An artificial neural network based approach for sentiment analysis of opinionated text
23 Oct 2012-pp 37-42
TL;DR: A sentiment classification model using back-propagation artificial neural network (BPANN) is proposed that combines the strength of BPANN in classification accuracy with utilizing intrinsic domain knowledge available in the sentiment lexicons.
Abstract: The Internet and Web 2.0 social media have emerged as an important medium for expressing sentiments, opinions, evaluations, and reviews. Sentiment analysis or opinion mining is becoming an open research domain due to the abundance of discussion forums, Weblogs, e-commerce portals, social networking and content sharing sites where people tend to express their opinions. Sentiment Analysis involves classifying text documents based on the opinion expressed being positive or negative about a given topic. This paper proposes a sentiment classification model using back-propagation artificial neural network (BPANN). Information Gain and three popular sentiment lexicons are used to extract sentiment representing features that are then used to train and test the BPANN. This novel approach combines the strength of BPANN in classification accuracy with utilizing intrinsic domain knowledge available in the sentiment lexicons. The results obtained on the movie-review corpora have shown that the proposed approach has been able to reduce dimensionality, while producing accurate sentiment based classification of text.
Citations
More filters
Dissertation•
01 Jan 2018
TL;DR: Arapca dil icin boyle buyuk olcekli verinin isleme ve hesaplama gereksinimleri, islenme ve analiz edilme gereklinimlerine uyan buyuk lceklini metin derlemi gibi.
Abstract: Bilgisayar tabanli teknolojilerinde toplanan verilerin kullanimi ve buyuklugu surekli
artimaktadir. Bu surekli artan buyuk verinin isleme ve hesaplama gereksinimleri,
ozellikle Dogal Dil Isleme NLP uygulamalarinda yeni bir zorluklar ortaya koymaktadir.
Bu zorluklardan biri, Duygu Analizi (DA) gibi NLP uygulamalarinda Buyuk Verilerin ele
alinma, islenme ve analiz edilme gereksinimlerine uyan buyuk olcekli metin derlemi gibi
zengin bir dilsel kaynagin saglanmasidir. Arapca dil icin boyle buyuk olcekli bir
kaynagin bulunmamasinin zorlugu cozmek icin, cevrimici haber Media'yi ve buyuk veri
kaynagi tarafindan uretilen acik kaynak meta verilerini kullanarak insa edilen GDELT
buyuk olcekli Arapca duygu analiz derlemimizi (GLASC) tanitmaktayiz. GLASC derlimi,
(Pozitif, Negatif ve Notr) kategorilerinde duzenlenen toplam 620.082 haber
makalesinden olusmaktadir ve ayni zamanda, derlemimizdeki her haber makalesinin
(-1 ve 1) araliginda bir duygu puani vardir. Ayrica, Makine ogrenme siniflandirma ve
regresyon yaklasimlarina dayali bir Arapca belge seviyesinde duygu analizi sistemi olusturmak icin GLASC derlemi kullanip bazi deneyler gerceklestirdik. Onerilen Makine
ogrenmesi modellerini egitmek icin, farkli oznitelik cikarma ve ozellik agirliklandirma
yontemlerini kullanarak GLASC derlemimizden farkli veri kumeleri urettik. Duygu analizi
gorevi icin sikca kullanilan siniflandirma ve regresyon, yontemlerinin testini iceren
karsilastirmali genis bir calisma gerceklestirilmistir. Buna ek olarak, cesitli kapsamli
deneyler kullanarak, duygu analizi icin siniflandirma performansinin iyilestirilmesinin
etkisini dogrulamak icin, (Cuvallama, Yukseltme, Rasgele altuzay ve Offekleme gibi)
topluluk ogrenme yontemlerinin cesitli turleri arastirilmistir. Bu calismada, makine
ogrenme yaklasimlarini ve kavrama dayali bir duyugu sozlugunu kullanarak, cumle
duzeyinde Arapca icin kavram tabanli bir duygu analiz sistemi sunulmustur. Yakin
zamanda cikan Ingilizce SenticNet_v4'u Arapca'ya cevirerek Arapca kavram temelli bir
duygu sozlugu uretmek icin bir yaklasim onerilmistir. Uretilen Arapca konsept temelli
duygu sozlugu Ar-SenticNet toplam 48k Arapca kavram icermektedir. Arapca cumleden
Konsepti cikarmak icin, anlamsal ayristirici olarak adlandirilan kural tabanli bir
kavramlari cikarma algoritmasi onerildi ve uygulanmistir. Ayrica, kavram tabanli cumle
duzeyinde Arapca duygu analizi sisteminin olusturulmasi icin farkli ozellikler cikarim ve
gosterim teknikleri sunurak kullandik. Kavram tabanli cumle duzeyinde Arapca duygu
analiz sisteminin karar modeli olusturmak icin, farkli siniflandirma yontemi ve
siniflandirici fuzyon modelleri kullanilarak, onerdigimiz ozellikler kumelerimizin farkli
kombinasyonlari ile kapsamli ve karsilastirmali deneyler yapilmistir. Elde edilen deney
sonuclarimiza dayanarak, onerilen Makine ogrenmesi tabanli Dokuman duzeyinde
Arapca duygu analiz sistemimiz icin, en iyi performans % 92.35 F-skoru degeri olan
SVM-HMM siniflandirici fuzyon modeliyle ve 0.183 RMSE degeri olan SVR regresyon
modeli ile, gerceklestirilmistir. Ote yandan, onerilen konsept tabanli cumle duzeyinde
Arapca duygu analiz sistemimiz icin, en iyi performans, %93.92'lik bir F-skoru degerine
sahip SVM-LR siniflayici fuzyon modeliyle ve 0.078 RMSE degeri olan SVR regresyon
modeli ile, gerceklestirilmistir.
