scholarly journals Curriculum Learning for Natural Answer Generation

2018 ◽  
Author(s):  
Cao Liu ◽  
Shizhu He ◽  
Kang Liu ◽  
Jun Zhao
Keyword(s):  
Full Advantage ◽  
Real World ◽  
Large Scale ◽  
Question Answering ◽  
Generative Models ◽  
The Internet ◽  
Basic Model ◽  
The Arts ◽  
Learning Data

By reason of being able to obtain natural language responses, natural answers are more favored in real-world Question Answering (QA) systems. Generative models learn to automatically generate natural answers from large-scale question answer pairs (QA-pairs). However, they are suffering from the uncontrollable and uneven quality of QA-pairs crawled from the Internet. To address this problem, we propose a curriculum learning based framework for natural answer generation (CL-NAG), which is able to take full advantage of the valuable learning data from a noisy and uneven-quality corpus. Specifically, we employ two practical measures to automatically measure the quality (complexity) of QA-pairs. Based on the measurements, CL-NAG firstly utilizes simple and low-quality QA-pairs to learn a basic model, and then gradually learns to produce better answers with richer contents and more complete syntaxes based on more complex and higher-quality QA-pairs. In this way, all valuable information in the noisy and uneven-quality corpus could be fully exploited. Experiments demonstrate that CL-NAG outperforms the state-of-the-arts, which increases 6.8% and 8.7% in the accuracy for simple and complex questions, respectively.

scholarly journals A Self-Aggregated Hierarchical Topic Model for Short Texts

2021 ◽  
Author(s):  
Yue Niu ◽  
Hongjie Zhang
Keyword(s):  
Real World ◽  
Semantic Information ◽  
Topic Model ◽  
Topic Models ◽  
The Internet ◽  
Ontology Learning ◽  
Learning Methods ◽  
Structured Information ◽  
Topic Hierarchy

With the growth of the internet, short texts such as tweets from Twitter, news titles from the RSS, or comments from Amazon have become very prevalent. Many tasks need to retrieve information hidden from the content of short texts. So ontology learning methods are proposed for retrieving structured information. Topic hierarchy is a typical ontology that consists of concepts and taxonomy relations between concepts. Current hierarchical topic models are not specially designed for short texts. These methods use word co-occurrence to construct concepts and general-special word relations to construct taxonomy topics. But in short texts, word cooccurrence is sparse and lacking general-special word relations. To overcome this two problems and provide an interpretable result, we designed a hierarchical topic model which aggregates short texts into long documents and constructing topics and relations. Because long documents add additional semantic information, our model can avoid the sparsity of word cooccurrence. In experiments, we measured the quality of concepts by topic coherence metric on four real-world short texts corpus. The result showed that our topic hierarchy is more interpretable than other methods.

Dynamic user modeling for expert recommendation in community question answering

2020 ◽  
Vol 39 (5) ◽  
pp. 7281-7292
Author(s):  
Tongze He ◽  
Caili Guo ◽  
Yunfei Chu ◽  
Yang Yang ◽  
Yanjun Wang
Keyword(s):  
Real World ◽  
Large Scale ◽  
Question Answering ◽  
Dynamic Change ◽  
Superior Performance ◽  
User Interest ◽  
Short Term ◽  
Expert Recommendation ◽  
Community Question Answering ◽  
User Expertise

Community Question Answering (CQA) websites has become an important channel for people to acquire knowledge. In CQA, one key issue is to recommend users with high expertise and willingness to answer the given questions, i.e., expert recommendation. However, a lot of existing methods consider the expert recommendation problem in a static context, ignoring that the real-world CQA websites are dynamic, with users’ interest and expertise changing over time. Although some methods that utilize time information have been proposed, their performance improvement can be limited due to fact that they fail they fail to consider the dynamic change of both user interests and expertise. To solve these problems, we propose a deep learning based framework for expert recommendation to exploit user interest and expertise in a dynamic environment. For user interest, we leverage Long Short-Term Memory (LSTM) to model user’s short-term interest so as to capture the dynamic change of users’ interests. For user expertise, we design user expertise network, which leverages feedback on users’ historical behavior to estimate their expertise on new question. We propose two methods in user expertise network according to whether the dynamic property of expertise is considered. The experimental results on a large-scale dataset from a real-world CQA site demonstrate the superior performance of our method.

