Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765275
Mohammad Ehsan Basiri, Neshat Safarian, Hadi Khosravi Farsani
Sentiment analysis of online reviews has attracted an increasing attention from both academia and industry. Although online reviews are valuable sources of information for detecting public opinion towards different aspects of products, they may be written by spammers with different purposes. In order to detect such spam reviews, several methods have been proposed for English language but no study has been reported on Persian spam detection so far. In the current study, Persian reviews of cell-phones are investigated to find spam type 1 and type 2 which are fake reviews and reviews only written about brands, respectively. In the proposed framework a labeled dataset, SpamPer, is first created using a majority voting on the answers of 11 questions previously designed for spam detection by human annotators. Then several preprocessing steps for Persian language are performed to refine the training data. Finally review-based and metadata features are extracted. The obtained results on 3000 reviews of SpamPer shows that the highest accuracy is obtained using the decision tree with 0.78 F1-measure. Moreover, the results reveal that SVM for unbalanced data and decision tree for balanced data achieve better performance when they are trained on the combination of metadata and review-based features.
{"title":"A Supervised Framework for Review Spam Detection in the Persian Language","authors":"Mohammad Ehsan Basiri, Neshat Safarian, Hadi Khosravi Farsani","doi":"10.1109/ICWR.2019.8765275","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765275","url":null,"abstract":"Sentiment analysis of online reviews has attracted an increasing attention from both academia and industry. Although online reviews are valuable sources of information for detecting public opinion towards different aspects of products, they may be written by spammers with different purposes. In order to detect such spam reviews, several methods have been proposed for English language but no study has been reported on Persian spam detection so far. In the current study, Persian reviews of cell-phones are investigated to find spam type 1 and type 2 which are fake reviews and reviews only written about brands, respectively. In the proposed framework a labeled dataset, SpamPer, is first created using a majority voting on the answers of 11 questions previously designed for spam detection by human annotators. Then several preprocessing steps for Persian language are performed to refine the training data. Finally review-based and metadata features are extracted. The obtained results on 3000 reviews of SpamPer shows that the highest accuracy is obtained using the decision tree with 0.78 F1-measure. Moreover, the results reveal that SVM for unbalanced data and decision tree for balanced data achieve better performance when they are trained on the combination of metadata and review-based features.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"47 1","pages":"203-207"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84532437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765294
M. Keyvanpour, Mehrnoush Barani Shirzad, Haniyeh Rashidghalam
Automatically capturing the main points from a single document or multiple documents is a challenging requirement. Extractive text summarization which refers to providing a brief summary extract significant sentences from text, deals with several issues. Recently a considerable amount of work has considered learning approaches as text summarization solutions. Intensive researches have surveyed different strategies for text summarization. This paper influenced by the merit performance of learning methods for this task, analytically reviewed current algorithms. In this paper we suggest a framework called "ELTS" including classification of existing learning based algorithm, introducing several criteria in order to make comparison between current models and an analysis based on these criteria. We offer ELTS with the aim to enhance future research which attempts to a) solve current methods defects, b) employ existing strategies according to their requirements or c) make analytical comparison between current and future work.
{"title":"ELTS: A Brief Review for Extractive Learning-Based Text Summarizatoin Algorithms","authors":"M. Keyvanpour, Mehrnoush Barani Shirzad, Haniyeh Rashidghalam","doi":"10.1109/ICWR.2019.8765294","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765294","url":null,"abstract":"Automatically capturing the main points from a single document or multiple documents is a challenging requirement. Extractive text summarization which refers to providing a brief summary extract significant sentences from text, deals with several issues. Recently a considerable amount of work has considered learning approaches as text summarization solutions. Intensive researches have surveyed different strategies for text summarization. This paper influenced by the merit performance of learning methods for this task, analytically reviewed current algorithms. In this paper we suggest a framework called \"ELTS\" including classification of existing learning based algorithm, introducing several criteria in order to make comparison between current models and an analysis based on these criteria. We offer ELTS with the aim to enhance future research which attempts to a) solve current methods defects, b) employ existing strategies according to their requirements or c) make analytical comparison between current and future work.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"28 1","pages":"234-239"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75946345","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765279
F. Kermani, Shirin Ghanbari
Automatic extractive text summarization is the process of condensing textual information while preserving the important concepts. The proposed method after performing pre-processing on input Persian news articles generates a feature vector of salient sentences from a combination of statistical, semantic and heuristic methods and that are scored and concatenated accordingly. The scoring of the salient features is based on the article’s title, proper nouns, pronouns, sentence length, keywords, topic words, sentence position, English words, and quotations. Experimental results on measurements including recall, F-measure, ROUGE-N are presented and compared to other Persian summarizers and shown to provide higher performance.