4 citations
Cites methods from "An artificial neural network based ..."
...In this work, we considered using five different ML based classification algorithms which are widely used in SA such as (SVM [15], HMM [69], NB [107], ANN [108] and KNN [109]), to build the classification model for the Arabic document level SA....
[...]
07 Sep 2015
TL;DR: A genetic algorithm able to mine comparative sentences from short sentences based on sequential patterns classification is proposed, reaching accuracy levels of 73% and better accuracy against literature baseline approaches are indicated.
Abstract: Comparative opinions represent a way of users express their preferences about two or more entities. In this paper we address the problem of comparative sentences mining focused on social medias. We propose a genetic algorithm able to mine comparative sentences from short sentences based on sequential patterns classification. A comparison among classifiers regarding comparative sentences analysis is also presented. Our results indicate better accuracy for the proposed technique against literature baseline approaches, reaching accuracy levels of 73%.
3 citations
Cites methods from "An artificial neural network based ..."
...We did three pre-processing steps: (1) stop words removal, (2) stemming and (3) 1000 features extraction based on information gain index [10]....
[...]
14 Oct 2014
TL;DR: In this paper corpus-based information extraction and opinion mining method is proposed for restaurant reviews, and the module is a part of a Russian knowledge-based recommendation system.
Abstract: In this paper corpus-based information extraction and opinion mining method is proposed. Our domain is restaurant reviews, and our information extraction and opinion mining module is a part of a Russian knowledge-based recommendation system.
3 citations
Additional excerpts
...Neural Networks [32, 34] 274 E....
[...]
01 Jan 2018
TL;DR: This approach focused on developing an LSTM RNN that could perform binary sentiment analysis for positively and negatively labeled sentences, and developed a collection of programs to classify individual sentences as either positive or negative.
Abstract: Sentiment analysis has taken on various machine learning approaches in order to optimize accuracy, precision, and recall. However, Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs) account for the context of a sentence by using previous predictions as additional input for future sentence predictions. Our approach focused on developing an LSTM RNN that could perform binary sentiment analysis for positively and negatively labeled sentences. In collaboration with Mariam Salloum, I developed a collection of programs to classify individual sentences as either positive or negative. This paper additionally looks into machine learning, neural networks, data preprocessing, implementation, and resulting comparisons.
3 citations
01 Jan 2014
TL;DR: A novel sentence ranking methodology based on the similarity score between a candidate sentence and benchmark summaries is introduced and the popular linear regression model achieved the best results in all evaluated datasets.
Abstract: Extractive summarization of text documents usually consists of ranking the document sentences and extracting the top-ranked sentences subject to the summary length constraints. In this paper, we explore the contribution of various supervised learning algorithms to the sentence ranking task. For this purpose, we introduce a novel sentence ranking methodology based on the similarity score between a candidate sentence and benchmark summaries. Our experiments are performed on three benchmark summarization corpora: DUC-2002, DUC2007 and MultiLing-2013. The popular linear regression model achieved the best results in all evaluated datasets. Additionally, the linear regression model, which included POS (Part-of-Speech)-based features, outperformed the one with statistical features only.
2 citations
References
More filters
22 Aug 2004
TL;DR: This research aims to mine and to summarize all the customer reviews of a product, and proposes several novel techniques to perform these tasks.
