Assessing the Contribution of Twitter's Textual Information to Graph-based Recommendation
Published 7 March 2017
Evgenia Wasserman Pritsker, Tsvi Kuflik, Einat Minkov
Citations10
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
An extended graph representation that includes socio-demographic and personal traits extracted from the content posted by the user on SM is proposed, demonstrating that processing unstructured textual information collected from Twitter and representing it in structured form in the graph improves recommendation performance.
Abstract
Graph-based recommendation approaches can model associations between users and items alongside additional contextual information. Recent studies demonstrated that representing features extracted from social media (SM) auxiliary data, like friendships, jointly with traditional users/items ratings in the graph, contribute to recommendation accuracy. In this work, we take a step further and propose an extended graph representation that includes socio-demographic and personal traits extracted from the content posted by the user on SM. Empirical results demonstrate that processing unstructured textual information collected from Twitter and representing it in structured form in the graph improves recommendation performance, especially in cold start conditions.
Keywords
Computer Science
Journal of Machine Learning ResearchLatent dirichlet allocation
27,049 Citations2003David M. Blei, Andrew Y. Ng +1 more
Collaborative Filtering for Implicit Feedback Datasets
3,167 Citations2008Yifan Hu, Yehuda Koren +1 more
This work identifies unique properties of implicit feedback datasets and proposes treating the data as indication of positive and negative preference associated with vastly varying confidence levels, which leads to a factor model which is especially tailored for implicit feedback recommenders.
PLoS ONEPersonality, Gender, and Age in the Language of Social Media: The Open-Vocabulary Approach
1,735 Citations2013H. Andrew Schwartz, Johannes C. Eichstaedt +9 more
This represents the largest study, by an order of magnitude, of language and personality, and found striking variations in language with personality, gender, and age.
Proceedings of the International AAAI Conference on Web and Social MediaPredicting Depression via Social Media
1,567 Citations2021Munmun De Choudhury, Michael Gamon +2 more
It is found that social media contains useful signals for characterizing the onset of depression in individuals, as measured through decrease in social activity, raised negative affect, highly clustered egonetworks, heightened relational and medicinal concerns, and greater expression of religious involvement.
IEEE Transactions on Knowledge and Data EngineeringTopic-sensitive pagerank: A context-sensitive ranking algorithm for web search
1,065 Citations2003Taher H. Haveliwala
It is shown that using linear combinations of these (precomputed) biased PageRank vectors to generate context-specific importance scores for pages at query time, can generate more accurate rankings than with a single, generic PageRank vector.
One-Class Collaborative Filtering
1,012 Citations2008Rong Pan, Y. Zhou +5 more
This paper considers the one-class problem under the CF setting, and proposes two frameworks to tackle OCCF, one based on weighted low rank approximation; the other based on negative example sampling.
Classifying latent user attributes in twitter
650 Citations2010Delip Rao, David Yarowsky +2 more
A novel investigation of stacked-SVM-based classification algorithms over a rich set of original features, applied to classifying these four user attributes, as distinct from the other primarily spoken genres previously studied in the user-property classification literature.
On social networks and collaborative recommendation
610 Citations2009Ioannis Konstas, Vassilios Stathopoulos +1 more
This work created a collaborative recommendation system that effectively adapts to the personal information needs of each user, and adopts the generic framework of Random Walk with Restarts in order to provide with a more natural and efficient way to represent social networks.
Predicting Personality from Twitter
604 Citations2011Jennifer Golbeck, Cristina Robles +2 more
This paper presents a method by which a user's personality can be accurately predicted through the publicly available information on their Twitter profile, and the implications this has for social media design, interface design, and broader domains.
Recommending twitter users to follow using content and collaborative filtering approaches
479 Citations2010John Hannon, Mike Bennett +1 more
A range of different profiling and recommendation strategies are evaluated, based on a large dataset of Twitter users and their tweets, to demonstrate the potential for effective and efficient followee recommendation.
Video suggestion and discovery for youtube
444 Citations2008Shumeet Baluja, Rohan Seth +6 more
A novel method based upon the analysis of the entire user-video graph to provide personalized video suggestions for users, termed Adsorption, provides a simple method to efficiently propagate preference information through a variety of graphs.
Short and tweet
413 Citations2010Jilin Chen, Rowan Nairn +3 more
This paper studied content recommendation on Twitter to better direct user attention and explored three separate dimensions in designing such a recommender: content sources, topic interest models for users, and social voting.
Computational IntelligenceUsing Hashtags to Capture Fine Emotion Categories from Tweets
412 Citations2014Saif M. Mohammad, Svetlana Kiritchenko
It is shown that emotion‐word hashtags are good manual labels of emotions in tweets and a method to generate a large lexicon of word–emotion associations from this emotion‐labeled tweet corpus is proposed, which is the first lexicon with real‐valued word‐emotion association scores.
International Joint Conference on Artificial IntelligenceItemRank: a random-walk based scoring algorithm for recommender engines
271 Citations2007Marco Gori, Augusto Pucci
This paper presents ItemRank, a random-walk based scoring algorithm, which can be used to rank products according to expected user preferences, in order to recommend top-rank items to potentially interested users.
Developing Age and Gender Predictive Lexica over Social Media
236 Citations2014Maarten Sap, Gregory Park +6 more
Predictive lexica (words and weights) for age and gender using regression and classification models from word usage in Facebook, blog, and Twitter data with associated demographic labels achieve state-of-the-art accuracy.
Addressing cold-start in app recommendation
171 Citations2013Jovian Lin, Kazunari Sugiyama +2 more
This paper describes a method that accounts for nascent information culled from Twitter to provide relevant recommendation in such cold-start situations and significantly outperforms other state-of-the-art recommendation techniques by up to 33%.
Proceedings of the AAAI Conference on Artificial IntelligenceInferring Latent User Properties from Texts Published in Social Media
163 Citations2015Svitlana Volkova, Yoram Bachrach +2 more
This work examines individual tweets to detect emotions and opinions emanating from them, and analyzes all the tweets published by a user to infer latent traits of that individual, focusing on Ekman’s six emotions.
Enhancing collaborative filtering systems with personality information
161 Citations2011Rong Hu, Pearl Pu
The results show that the proposed CF variations, which consider personality characteristics, can significantly improve the performance of the traditional rating-based CF in terms of the evaluation metrics MAE and ROC sensitivity.
Predicting purchase behaviors from social media
137 Citations2013Yongzheng Zhang, Marco Pennacchiotti
The paper provides an extensive analysis on how users' Facebook profile information correlates to purchases on eBay, and analyzes the performance of different feature sets and learning algorithms on the task of purchase behavior prediction.
User Modeling and User-Adapted InteractionFacebook single and cross domain data for recommendation systems
130 Citations2012Bracha Shapira, Lior Rokach +1 more
This paper investigates the feasibility and effectiveness of utilizing existing available data from social networks for the recommendation process, specifically from Facebook, and highlights the benefits of utilizing cross domain Facebook data to achieve improvement in recommendation performance.
User Modeling and User-Adapted InteractionAlleviating the new user problem in collaborative filtering by exploiting personality information
127 Citations2016Ignacio Fernández-Tobías, Matthias Braunhofer +3 more
Ghent University Academic Bibliography (Ghent University)MovieTweetings: a movie rating dataset collected from twitter
121 Citations2013Simon Dooms, