880 Hits in 5.8 sec

Bad Company - Neighborhoods in Neural Embedding Spaces Considered Harmful

Johannes Hellrich, Udo Hahn
2016 International Conference on Computational Linguistics  
The overall low reliability we observe, nevertheless, casts doubt on the suitability of word neighborhoods in embedding spaces as a basis for qualitative conclusions on synchronic and diachronic lexico-semantic  ...  We assess the reliability and accuracy of (neural) word embeddings for both modern and historical English and German.  ...  neighborhoods in embedding space for preselected words (Jo, 2016) .  ... 
dblp:conf/coling/HellrichH16 fatcat:x5aoeihwpnb7xnxxppooaehva4


2014 International Journal of Computer Science and Informatics  
Due to the rapid increase of electronic mail (or e-mail), several people and companies found it an easy way to distribute a massive amount of undesired messages to a tremendous number of users at a very  ...  Unethical e-mail senders bear little or no cost for mass distribution of messages, yet normal e-mail users are forced to spend time and effort in reading undesirable messages from their mailboxes.  ...  , the Bayesian approach considers all the evidence in the email, both good and bad.  ... 
doi:10.47893/ijcsi.2014.1164 fatcat:qjo2p6kqlrgz7j6w2tf7omo3we

Focus on what matters: Applying Discourse Coherence Theory to Cross Document Coreference [article]

William Held, Dan Iter, Dan Jurafsky
2021 arXiv   pre-print
We model the entities/events in a reader's focus as a neighborhood within a learned latent embedding space which minimizes the distance between mentions and the centroids of their gold coreference clusters  ...  Existing approaches simplify by considering coreference only within document clusters, but this fails to handle inter-cluster coreference, common in many applications.  ...  Training Pair Generation We use K nearest neighbors in the bi-encoder embedding space to generate training data for the pairwise classifier.  ... 
arXiv:2110.05362v1 fatcat:oc5o6e3j7ncf3jwnuc4fio3424

Content-based Influence Modeling for Opinion Behavior Prediction

Chengyao Chen, Zhitao Wang, Yu Lei, Wenjie Li
2016 International Conference on Computational Linguistics  
Nowadays, social media has become a popular platform for companies to understand their customers.  ...  In the experiments conducted on Twitter datasets, our model significantly outperforms other popular opinion formation models.  ...  Acknowledgments The work described in this paper was supported by Research Grants Council of Hong Kong (PolyU 5202/12E, PolyU 152094/14E), National Natural Science Foundation of China (61272291 and 61672445  ... 
dblp:conf/coling/ChenWLL16 fatcat:3vqzmcc2ona3zl22hvefi3epnm

Named Entity Recognition for Hungarian Using Various Machine Learning Algorithms

Richárd Farkas, György Szarvas, András Kocsor
2006 Acta Cybernetica  
In this paper we introduce a statistical Named Entity recognizer (NER) system for the Hungarian language.  ...  We examined three methods for identifying and disambiguating proper nouns (Artificial Neural Network, Support Vector Machine, C4.5 Decision Tree), their combinations and the effects of dimensionality reduction  ...  Since C4.5 considers attribute vectors as points in an ndimensional space, using continuous sample attributes naturally makes sense.  ... 
dblp:journals/actaC/FarkasSK06 fatcat:wo4usonctjb5nbny7y6nocfekm

From Symbols to Embeddings: A Tale of Two Representations in Computational Social Science

Huimin Chen, Cheng Yang, Xuanming Zhang, Zhiyuan Liu, Maosong Sun, Jianbin Jin
2021 Journal of Social Computing  
To explore the answer, we give a thorough review of data representations in CSS for both text and network.  ...  From the statistics of these applications, we unearth the strength of each kind of representations and discover the tendency that embedding-based representations are emerging and obtaining increasing attention  ...  The similarity task can be tackled by measuring the distance or similarity of neural-based sentence representations in embedding space.  ... 
doi:10.23919/jsc.2021.0011 fatcat:sczl7racpng75agdrddgnrex3q

On the Safety of Machine Learning: Cyber-Physical Systems, Decision Sciences, and Data Products [article]

Kush R. Varshney, Homa Alemzadeh
2017 arXiv   pre-print
In this paper, we do so by defining machine learning safety in terms of risk, epistemic uncertainty, and the harm incurred by unwanted outcomes.  ...  We then discuss example techniques that can be adopted in each category, such as considering interpretability and causality of predictive models, objective functions beyond expected prediction accuracy  ...  Modern techniques such as extreme gradient boosting and deep neural networks may exploit these biases and achieve high accuracy, but they may fail in making safe predictions due to unknown shifts in the  ... 
arXiv:1610.01256v2 fatcat:gjekxecrnze4hgbylmogjv4b3y

Exploring Echo-Systems: How Algorithms Shape Immersive Media Environments

2018 Journal of Media Literacy Education  
The algorithm systems are embedded programs that analyze past user data and search history in combination with other users' searches and history to calculate digital outcomes, anticipate possible recommendations  ...  This paper analyzes how the algorithm itself should be considered an immersive media environment that permits users to consume unique media feeds that may affect civic actions.  ...  YouTube CEO Susan Wojcicki explained in a formal YouTube blog post that "bad actors are exploiting our openness to mislead, manipulate, harass or even harm" (Wojcicki, 2017) .  ... 
doi:10.23860/jmle-2018-10-2-8 fatcat:c2lq6vfmzjhhzmvogxfmkqodfi

From Symbols to Embeddings: A Tale of Two Representations in Computational Social Science [article]

Huimin Chen, Cheng Yang, Xuanming Zhang, Zhiyuan Liu, Maosong Sun, Jianbin Jin
2021 arXiv   pre-print
To explore the answer, we give a thorough review of data representations in CSS for both text and network.  ...  However, these large-scale and multi-modal data also present researchers with a great challenge: how to represent data effectively to mine the meanings we want in CSS?  ...  The similarity task can be tackled by measuring the distance or similarity of neural-based sentence representations in embedding space.  ... 
arXiv:2106.14198v1 fatcat:dvy5awnfuvbnnkzusjl5wbhfki

Optimization Modeling and Empirical Research on Gasoline Octane Loss Based on Data Analysis

Ji Guo, Yujia Lou, Wanyi Wang, Xianhua Wu, Chi-Hua Chen
2021 Journal of Advanced Transportation  
Gasoline is one of the most consumed light petroleum products in transportation and other industries.  ...  Firstly, the data are screened and the high-dimensional data are reduced to construct the neural network prediction model optimized by genetic algorithm.  ...  a very small local neighborhood in Euclidean space, and that a certain point can be represented by linear least squares of its surrounding points.  ... 
doi:10.1155/2021/5553069 fatcat:tfghzb2475c7zfs76hsx7sceeq

A Wind of Change: Detecting and Evaluating Lexical Semantic Change across Times and Domains

Dominik Schlechtweg, Anna Hätty, Marco Del Tredici, Sabine Schulte im Walde
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
We perform an interdisciplinary large-scale evaluation for detecting lexical semantic divergences in a diachronic and in a synchronic task: semantic sense changes across time, and semantic sense changes  ...  In addition, we demonstrate that the same evaluation task and modelling approaches can successfully be utilised for the synchronic detection of domain-specific sense divergences in the field of term extraction  ...  Bad company-Neighborhoods in neural embedding spaces considered harmful. In Proceedings of the International Conference on Computational Linguistics 2016, pages 2785-2796, Osaka, Japan.  ... 
doi:10.18653/v1/p19-1072 dblp:conf/acl/SchlechtwegHTW19 fatcat:6t4umurbjvgo5ckdnvhcslgvyq

Transfer Meets Hybrid: A Synthetic Approach for Cross-Domain Collaborative Filtering with Text [article]

Guangneng Hu, Yu Zhang, Qiang Yang
2019 arXiv   pre-print
We propose a novel neural model to smoothly enable Transfer Meeting Hybrid (TMH) methods for cross-domain recommendation with unstructured text in an end-to-end manner.  ...  In real-world life, no single service can satisfy a user's all information needs. Thus it motivates us to exploit both auxiliary and source information for RSs in this paper.  ...  This may have a risk in transferring the noise and harm the performance, as pointed out in its sparse variant [21] .  ... 
arXiv:1901.07199v1 fatcat:ti7l7rv2vzca7cauwh4iidaceq

Topic Compositional Neural Language Model [article]

Wenlin Wang, Zhe Gan, Wenqi Wang, Dinghan Shen, Jiaji Huang, Wei Ping, Sanjeev Satheesh, Lawrence Carin
2018 arXiv   pre-print
We propose a Topic Compositional Neural Language Model (TCNLM), a novel method designed to simultaneously capture both the global semantic meaning and the local word ordering structure in a document.  ...  In order to train the MoE model efficiently, a matrix factorization method is applied, by extending each weight matrix of the RNN to be an ensemble of topic-dependent weight matrices.  ...  negative • the movie reinforces my token bad ratings -it 's the worst movie i have ever seen . • it was pretty bad , but aside from a show to the 2 idiots in their cast members , i 'm psychotic. • we had  ... 
arXiv:1712.09783v3 fatcat:4kglybxj7nafnauk3gp26uf5hu

Artificial Intelligence and Community Well-being: A Proposal for an Emerging Area of Research

Laura Musikanski, Bogdana Rakova, James Bradbury, Rhonda Phillips, Margaret Manson
2020 International Journal of Community Well-Being  
Three components of this research we propose are (1) the development and use of well-being metrics to measure the impacts of AI; (2) the use of community-based approaches in the development of AI; and  ...  No. research involving human or animal participants was involved in the formation of this essay. All relevant ethical standards were observed.  ...  Imagined and real uses by communities of cross-media AI include community asset mapping and building, fostering cohesion among neighborhood residents through use and control of public spaces, and engagement  ... 
doi:10.1007/s42413-019-00054-6 fatcat:ohvneqw24rgnfjpkqg3bnztmxa

Flexible and Context-Specific AI Explainability: A Multidisciplinary Approach [article]

Valérie Beaudouin, David Bounie (IP Paris, ECOGE, SES), Stéphan Clémençon, Florence d'Alché-Buc, James Eagan, Jayneel Parekh
2020 arXiv   pre-print
The recent enthusiasm for artificial intelligence (AI) is due principally to advances in deep learning.  ...  We propose three logical steps: First, define the main contextual factors, such as who the audience of the explanation is, the operational context, the level of harm that the system could cause, and the  ...  -and assessing the severity of harm for each possible bad event, and the probability of each bad event occurring.  ... 
arXiv:2003.07703v1 fatcat:knsvo6eftzf2fe3eeoekck5xaa
« Previous Showing results 1 — 15 out of 880 results