Abstract: Merchants selling products on the Web often ask their customers to review the products that they have purchased and the associated services. As e-commerce is becoming more and more popular, the number of customer reviews that a product receives grows rapidly. For a popular product, the number of reviews can be in hundreds or even thousands. This makes it difficult for a potential customer to read them to make an informed decision on whether to purchase the product. It also makes it difficult for the manufacturer of the product to keep track and to manage customer opinions. For the manufacturer, there are additional difficulties because many merchant sites may sell the same product and the manufacturer normally produces many kinds of products. In this research, we aim to mine and to summarize all the customer reviews of a product. This summarization task is different from traditional text summarization because we only mine the features of the product on which the customers have expressed their opinions and whether the opinions are positive or negative. We do not summarize the reviews by selecting a subset or rewrite some of the original sentences from the reviews to capture the main points as in the classic text summarization. Our task is performed in three steps: (1) mining product features that have been commented on by customers; (2) identifying opinion sentences in each review and deciding whether each opinion sentence is positive or negative; (3) summarizing the results. This paper proposes several novel techniques to perform these tasks. Our experimental results using reviews of a number of products sold online demonstrate the effectiveness of the techniques.
7,330 citations
"An artificial neural network based ..." refers background in this paper
...Positive and negative sentiment based summaries for product features from reviews were proposed by Hu and Liu (2004)....
[...]
01 Jan 2002
TL;DR: In this paper, the problem of classifying documents not by topic, but by overall sentiment, e.g., determining whether a review is positive or negative, was considered and three machine learning methods (Naive Bayes, maximum entropy classiflcation, and support vector machines) were employed.
Abstract: We consider the problem of classifying documents not by topic, but by overall sentiment, e.g., determining whether a review is positive or negative. Using movie reviews as data, we flnd that standard machine learning techniques deflnitively outperform human-produced baselines. However, the three machine learning methods we employed (Naive Bayes, maximum entropy classiflcation, and support vector machines) do not perform as well on sentiment classiflcation as on traditional topic-based categorization. We conclude by examining factors that make the sentiment classiflcation problem more challenging.
6,980 citations
06 Jul 2002
TL;DR: This work considers the problem of classifying documents not by topic, but by overall sentiment, e.g., determining whether a review is positive or negative, and concludes by examining factors that make the sentiment classification problem more challenging.
Abstract: We consider the problem of classifying documents not by topic, but by overall sentiment, e.g., determining whether a review is positive or negative. Using movie reviews as data, we find that standard machine learning techniques definitively outperform human-produced baselines. However, the three machine learning methods we employed (Naive Bayes, maximum entropy classification, and support vector machines) do not perform as well on sentiment classification as on traditional topic-based categorization. We conclude by examining factors that make the sentiment classification problem more challenging.
6,626 citations
Posted Content•
TL;DR: A simple unsupervised learning algorithm for classifying reviews as recommended (thumbs up) or not recommended (Thumbs down) if the average semantic orientation of its phrases is positive.
Abstract: This paper presents a simple unsupervised learning algorithm for classifying reviews as recommended (thumbs up) or not recommended (thumbs down). The classification of a review is predicted by the average semantic orientation of the phrases in the review that contain adjectives or adverbs. A phrase has a positive semantic orientation when it has good associations (e.g., "subtle nuances") and a negative semantic orientation when it has bad associations (e.g., "very cavalier"). In this paper, the semantic orientation of a phrase is calculated as the mutual information between the given phrase and the word "excellent" minus the mutual information between the given phrase and the word "poor". A review is classified as recommended if the average semantic orientation of its phrases is positive. The algorithm achieves an average accuracy of 74% when evaluated on 410 reviews from Epinions, sampled from four different domains (reviews of automobiles, banks, movies, and travel destinations). The accuracy ranges from 84% for automobile reviews to 66% for movie reviews.
4,526 citations
TL;DR: The authors examine the effect of consumer reviews on relative sales of books at Amazon.com and Barnesandnoble.com, and find that reviews are overwhelmingly positive at both sites, but there are more reviews and longer reviews at Amazon and that an improvement in a book's reviews leads to an increase in relative sales.
Abstract: The authors examine the effect of consumer reviews on relative sales of books at Amazon.com and Barnesandnoble.com. The authors find that (1) reviews are overwhelmingly positive at both sites, but there are more reviews and longer reviews at Amazon.com; (2) an improvement in a book's reviews leads to an increase in relative sales at that site; (3) for most samples in the study, the impact of one-star reviews is greater than the impact of five-star reviews; and (4) evidence from review-length data suggests that customers read review text rather than relying only on summary statistics.
4,180 citations