Coreference Resolution on Blogs and Commented News

This paper describes the process of adapting the Stanford Coreference resolution module to the Basque language, taking into account the characteristics of the language. The module has been integrated in a linguistic analysis pipeline obtaining an end-to-end coreference resolution system for the Basque language. The adaptation process explained can benefit and facilitate other languages with similar characteristics in the implementation of their coreference resolution systems. During the experimentation phase, we have demonstrated that language-specific features have a noteworthy effect on coreference resolution, obtaining a gain in CoNLL score of 7.07 with respect to the baseline system. We have also analysed the effect that preprocessing has in coreference resolution, comparing the results obtained with automatic mentions versus gold mentions. When gold mentions are provided, the results increase 11.5 points in CoNLL score in comparison with results obtained when automatic mentions are used. The contribution of each sieve is analysed concluding that morphology is essential for agglutinative languages to obtain good performance in coreference resolution. Finally, an error analysis of the coreference resolution system is presented which have revealed our system’s weak points and help to determine the improvements of the system. As a result of the error analysis, we have enriched the Basque coreference resolution adding new two sieves, obtaining an improvement of 0.24 points in CoNLL F1 when automatic mentions are used and of 0.39 points when the gold mentions are provided.

PurposeUp to date development in sentiment analysis has resulted in a symbolic growth in the volume of study, especially on more subjective text types, namely, product or movie reviews. The key difference between these texts with news articles is that their target is defined and unique across the text. Hence, the reviews on newspaper articles can deal with three subtasks: correctly spotting the target, splitting the good and bad content from the reviews on the concerned target and evaluating different opinions provided in a detailed manner. On defining these tasks, this paper aims to implement a new sentiment analysis model for article reviews from the newspaper.Design/methodology/approachHere, tweets from various newspaper articles are taken and the sentiment analysis process is done with pre-processing, semantic word extraction, feature extraction and classification. Initially, the pre-processing phase is performed, in which different steps such as stop word removal, stemming, blank space removal are carried out and it results in producing the keywords that speak about positive, negative or neutral. Further, semantic words (similar) are extracted from the available dictionary by matching the keywords. Next, the feature extraction is done for the extracted keywords and semantic words using holoentropy to attain information statistics, which results in the attainment of maximum related information. Here, two categories of holoentropy features are extracted: joint holoentropy and cross holoentropy. These extracted features of entire keywords are finally subjected to a hybrid classifier, which merges the beneficial concepts of neural network (NN), and deep belief network (DBN). For improving the performance of sentiment classification, modification is done by inducing the idea of a modified rider optimization algorithm (ROA), so-called new steering updated ROA (NSU-ROA) into NN and DBN for weight update. Hence, the average of both improved classifiers will provide the classified sentiment as positive, negative or neutral from the reviews of newspaper articles effectively.FindingsThree data sets were considered for experimentation. The results have shown that the developed NSU-ROA + DBN + NN attained high accuracy, which was 2.6% superior to particle swarm optimization, 3% superior to FireFly, 3.8% superior to grey wolf optimization, 5.5% superior to whale optimization algorithm and 3.2% superior to ROA-based DBN + NN from data set 1. The classification analysis has shown that the accuracy of the proposed NSU − DBN + NN was 3.4% enhanced than DBN + NN, 25% enhanced than DBN and 28.5% enhanced than NN and 32.3% enhanced than support vector machine from data set 2. Thus, the effective performance of the proposed NSU − ROA + DBN + NN on sentiment analysis of newspaper articles has been proved.Originality/valueThis paper adopts the latest optimization algorithm called the NSU-ROA to effectively recognize the sentiments of the newspapers with NN and DBN. This is the first work that uses NSU-ROA-based optimization for accurate identification of sentiments from newspaper articles.

The task of coreference resolution concerns the clustering of words and phrases referring to the same entity in text, either in the same document or across multiple documents. The task is challenging, as it concerns elements of named entity recognition and reading comprehension, as well as others. In this paper, we introduce DutchParliament, a new Dutch coreference resolution dataset obtained through the manual annotation of 74 government debates, expanded with a domain-specific class. In contrast to existing datasets, which are often composed of news articles, blogs or other documents, the debates in DutchParliament are transcriptions of speech, and therefore offer a unique structure and way of referencing compared to other datasets. By constructing and releasing this dataset, we hope to facilitate the research on coreference resolution in niche domains, with different characteristics than traditional datasets. The DutchParliament dataset was compared to SoNaR-1 and RiddleCoref, two other existing Dutch coreference resolution corpora, to highlight its particularities and differences from existing datasets. Furthermore, two coreference resolution models for Dutch, the rule-based DutchCoref model and the neural e2eDutch model, were evaluated on the DutchParliament dataset to examine their performance on the DutchParliament dataset. It was found that the characteristics of the DutchParliament dataset are quite different from that of the other two datasets, although the performance of the e2eDutch model does not seem to be significantly affected by this. Furthermore, experiments were conducted by utilizing the metadata present in the DutchParliament corpus to improve the performance of the e2eDutch model. The results indicate that the addition of available metadata about speakers has a beneficial effect on the performance of the model, although the addition of the gender of speakers seems to have a limited effect.