{"title":"Extractive Persian Summarizer for News Websites","authors":"F. Kermani, Shirin Ghanbari","doi":"10.1109/ICWR.2019.8765279","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765279","url":null,"abstract":"Automatic extractive text summarization is the process of condensing textual information while preserving the important concepts. The proposed method after performing pre-processing on input Persian news articles generates a feature vector of salient sentences from a combination of statistical, semantic and heuristic methods and that are scored and concatenated accordingly. The scoring of the salient features is based on the article’s title, proper nouns, pronouns, sentence length, keywords, topic words, sentence position, English words, and quotations. Experimental results on measurements including recall, F-measure, ROUGE-N are presented and compared to other Persian summarizers and shown to provide higher performance.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"7 1","pages":"85-89"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88891498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765274
Alireza Bitarafan, Chitra Dadkhah
Online stores and e-commerce platforms have become increasingly popular in recent years, and a reasonable approach to compare the available products is to use comments or feedbacks written by other online users for each product. Therefore, these platforms can be a great opportunity for spammers to promote or demote their target products with fake reviews. So far, there is plenty of studies done with the purpose of distinguishing spam reviews or spammers from genuine ones, but it should not be neglected that often spammers work in collusion with each other to control the rating score of a product more naturally. Hence, this article focuses on the latter aspect i.e., review spammer group detection. In most of the previous works, Frequent Item set Mining (FIM) is applied in the early stage to find candidate groups and then an unsupervised ranking procedure is done based on some predefined features. Although, FIM methods mostly suffer from threshold setting, i.e., using low support values causes inefficiency and high support values ignore some useful patterns. Furthermore, instead of unsupervised methods, semi-supervised ones which don’t need many labeled data, can improve the accuracy of detection greatly. In this article, we tackle the above-mentioned challenges taking advantage of some labeled instances in a Heterogeneous Information Network (HIN). Using a HIN can preserve the semantics between different kinds of nodes in the network. Also, we extract candidate groups using spammer behaviors and their relations which makes it a robust approach when spammers decide to be more intelligent. Experiments on a real-life Yelp dataset show the efficiency of our approach.
{"title":"SPGD_HIN: Spammer Group Detection based on Heterogeneous Information Network","authors":"Alireza Bitarafan, Chitra Dadkhah","doi":"10.1109/ICWR.2019.8765274","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765274","url":null,"abstract":"Online stores and e-commerce platforms have become increasingly popular in recent years, and a reasonable approach to compare the available products is to use comments or feedbacks written by other online users for each product. Therefore, these platforms can be a great opportunity for spammers to promote or demote their target products with fake reviews. So far, there is plenty of studies done with the purpose of distinguishing spam reviews or spammers from genuine ones, but it should not be neglected that often spammers work in collusion with each other to control the rating score of a product more naturally. Hence, this article focuses on the latter aspect i.e., review spammer group detection. In most of the previous works, Frequent Item set Mining (FIM) is applied in the early stage to find candidate groups and then an unsupervised ranking procedure is done based on some predefined features. Although, FIM methods mostly suffer from threshold setting, i.e., using low support values causes inefficiency and high support values ignore some useful patterns. Furthermore, instead of unsupervised methods, semi-supervised ones which don’t need many labeled data, can improve the accuracy of detection greatly. In this article, we tackle the above-mentioned challenges taking advantage of some labeled instances in a Heterogeneous Information Network (HIN). Using a HIN can preserve the semantics between different kinds of nodes in the network. Also, we extract candidate groups using spammer behaviors and their relations which makes it a robust approach when spammers decide to be more intelligent. Experiments on a real-life Yelp dataset show the efficiency of our approach.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"126 1","pages":"228-233"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76704767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765253
Elham Sodagari, M. Keyvanpour
Requirements engineering is an important process in software engineering. An important task in the area is to select and optimize the requirements by considering requirement interaction management. Actually, the main goal of this area is to make best and most optimal choices among all the possible requirements, taking into account the dependencies between requirements. In this way, there are challenges and threats at all stages, including requirements engineering, search-based software engineering, requirements interactive management, and selection and optimization of requirements. The Identification and classification of challenges help to better understand the problem and find better solutions. We intend to examine and classify the main challenges in the papers in this area. Our goal in this article is to classify the challenges in this area from a new perspective.
