Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
19 results
Search Results
Item Analysis of stock prices and its associated volume of social network activity(Springer Science and Business Media Deutschland GmbH, 2018) Kinnal, K.; Sanjeev, K.V.; Chandrasekaran, K.Stock market prediction has been a convenient testing ground and a highly cited example for applying machine learning techniques to real-life scenarios. However, most of these problems using twitter feeds to analyze stock market prices make use of techniques such as sentimental analysis, mood scoring, financial behavioral analysis, and other such similar methods. In this paper, we propose to discover a correlation between the stock market prices and their associated twitter activity. It is always observed that whenever there are spikes in the stock market prices, there is a preceding twitter activity indicating the imminent spike in the aforementioned stock price. Our objective is to discover this existence of a correlation between the volumes of tweets observed when a market indices’ stock price spikes and the amount by which the stock price changes, with the help of machine learning techniques. If this correlation does exist, then an attempt is made to figure out a mathematical relationship between the two factors. © Springer Nature Singapore Pte Ltd. 2018.Item An iterative MapReduce framework for sports-based tweet clustering(Association for Computing Machinery acmhelp@acm.org, 2015) Saxena, G.; Santurkar, S.In recent years, social media has evolved into a vital source for real-time information. Sports is one of the most popular topics on social media and attracts the attention of users all over the world. However, a large amount of data is generated on a daily basis, making it difficult for the fans to follow the topics of their interest. Clustering of these posts can resolve this issue by retrieving unambiguous and distinct topics. MapReduce is a programming paradigm that is very effective in designing distributed applications that can be deployed on the cloud. Clustering algorithms are generally iterative in nature. The performance gain offered by MapReduce cannot be completely realized by these algorithms due to the inherent architectural bottlenecks associated with iterative tasks. Twister is a MapReduce-based framework designed to minimize these bottlenecks. In this paper, we propose a distributed framework that gathers sports-related tweets and clusters them into distinct topics using the DB-SCAN algorithm customized for Twister. The accuracy of the framework was analysed using the precision-recall scoring mechanism to determine the set of DBSCAN and framework parameters that result in the best set of clusters. The performance of our framework is evaluated based on our clustering results and simulations using the MRSim simulator. We expect that this framework could be used as a model for performing topic detection over generic tweets. We have used the domain of sports to establish the proof of this concept. © 2015 ACM.Item Automatic detection of rumoured tweets and finding its origin(Institute of Electrical and Electronics Engineers Inc., 2016) Sahana, V.P.; Pias, A.R.; Shastri, R.; Mandloi, S.The number of people using social networking sites such as Twitter is exponentially increasing every day. These sites not only act as a platform for staying connected with friends and exchanging opinions and ideologies, but also help to share and disseminate information. Important events in the real world cause a huge flood of tweets on Twitter. During many such events, Twitter has been used to spread rumours and cause havoc and panic among the people, thus worsening the situation. In this project, we try to automatically detect the rumours spreading on Twitter and identify its source. Using some of the rumoured tweets posted during the London Riots in 2011 and some non-rumoured tweets, we trained a classifier. Our classifier correctly classifies the tweets with high accuracy. We show that, in rumour detection, the information propagated by the user becomes more important than the identity of the user by showing that tweet-based features play a much higher and significant role than user-based features. We propose an algorithm to find the origin of the rumoured tweets i.e. obtain the account information of the user who first started spreading rumours on Twitter. © 2015 IEEE.Item Twitter-user recommender system using tweets: A content-based approach(Institute of Electrical and Electronics Engineers Inc., 2017) Nidhi, R.H.; Annappa, B.With the advent of the internet into our everyday lives, online social networks such as Facebook and Twitter have taken up a major role in networking, information deployment and entertainment. As of 2017, Twitter's outreach is over 317M monthly active users generating more than 320M tweets every day, thus making it one of the fastest information deployment mediums of this era. In order to aid data distribution without causing a glut of information to the users, we develop a recommender system focusing on a vital aspect of social media - relationships among users, addressing a popular problem of users - who to follow/befriend? By choosing the right accounts and users to follow, the sources of information can be controlled as desired. The information collected from the most recent tweets of a user is used to find other users whose recent tweets contain similar information, ensuring there is at least one mutual friend among users. By making use of the continuous and real time updating of data on social networks, we develop a method to ensure our training sets consist of relevant information for classification, thus preserving accuracy while reducing training set sizes for probabilistic learning models. We use two algorithms to detect tweets of common topics, namely a Noun Phrase detector and a Naïve Bayes Text (Topic) Classifier and further compare their complexity and accuracy. The Naive Bayes Classifier, despite being probabilistic, functioned well with a relatively small training set. This is only with the exception of Twitter as it is a real-time updating framework. Exact matches were hard to obtain with the Noun phrase detector, as we are going only one level deep due to limited compute. However, when matches were found, it is upto 90% accurate. Experiments on tweets of random public users have found that Naive Bayes Classifier with a small but recent training data set can work as well as or better than a Collaborative filter without the assumptions of the Collaborative model. © 2017 IEEE.Item Friendship recommendation system using topological structure of social networks(Springer Verlag service@springer.de, 2018) Kumar, P.; Guddeti, G.Popularity and importance of Recommendation System is being increased day by day in both commercial and research community. Social networks (SNs) like Facebook, Twitter, and LinkedIn draw more attention since without any