While it has been claimed that anaphora or coreference resolution plays an important role in opinion mining, it is not clear to what extent coreference resolution actually boosts performance, if at all. In this paper, we investigate the potential added value of coreference resolution for the aspect-based sentiment analysis of restaurant reviews in two languages, English and Dutch. We focus on the task of aspect category classification and investigate whether including coreference information prior to classification to resolve implicit aspect mentions is beneficial. Because coreference resolution is not a solved task in NLP, we rely on both automatically-derived and gold-standard coreference relations, allowing us to investigate the true upper bound. By training a classifier on a combination of lexical and semantic features, we show that resolving the coreferential relations prior to classification is beneficial in a joint optimization setup. However, this is only the case when relying on gold-standard relations and the result is more outspoken for English than for Dutch. When validating the optimal models, however, we found that only the Dutch pipeline is able to achieve a satisfying performance on a held-out test set and does so regardless of whether coreference information was included.

Coreference resolution (CR) is a key task in the automated analysis of characters in stories. Standard CR systems usually trained on newspaper texts have difficulties with literary texts, even with novels; a comparison with newspaper texts showed that average sentence length is greater in novels and the number of pronouns, as well as the percentage of direct speech is higher. We report promising evaluation results for a rule-based system similar to [Lee et al. 2011], but tailored to the domain which recognizes coreference chains in novels much better than CR systems like CorZu. Rule-based systems performed best on the CoNLL 2011 challenge [Pradhan et al. 2011]. Recent work in machine learning showed similar results as rule-based systems [Durett et al. 2013]. The latter has the advantage that its explanation component facilitates a fine grained error analysis for incremental refinement of the rules.

With the accelerated evolution of social networks, there is a tremendous increase in opinions by the people about products or services. While this user-generated content in natural language is intended to be valuable, its large amounts require use of content mining methods and NLP to uncover the knowledge for various tasks. In this study, sentiment analysis is used to analyze and understand the opinions of users using statistical approaches, knowledge-based approaches, hybrid approaches, and concept-based ontologies. Unfortunately, sentiment analysis also experiences a range of difficulties like colloquial words, negation handling, ambiguity in word sense, coreference resolution, which highlight another perspective emphasizing that sentiment analysis is certainly a restricted NLP problem. The purpose of this chapter is to discover how sentiment analysis is a restricted NLP problem. Thus, this chapter discussed the concept of sentiment analysis in the field of NLP and explored that sentiment analysis is a restricted NLP problem due to the sophisticated nature of natural language.

We aim to shed light on the state-of-the-art in NP coreference resolution by teasing apart the differences in the MUC and ACE task definitions, the assumptions made in evaluation methodologies, and inherent differences in text corpora. First, we examine three subproblems that play a role in coreference resolution: named entity recognition, anaphoricity determination, and coreference element detection. We measure the impact of each subproblem on coreference resolution and confirm that certain assumptions regarding these subproblems in the evaluation methodology can dramatically simplify the overall task. Second, we measure the performance of a state-of-the-art coreference resolver on several classes of anaphora and use these results to develop a quantitative measure for estimating coreference resolution performance on new data sets.

We live in an era of information explosion where media for information communication and information sharing have proliferated due to the onset of a digital age. One such important source and medium of consolidated, comprehensive and reliable information, which has persisted through centuries, is news articles. News articles have been considered as a primary source of information, ranging from local to global scope, and encompassing a wide array of domains, by most of the human population. However, as the technology for dissemination of information attracts innovations, it becomes equally important that the consumption and utilization of this information becomes smart as well. If processed properly and treated in an intelligent manner, the information contained in news articles can reveal useful insights and hidden patterns that might prove to be of critical importance in certain decision making pertaining to the topics and domains in question. The paper puts forward a novel framework created by combining Natural Language Processing methodologies like Sentiment Analysis and Information Retrieval, which are performed on top of a Deep Learning based classification with aptly defined classes for generating useful insights from a news article dataset. The paper contains the details regarding the implementation of this system and also a presentation of results on a particular data set of news articles - articles on HIV-AIDS, which have been categorized into meaningful classes and have been processed using aforementioned NLP tasks. Through this paper, a curated dataset of 3,000 news articles related to HIV-AIDS has also been introduced, which will be open sourced so as to inspire further research and experimentation in this specific domain.

Coreference resolution is the problem of determining which mention in a text refer to the same entities, and is a crucial and difficult step in every natural language processing task. Despite the efforts that have been made in the past to solve this problem, its performance still does not meet today’s applications requirements. Given the importance of the verbs in sentences, in this work we tried to incorporate three types of their information on coreference resolution problem, namely, selectional restriction of verbs on their arguments, semantic relation between verb pairs, and the truth that arguments of a verb cannot be coreferent of each other. As a needed resource for supporting our model, we generate a repository of semantic relations between verb pairs automatically using Distributional Memory (DM), a state-of-the-art framework for distributional semantics. This resource consists of pairs of verbs associated with their probable arguments, their role mapping, and significance scores based on our measures. Our proposed model for coreference resolution encodes verbs’ knowledge with Markov logic network rules on top of deterministic Stanford coreference resolution system. Experiment results show that this semantic layer can improve the recall of the Stanford system while preserves its precision and improves it slightly.