{"title":"Challenges Classification of Software Requirements Interaction Management Using Search-Based Methods","authors":"Elham Sodagari, M. Keyvanpour","doi":"10.1109/ICWR.2019.8765253","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765253","url":null,"abstract":"Requirements engineering is an important process in software engineering. An important task in the area is to select and optimize the requirements by considering requirement interaction management. Actually, the main goal of this area is to make best and most optimal choices among all the possible requirements, taking into account the dependencies between requirements. In this way, there are challenges and threats at all stages, including requirements engineering, search-based software engineering, requirements interactive management, and selection and optimization of requirements. The Identification and classification of challenges help to better understand the problem and find better solutions. We intend to examine and classify the main challenges in the papers in this area. Our goal in this article is to classify the challenges in this area from a new perspective.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"104 1","pages":"246-251"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80486288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765268
Sajjad Salehi, F. Taghiyareh
Individuals may change their opinion in effect of a wide range of factors like interaction with peer groups, governmental policies and personal intentions. Works in this area mainly focus on individuals in social network and their interactions while neglect other factors. In this paper we have introduced an opinion formation model that consider the internal tendency as a personal feature of individuals in social network. In this model agents may trust, distrust or be neutral to their neighbors. They modify their opinion based on the opinion of their neighbors, trust/distrust to them while considering the internal tendency. The results of simulation show that this model can predict the opinion of social network especially when the average of nodal degree and clustering coefficient are high enough. Since this model can predict the preferences of individuals in market, it can be used to define marketing and production strategy.
{"title":"Introspective Agents in Opinion Formation Modeling to Predict Social Market","authors":"Sajjad Salehi, F. Taghiyareh","doi":"10.1109/ICWR.2019.8765268","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765268","url":null,"abstract":"Individuals may change their opinion in effect of a wide range of factors like interaction with peer groups, governmental policies and personal intentions. Works in this area mainly focus on individuals in social network and their interactions while neglect other factors. In this paper we have introduced an opinion formation model that consider the internal tendency as a personal feature of individuals in social network. In this model agents may trust, distrust or be neutral to their neighbors. They modify their opinion based on the opinion of their neighbors, trust/distrust to them while considering the internal tendency. The results of simulation show that this model can predict the opinion of social network especially when the average of nodal degree and clustering coefficient are high enough. Since this model can predict the preferences of individuals in market, it can be used to define marketing and production strategy.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"53 14","pages":"28-34"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91483747","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765259
M. Asemani, Fatemeh Abdollahei, Fatemeh Jabbari
Internet of things (IoT) offers some advanced vertical services through data capture and processing. To provide the services to the end user, some other services such as data analytics, device management, and connection management should be delivered in the IoT ecosystem. IoT platform is the element, which delivers the central process and management, and the vertical services to the end users, by providing some tools, and computation for device management and data lifecycle management (from sensors networks to the end users). Although there are lots of IoT platform products in the market, there is not any unique, precise, or standard definition with the detailed description of IoT platform, which includes various definitions and functionalities of IoT platform from scientific and market perspective, on both cloud and fog computing resources. In this paper, a novel, comprehensive definition for IoT platform and its attributes in Cloud and Fog layer is proposed, which is extracted from scientific definitions in academic papers, the definition, and features for commercial products provided by IoT leader companies, as well as the description of IoT platform in some open source projects.