previous knowledge a lot of connections have been established. The creation of relationship between users is the key feature of a social network. Therefore, it is important for researchers to look for a new way to provide recommendations with more relevance. This paper proposes two algorithms for recommending a new friend in online social networks. The first algorithm is based on the number of mutual friends and second is based on influence score. These recommendation algorithms use collaborative filtering and provide the idea of doing recommendations (e.g., Facebook recommend friends, Netflix suggest movies, Amazon recommend products, etc.). Obtained results and analysis indicate that influence-based recommendation system is more accurate as compared to mutual friend-based recommendation. These proposed recommendation algorithms can be used for the development of an effective social networking or e-commerce site and thereby providing a better experience to users. © Springer Nature Singapore Pte Ltd. 2018.Item A Bag-of-Phonetic-Codes Modelfor Cyber-Bullying Detection in Twitter(Institute of Electrical and Electronics Engineers Inc., 2018) Shekhar, A.; Venkatesan, M.Social networking sites such as Twitter, Facebook, MySpace, Instagram are emerging as a strong medium of communication these days. These have become a part and parcel of daily life. People can express their thoughts and activities among their social circle with brings them closer to their community. However this freedom of expression has its drawbacks. Sometimes people show their aggression on Social Media which in turn hurts the sentiments of the targeted victims. Certain forms of cyber-bullying are sexual, racial and physical disability based. Hence a proper surveillance is necessary to tackle such situations. Twitter as a micro-blogging site sees cyber abuse on a daily basis. However, tweets are raw texts; containing a lot of misspelled words and censored words. This paper proposes a novel method to detect cyber-bullying, a Bag-of-Phonetic-Codes model. Using pronunciation of words as features can rectify misspelled words and can identify censored words. Correctly identifying duplicate words can lead to smaller vocabulary of words, thereby reducing the feature space. The inspiration for this proposed work is drawn from the famous Bag-of-Words model for extracting textual features. Phonetic code generation has been done using the Soundex Algorithm. Besides the proposed model, experiments were carried out with both supervised and unsupervised machine learning approaches on multiple datasets to understand the approaches and challenges in the domain of cyber-bullying detection. © 2018 IEEE.Item Predicting Influenza Outbreak using Constrained Static and dynamic Feature(Institute of Electrical and Electronics Engineers Inc., 2018) Dofadar, S.; Venkatesan, M.Twitter is a free social networking and micro-blogging service that gives the opportunity to write and read each others tweet to its 330 million users all over the world with a limitation of 280 characters in each tweet. As a result, Twitter can provide a huge amount of data regarding what is happening at a particular time in all over the world. One of those is epidemic event detection and prediction from the twitter data. In this study, the use of Twitter data to detect influenza outbreak is examined. The result from this experiment shows that estimate of influenza outbreak can be derived from twitter correctly combining constrained supervised and unsupervised features and then using a prediction model. © 2018 IEEE.Item Trigger Event and Hate Content: Insights from Twitter Analytics(Institute of Electrical and Electronics Engineers Inc., 2019) Chetty, N.; Alathur, S.The problematic act damages the target and seeds the fear in the neighborhood. Social media sites are used for planning and coordinating problematic acts. The problematic act is a trigger event which influences hatred feeling. The objective of the paper is to analyze the aftermath of a recent problematic incident in the southern part of the Asian continent from Twitter content. After the problematic incident, citizens used to share their views over social media sites. A total of 48,819 opinions shared through Twitter social media are collected and analyzed using the software developed in the R programming language. The results show hatred against the problematic act through different related emotions. Results also contain more negative tweets which are almost thrice the positive tweets. Fear and anger emotions exhibit a high degree of emotions than the other. © 2019 IEEE.Item Developing Indian Smart Cities: Insights from Social Media(Springer Science and Business Media Deutschland GmbH, 2020) Chetty, N.; Alathur, S.Smart cities play an important role in overall development of a nation by progressing with economic, environmental and social domains. India has projected to create 100 smart cities in near future. The purpose of the paper is to identify the key influencing components and the social media users’ expectations for smart cities development in India. The Twitter social media content of smart cities council and the user posts on smart cities are collected through Twitter application programming interface. The collected tweets are cleaned by pre-processing methods and analyzed for insights. Technology, infrastructure, innovation, transport, mobility and management are the key influencing components for smart cities development in India. The social media users are expecting to emphasize on combating the issues like Covid-19 and use of IoT technology for the success of smart cities project. The integration of different components could increase the success of the project. The analysis of the content shared by the groups (smart cities council and the social media users) which are at different sides of smart cities’ development project, increases the novelty of the study. © 2020, IFIP International Federation for Information Processing.Item Climate Change and COVID-19 Metaphors: Environmental Consciousness in Social Media(Institute of Electrical and Electronics Engineers Inc., 2021) Chetty, N.; Alathur, S.The changes in environmental conditions have reportable impacts on the Covid-19 pandemic and vice versa. About these impacts, the people are deliberating metaphorically on climate change and Covid-19. The purpose of the paper is to analyze and identify the people's consciousness about the interrelationship between climate change and Covid-19. The existing literature on climate change and impacts on Covid-19 are reviewed, and inferences have been drawn from the result. Apart from the literature review, the Twitter social media content is analyzed and interpreted. The results have shown that there are noticeable impacts of climate change on Covid-19 and vice-versa. The people are conscious of the relationship between climate change and the Covid-19 pandemic. © 2021 IEEE.