In this paper we have used sentiment analysis on news articles to see its effect on stock prices. We collected our dataset using Bing API which gave us links to news articles about a specific company. As no pre-existing sentiment dictionary specifically for stock articles exited, we created a specialized sentiment dictionary only meant to analyze stock articles. Two different machine learning algorithms were applied to the dataset and the accuracy of the two was compared. In order to test our results we attached an overall sentiment to each article in our data set which was compared to the predicted sentiment by the algorithm. We also compared the predicted results with the actual change in the stock prices on the market.

Opinion Mining and Sentiment Analysis in Natural Language Processing (NLP) are challenging, as they require deep understanding. Understanding involves methods that could differentiate between the facts of explicit and implicit, regular and irregular, syntactical and semantic language rules. Researches oriented towards Natural Language Processing and Sentiment Analysis have many unresolved problems like co-reference resolution, negation handling, anaphora resolution, named-entity recognition, and word-sense disambiguation. This paper is proposed to develop an Optimized Partial Ancestral Graph (O-PAG) which is capable of mining patterns in human interactions and compare it with an existing tree based pattern mining approach. The experimental results are exposed to number of frequent interactions made and execution time. Results indicate that the overall performance can reach considerable improvements on using O-PAG approach.

In this brief research note, we propose a set of concepts and methods for identifying and operationalizing controversial news items. Based on an analysis of online readers’ comments posted in response to political news articles in the lead-up to the most recent South Korean presidential election (May 9, 2017), we develop what we term as “controversy indicator.” We calculate a controversy score using the total number of reader comments and the proportion of “upvotes” (indicating approval) and “downvotes” (indicating disapproval) a given news article elicits. Our findings demonstrate the potential usefulness of the controversy indicator in understanding the contemporary news environment, which is becoming increasingly divisive and polluted with disinformation.

Reviews on the Google Play website are an application's user ratings, which contain users' ratings and comments. Review data describes user sentiment towards the application according to the ratings and comments are given. In practice, there is often a discrepancy between the rating and the comments given, resulting in a biased sentiment, so it is necessary to analyze the review to find out the sentiment contained therein. The Naïve Bayes method and the Support Vector Machine can be one of the methods that are often used to carry out sentiment analysis, because they have a good level of accuracy in the sentiment analysis process. In collecting data from the Google Play site using the Web Scraping technique with the google- play-scraper package from Python. Reviews that are successfully scraped then go through the preprocessing stage so that the data set is more structured. In the next stage, the data set is labeled based on the rating, and given a weight using TF-IDF. After classifying using the Naïve Bayes and Support Vector Machine methods, then evaluating using the confusion matrix, and validating using K-Fold Cross Validation. Research results using the Naïve Bayes method and Support Vector Machine for sentiment analysis on the Google Play website, the Naïve Bayes method produces 87.82% accuracy, 58.90% precision, 60.08% recall, while the Support Vector Machine method produces 90% accuracy .01% , precision 61.89%, recall 60.18%.

Sentiment analysis (a branch of Natural Language Processing) is one of the most crucial components of assessing online contents in current digital age. One of the leading competitors in the digital content market is YouTube, where thousands of new videos are uploaded every day. To assess the opinions of the users and customize advertisements and future contents accordingly, an appropriate sentiment analysis model is very crucial. Therefore, this study proposes a novel combination of deep learning technique called Encoder-decoder based Attention model to conduct sentiment analysis on YouTube reviews. In the proposed approach, the squeeze-and-excitation attention layer is utilized. Nearly 7,00,000 user comments from 8000 YouTube channels are utilized to train and test the proposed model in this study. Three different sentiments-positive, negative, and neutral—are assigned to the comments in the dataset for training purposes. The results show that the maximum accuracy of the model is 92.8% whereas maximum F1-score is 91.9%, surpassing several ML based state-of-the-art approaches. The evaluated metrics verify the balanced performance of the proposed approach. Moreover, the impact of the attention mechanism on the proposed approach is also assessed. Model accuracy for sentiment analysis is significantly improved by combining attention mechanism with encoder-decoder. As a generalized framework, this research can serve as a valuable guideline for future sentiment analysis on different languages.

Recommender systems gather user profile data either explicitly (users enter it) or implicitly (online behavior tracking).Surprisingly, given the prevalence of social media forums, which contain a rich set of user comments, there have been very few attempts to analyze the content of these comments to build up a user profile. In this paper, we compare and contrast a number of strategies for using text analysis to automatically gather profile data from user comments on news articles. We use this data to prototype a news recommender system based on the Guardian newspaper's 'Comment is Free' forum. The paper shows the feasibility of the approach: in a user study with fifty participants, our recommender outperforms a commercial 'best-in-class' system. Furthermore, we show that user comments allow recommender systems to track an evolving conversation related to a news article and can thus provide recommendations that better match the topics of conversation in comments, which maybe quite different from those in the original news article.