{"title":"Understanding IoT Platforms : Towards a comprehensive definition and main characteristic description","authors":"M. Asemani, Fatemeh Abdollahei, Fatemeh Jabbari","doi":"10.1109/ICWR.2019.8765259","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765259","url":null,"abstract":"Internet of things (IoT) offers some advanced vertical services through data capture and processing. To provide the services to the end user, some other services such as data analytics, device management, and connection management should be delivered in the IoT ecosystem. IoT platform is the element, which delivers the central process and management, and the vertical services to the end users, by providing some tools, and computation for device management and data lifecycle management (from sensors networks to the end users). Although there are lots of IoT platform products in the market, there is not any unique, precise, or standard definition with the detailed description of IoT platform, which includes various definitions and functionalities of IoT platform from scientific and market perspective, on both cloud and fog computing resources. In this paper, a novel, comprehensive definition for IoT platform and its attributes in Cloud and Fog layer is proposed, which is extracted from scientific definitions in academic papers, the definition, and features for commercial products provided by IoT leader companies, as well as the description of IoT platform in some open source projects.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"28 1","pages":"172-177"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82052735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765247
Morteza Moradi, Farhad Bayat, M. Charmi
Reduced output quality and being unaware of content are among major issues with traditional image compression techniques. Such issues cause some critical problems when it comes to quality-intensive applications, including object/face detection and recognition, Web-based image viewers and management systems, etc. On the other side, efficiency of Web-based image search engines and retrieval systems in terms of user experience and usability could be affected. In order to cope with these challenges, a novel image compression method is proposed that takes advantages of collective human cognitive intelligence to detect the salient object(s) based on the recognized key concept(s). Then, other less-important regions/objects will be subject to the safe compression. Such an approach, besides preserving semantic aspects of the images that will result in smart (concept-aware) compression, could provide some crowdsourced labels for more efficient indexing and annotating of images. In this regard, two birds could be beaten with one stone: compressing Web images with respect to their content/concept and annotating them with crowd-suggested labels. The experimental results as well as user acceptance evaluation proved the efficacy of the introduced method.
{"title":"Concept-Aware Web Image Compression Based on Crowdsourced Salient Object Detection","authors":"Morteza Moradi, Farhad Bayat, M. Charmi","doi":"10.1109/ICWR.2019.8765247","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765247","url":null,"abstract":"Reduced output quality and being unaware of content are among major issues with traditional image compression techniques. Such issues cause some critical problems when it comes to quality-intensive applications, including object/face detection and recognition, Web-based image viewers and management systems, etc. On the other side, efficiency of Web-based image search engines and retrieval systems in terms of user experience and usability could be affected. In order to cope with these challenges, a novel image compression method is proposed that takes advantages of collective human cognitive intelligence to detect the salient object(s) based on the recognized key concept(s). Then, other less-important regions/objects will be subject to the safe compression. Such an approach, besides preserving semantic aspects of the images that will result in smart (concept-aware) compression, could provide some crowdsourced labels for more efficient indexing and annotating of images. In this regard, two birds could be beaten with one stone: compressing Web images with respect to their content/concept and annotating them with crowd-suggested labels. The experimental results as well as user acceptance evaluation proved the efficacy of the introduced method.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"36 1","pages":"221-227"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86718777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765251
Kaveh Bakhtiyari, M. Taghavi, Milad Taghavi, J. Bentahar
Computational feature recognition is an essential component for intelligent systems to sense the objects and environments. This paper proposes a novel conceptual model, named Ambiance Signal Processing (AmSiP), to identify objects’ features when they are not directly accessible by sensors. AmSiP analyzes the surrounding and ambiance of objects/subjects collaboratively to recognize the object’s features instead of concentrating on each individual and accessible object. To validate the proposed model, this study runs an experiment with 50 participants, whose emotional state variations are estimated by measuring the surroundings features and the emotions of other people in the same environment. The results of a t-Test on the data collected from this experiment showed that users’ emotions were being changed in a course of time during the experiment; however, AmSiP could estimate subjects’ emotions reliably according to the environmental characteristics and similar patterns. To evaluate the reliability and efficiency of this model, a collaborative affective computing system was implemented using keyboard keystroke dynamics and mouse interactions of the users whose emotions were affected by different types of music. In comparison with other conventional techniques (explicit access), the prediction was reliable. Although the developed model sacrifices a minor accuracy, it earns the superiority of uncovering blind knowledge about the subjects out of the sensors’ direct access.