Automatically Measuring Question Authenticity in Real-World Classrooms

2018 ◽  
Vol 47 (7) ◽  
pp. 451-464 ◽  
Author(s):  
Sean Kelly ◽  
Andrew M. Olney ◽  
Patrick Donnelly ◽  
Martin Nystrand ◽  
Sidney K. D’Mello
Keyword(s):  
Language Processing ◽  
Real World ◽  
Large Scale ◽  
Teaching Effectiveness ◽  
Teacher Questions ◽  
Data Collection And Analysis ◽  
Audio Recordings ◽  
Valuable Complement ◽  
Archival Database

Analyzing the quality of classroom talk is central to educational research and improvement efforts. In particular, the presence of authentic teacher questions, where answers are not predetermined by the teacher, helps constitute and serves as a marker of productive classroom discourse. Further, authentic questions can be cultivated to improve teaching effectiveness and consequently student achievement. Unfortunately, current methods to measure question authenticity do not scale because they rely on human observations or coding of teacher discourse. To address this challenge, we set out to use automatic speech recognition, natural language processing, and machine learning to train computers to detect authentic questions in real-world classrooms automatically. Our methods were iteratively refined using classroom audio and human-coded observational data from two sources: (a) a large archival database of text transcripts of 451 observations from 112 classrooms; and (b) a newly collected sample of 132 high-quality audio recordings from 27 classrooms, obtained under technical constraints that anticipate large-scale automated data collection and analysis. Correlations between human-coded and computer-coded authenticity at the classroom level were sufficiently high ( r = .602 for archival transcripts and .687 for audio recordings) to provide a valuable complement to human coding in research efforts.

scholarly journals Developing a fuzzy logic based system for monitoring and early detection of residential fire based on thermistor sensors

2015 ◽  
Vol 12 (1) ◽  
pp. 63-89 ◽  
Author(s):  
Mirjana Maksimovic ◽  
Vladimir Vujovic ◽  
Branko Perisic ◽  
Vladimir Milosevic
Keyword(s):  
Fuzzy Logic ◽  
Real World ◽  
Network Performance ◽  
Performance Monitoring ◽  
Sensor Nodes ◽  
Sensor Data ◽  
Critical Event ◽  
The Internet ◽  
Network Performance Monitoring

The recent proliferation of global networking has an enormous impact on the cooperation of smart elements, of arbitrary kind and purpose that can be located anywhere and interact with each other according to the predefined protocol. Furthermore, these elements have to be intelligently orchestrated in order to support distributed sensing and/or monitoring/control of real world phenomena. That is why the Internet of Things (IoT) concept raises like a new, promising paradigm for Future Internet development. Considering that Wireless Sensor Networks (WSNs) are envisioned as integral part of arbitrary IoTs, and the potentially huge number of cooperating IoTs that are usually used in the real world phenomena monitoring and management, the reliability of individual sensor nodes and the overall network performance monitoring and improvement are definitely challenging issues. One of the most interesting real world phenomena that can be monitored by WSN is indoor or outdoor fire. The incorporation of soft computing technologies, like fuzzy logic, in sensor nodes has to be investigated in order to gain the manageable network performance monitoring/control and the maximal extension of components life cycle. Many aspects, such as routes, channel access, locating, energy efficiency, coverage, network capacity, data aggregation and Quality of Services (QoS) have been explored extensively. In this article two fuzzy logic approaches, with temporal characteristics, are proposed for monitoring and determining confidence of fire in order to optimize and reduce the number of rules that have to be checked to make the correct decisions. We assume that this reduction may lower sensor activities without relevant impact on quality of operation and extend battery life directly contributing the efficiency, robustness and cost effectiveness of sensing network. In order to get a real time verification of proposed approaches a prototype sensor web node, based on Representational State Transfer (RESTful) services, is created as an infrastructure that supports fast critical event signaling and remote access to sensor data via the Internet.

Supporting Quality of Service for Internet Multimedia Applications

2011 ◽  
pp. 3622-3628
Author(s):  
Yew-Hock Ang
Keyword(s):  
Quality Of Service ◽  
Large Scale ◽  
The Internet ◽  
Digital Audio ◽  
Data Traffic ◽  
Service Model ◽  
The Past ◽  
Ip Multicasting ◽  
Scale Experiment

The Internet has gone from near-invisibility to near-ubiquity and penetrated into every aspect of society in the past decades (Department of Commerce, 1998). The application scenarios have also changed dramatically, and now demand a more sophisticated service model from the network. In the early 1990s, there was a large-scale experiment in sending digitized voice and video across the Internet through a packetswitched infrastructure (Braden, Clark, & Shenker, 1994). These highly-visible experiments have depended upon three enabling technologies: (1) Many modern workstations now come equipped with built-in multimedia hardware, (2) IP multicasting, which was not yet generally available in commercial routers, and (3) Highly-sophisticated digital audio and video applications have been developed. It became clear from these experiments that an important technical element of the Internet is still missing: multimedia, which dominate increasing proportion of today’s data traffic, are not well supported on the Internet.

