Conversation task [10] ) where the number of topics was actually determined using topic set size design with  ... 
dblp:conf/ntcir/SakaiS16 fatcat:jf3xsr6kzfg6rde2n6y6s6cbha

Examining LDA2Vec and Tweet Pooling for Topic Modeling on Twitter Data

Kristofferson Culmer, Jeffrey Uhlmann
2021 WSEAS Transactions on Information Science and Applications  
specifically for topic modeling on short text documents.  ...  The short lengths of tweets present a challenge for topic modeling to extend beyond what is provided explicitly from hashtag information.  ...  This behavior is consistent with the expectation that LDA2Vec should perform better with a larger document size. Biterm had the best coher ence score for the C p measure.  ... 
doi:10.37394/23209.2021.18.13 fatcat:dzk5r3lv3zc6bolj6mszbaqb4y

Structural analysis of chat messages for topic detection

Haichao Dong, Siu Cheung Hui, Yulan He
2006 Online information review (Print)  
In the proposed approach, different techniques such as sessionalization of chat messages and extraction of features from icon texts and URLs are incorporated for message pre-processing.  ...  The primary objective of chat message characterization is to understand the properties of chat messages for effective message analysis such as message topic detection.  ...  The remaining 30% was then used as the testing data set for performance evaluation of topic categorization.  ... 
doi:10.1108/14684520610706398 fatcat:5phkryqc2bdg7nw6xezeusqojy

Visual Text Analytics for Online Conversations

Enamul Hoque
2016 Companion Publication of the 21st International Conference on Intelligent User Interfaces - IUI '16 Companion  
coauthors: Chapter 2 is based on the article ConVis: A visual text analytic system for exploring blog conversations, by Enamul Hoque and Giuseppe  ...  Often many people contribute to the discussion, which become very long with hundreds of comments, making it difficult for users to get insights about the discussion.  ...  Following the design study methodology in InfoVis, we started with a user requirement analysis for the domain of blog conversations to derive a set of design principles.  ... 
doi:10.1145/2876456.2876461 dblp:conf/iui/Hoque16 fatcat:ob7o4r6v7nbx7mxeizkuovmuqa

affinity: A System for Latent User Similarity Comparison on Texting Data [article]

Tobias Eichinger and Felix Beierle and Sumsam Ullah Khan and Robin Middelanis and Veeraraghavan Sekar and Sam Tabibzadeh
2019 arXiv   pre-print
Third, assessing the quality of a similarity measure on text messaging data representing a potentially infinite set of topics is non-trivial.  ...  Second, the definition of an appropriate privacy-preserving similarity measure is non-trivial.  ...  Most users have a diverse set of people they text with on probably an even more diverse set of topics. Hence texting histories will not harbor a single topic yet rather be a mixed lot of topics.  ... 
arXiv:1904.01897v1 fatcat:jq4pjilb4fappbr7bamji2acdu

A Review on Dyadic Conversation Visualizations - Purposes, Data, Lens of Analysis [article]

Joshua Y. Kim, Rafael A. Calvo, Kalina Yacef, N.J. Enfield
2019 arXiv   pre-print
One, we summarize the current practices in the domain of visualizing dyadic conversations. Two, we provide suggestions for future dialogue visualization research.  ...  Many professional services are provided through text and voice systems, from voice calls over the internet to messaging and emails.  ...  The crisis counselor participates in a text messaging conversation with the caller. Words in the text messages each contribute towards the topic-buckets.  ... 
arXiv:1905.00653v1 fatcat:mwx27qzwarho7ibs5j7cfskqoe

The Influence of Technology Delivery Mode on Intervention Outcomes: Analysis of a Theory-Based Sexual Health Program

Nicole Levitz, Erica Wood, Leslie Kantor
2018 Journal of Medical Internet Research  
Educator confidence was significantly associated with all the topics discussed.  ...  All the 3 modalities had significant associations with educator confidence and showed similar effect sizes to those of user confidence.  ...  The authors would also like to thank Dr James Jaccard and Dr Lindsay Bornheimer for their assistance.  ... 
doi:10.2196/10398 pmid:30158100 fatcat:vhh5tnmtbjepbbwublpl2hmclq

Topical clustering of search results

Ugo Scaiella, Paolo Ferragina, Andrea Marino, Massimiliano Ciaramita
2012 Proceedings of the fifth ACM international conference on Web search and data mining - WSDM '12  
We test several standard measures for evaluating the performance of all systems and show a relative improvement of up to 20%.  ...  clusters with meaningful phrases describing the topics of the results included in them.  ...  The authors thank professors F. Romani and G. Del Corso for insightful discussions about Laplacian matrices.  ... 
doi:10.1145/2124295.2124324 dblp:conf/wsdm/ScaiellaFMC12 fatcat:eql7et76djc73kul6cylwv7pke

ConCET: Entity-Aware Topic Classification for Open-Domain Conversational Agents [article]

Ali Ahmadvand, Harshita Sahijwani, Jason Ingyu Choi, Eugene Agichtein
2020 arXiv   pre-print
methods; second, we evaluate ConCET on a large dataset of human-machine conversations with real users, collected as part of the Amazon Alexa Prize.  ...  Identifying the topic (domain) of each user's utterance in open-domain conversational systems is a crucial step for all subsequent language understanding and response tasks.  ...  We gratefully acknowledge the financial and computing support from the Amazon Alexa Prize 2018.  ... 
arXiv:2005.13798v1 fatcat:y3kquj4kzva6tdkkerxadjodxq

Topic identification based extrinsic evaluation of summarization techniques applied to conversational speech

David Harwath, Timothy J. Hazen
2012 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We show that these results appear to be correlated with the performance of an automated topic identification system, and argue that this automated system can act as a low-cost proxy for a human evaluation  ...  In this paper, we use topic identification as a proxy for relevancy determination in the context of an information retrieval task, and a summary is deemed effective if it enables a user to determine the  ...  ACKNOWLEDGEMENTS The authors would like to thank Jim Glass and Jackie Lee of MIT CSAIL for their assistance in conducting the Mechanical Turk experiments.  ... 
doi:10.1109/icassp.2012.6289061 dblp:conf/icassp/HarwathH12 fatcat:2kxqlaznszdf7mxc62g6v4tbma

Challenges of Building an Intelligent Chatbot

Anna Chizhik, Yulia Zherebtsova
2020 International Conference "Internet and Modern Society"  
As systems designed for personalized interaction with users, conversational chatbots are becoming increasingly sophisticated in an attempt to mimic human dialogue.  ...  With continued growth in messaging applications and increasing demand for machine-based communications, conversational chatbot is likely to play a large part in companies' customer experience strategy.  ...  Introduction Intelligent dialogue agents are designed to conduct a coherent and emotionally engaging conversation with users.  ... 
dblp:conf/ims2/ChizhikZ20 fatcat:tptdmej6rrd4hp3s4rf2qsdcmq

Topic-aware Pointer-Generator Networks for Summarizing Spoken Conversations [article]

Zhengyuan Liu, Angela Ng, Sheldon Lee, Ai Ti Aw, Nancy F. Chen
2019 arXiv   pre-print
To the best of our knowledge, to date no one has adopted it for summarizing conversations.  ...  Due to the lack of publicly available resources, conversation summarization has received far less attention than text summarization.  ...  The task is to learn a function f with a parameter set θ that maximizes the probability to generate readable and meaningful output text.  ... 
arXiv:1910.01335v1 fatcat:wqnnux2oj5a7bo7buvzuil4q4q
« Previous Showing results 1 — 15 out of 122,851 results