{"title":"Ambiance Signal Processing: A Study on Collaborative Affective Computing","authors":"Kaveh Bakhtiyari, M. Taghavi, Milad Taghavi, J. Bentahar","doi":"10.1109/ICWR.2019.8765251","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765251","url":null,"abstract":"Computational feature recognition is an essential component for intelligent systems to sense the objects and environments. This paper proposes a novel conceptual model, named Ambiance Signal Processing (AmSiP), to identify objects’ features when they are not directly accessible by sensors. AmSiP analyzes the surrounding and ambiance of objects/subjects collaboratively to recognize the object’s features instead of concentrating on each individual and accessible object. To validate the proposed model, this study runs an experiment with 50 participants, whose emotional state variations are estimated by measuring the surroundings features and the emotions of other people in the same environment. The results of a t-Test on the data collected from this experiment showed that users’ emotions were being changed in a course of time during the experiment; however, AmSiP could estimate subjects’ emotions reliably according to the environmental characteristics and similar patterns. To evaluate the reliability and efficiency of this model, a collaborative affective computing system was implemented using keyboard keystroke dynamics and mouse interactions of the users whose emotions were affected by different types of music. In comparison with other conventional techniques (explicit access), the prediction was reliable. Although the developed model sacrifices a minor accuracy, it earns the superiority of uncovering blind knowledge about the subjects out of the sensors’ direct access.","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"11 1","pages":"35-40"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85341351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-04-01DOI: 10.1109/ICWR.2019.8765282
Mitra Bavakhani, Alireza Yari, A. Sharifi
Due to the expansion of social networks and media such as Tweeter, Facebook, LinkedIn, and different weblogs, and the great increase in information sharing and comments, Which typically are in the form of text data, big enough to be recognized as big data., and with respect to the importance of these data for the analysis of customers’ priorities, needs and their attitudes toward different products, finding and extracting data from their comments, are the primary goals of this research. To serve this purpose, this research has used deep learning approach, and multilayer neural network methods in order to extract the polarity of customers’ opinions and comments in two domains of products/services ranging from restaurant to laptop.The findings of this study indicate that the proposed model using the potencies of the long short-term-memory networks, is able to determine the comments’ polarity with 85 % and 84.62 % precision for restaurant and laptop domains respectively, in such a way that the results are relatively more accurate than the results of other methods
{"title":"A Deep Learning Approach for Extracting Polarity from Customers’ Reviews","authors":"Mitra Bavakhani, Alireza Yari, A. Sharifi","doi":"10.1109/ICWR.2019.8765282","DOIUrl":"https://doi.org/10.1109/ICWR.2019.8765282","url":null,"abstract":"Due to the expansion of social networks and media such as Tweeter, Facebook, LinkedIn, and different weblogs, and the great increase in information sharing and comments, Which typically are in the form of text data, big enough to be recognized as big data., and with respect to the importance of these data for the analysis of customers’ priorities, needs and their attitudes toward different products, finding and extracting data from their comments, are the primary goals of this research. To serve this purpose, this research has used deep learning approach, and multilayer neural network methods in order to extract the polarity of customers’ opinions and comments in two domains of products/services ranging from restaurant to laptop.The findings of this study indicate that the proposed model using the potencies of the long short-term-memory networks, is able to determine the comments’ polarity with 85 % and 84.62 % precision for restaurant and laptop domains respectively, in such a way that the results are relatively more accurate than the results of other methods","PeriodicalId":6680,"journal":{"name":"2019 5th International Conference on Web Research (ICWR)","volume":"12 1","pages":"276-280"},"PeriodicalIF":0.0,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85875530","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：