Data Quality in Cooperative Information Systems

2011 ◽  
pp. 297-301
Author(s):  
Carla Marchetti ◽  
Massimo Mecella ◽  
Monica Scannapieco ◽  
Antoninio Virgillito
Keyword(s):  
Information System ◽  
Information Systems ◽  
Data Quality ◽  
Real World ◽  
Large Scale ◽  
Quality Certification ◽  
Geographically Distributed ◽  
Cooperative Information Systems

A Cooperative Information System (CIS) is a large-scale information system that interconnects various systems of different and autonomous organizations, geographically distributed and sharing common objectives (De Michelis et al., 1997). Among the different resources that are shared by organizations, data are fundamental; in real world scenarios, organization A may not request data from organization B, if it does not trust B’s data (i.e., if A does not know that the quality of the data that B can provide is high). As an example, in an e-government scenario in which public administrations cooperate in order to fulfill service requests from citizens and enterprises (Batini & Mecella, 2001), administrations very often prefer asking citizens for data rather than from other administrations that have stored the same data, because the quality of such data is not known. Therefore, lack of cooperation may occur due to lack of quality certification.

Adaptive Reinforcement Learning Integrating Exploitation-and Exploration-oriented Learning

1999 ◽  
Vol 3 (6) ◽  
pp. 474-478
Author(s):  
Satoshi Kurihara ◽  
◽  
Rikio Onai ◽  
Toshiharu Sugawara ◽  
Keyword(s):  
Reinforcement Learning ◽  
Mobile Robots ◽  
Real World ◽  
Large Scale ◽  
Autonomous Systems ◽  
Learning System ◽  
The Internet ◽  
Changing Environment ◽  
The Real ◽  
Exploitation And Exploration

We propose and evaluate an adaptive reinforcement learning system that integrates both exploitation- and exploration-oriented learning (ArLee). Compared to conventional reinforcement learning, ArLee is more robust in a dynamically changing environment and conducts exploration-oriented learning efficiently even in a large-scale environment. It is thus well suited for autonomous systems, for example, software agents and mobile robots, that operate in dynamic, large-scale environments, such as the real world and the Internet. Simulation demonstrates the learning system’s basic effectiveness.

Evaluating citizens concern about the quality of their drinking water

2005 ◽  
Vol 5 (2) ◽  
pp. 17-22 ◽  
Author(s):  
A. Contu ◽  
M. Carlini ◽  
A. Maccioni ◽  
P. Meloni ◽  
M. Schintu
Keyword(s):  
Drinking Water ◽  
General Public ◽  
Large Scale ◽  
Tap Water ◽  
General Information ◽  
The Internet ◽  
Human Consumption ◽  
Experimental Protocol ◽  
Online Queries

Citizens' concern about the quality of their drinking water stems especially from poor or wrong information. A tool allowing consumers to consult the Internet for the “label characteristics” of the tap water they are drinking, and to look for general information on water for human consumption has been developed. The tool allows to carry out online queries contributing to optimising management and using the water resource. Thanks to the implementation of an experimental protocol in secondary schools in Sardinia (Italy), it has been possible to test the efficacy of the tool by adapting its content to a large scale of users and to carry out a didactic-educational drill on the theme of water resources. To assess the weight of the aesthetics (taste, colour, and odour) in determining the degree of acceptability of the water, two tests were carried on both resident and non-resident assessors. This study allowed to survey the perception of the general public and the concerns arising from them.

scholarly journals A survey on VV&A of large-scale simulations

2019 ◽  
Vol 3 (1) ◽  
pp. 63-86 ◽  
Author(s):  
Yanan Wang ◽  
Jianqiang Li ◽  
Sun Hongbo ◽  
Yuan Li ◽  
Faheem Akhtar ◽  
...  
Keyword(s):  
Real World ◽  
Design Methodology ◽  
Large Scale ◽  
Content Type ◽  
Related Research ◽  
Network Simulations ◽  
Large Scale Simulations ◽  
Research Studies

Purpose Simulation is a well-known technique for using computers to imitate or simulate the operations of various kinds of real-world facilities or processes. The facility or process of interest is usually called a system, and to study it scientifically, we often have to make a set of assumptions about how it works. These assumptions, which usually take the form of mathematical or logical relationships, constitute a model that is used to gain some understanding of how the corresponding system behaves, and the quality of these understandings essentially depends on the credibility of given assumptions or models, known as VV&A (verification, validation and accreditation). The main purpose of this paper is to present an in-depth theoretical review and analysis for the application of VV&A in large-scale simulations. Design/methodology/approach After summarizing the VV&A of related research studies, the standards, frameworks, techniques, methods and tools have been discussed according to the characteristics of large-scale simulations (such as crowd network simulations). Findings The contributions of this paper will be useful for both academics and practitioners for formulating VV&A in large-scale simulations (such as crowd network simulations). Originality/value This paper will help researchers to provide support of a recommendation for formulating VV&A in large-scale simulations (such as crowd network simulations).

Sign in / Sign up

Export Citation Format

Share Document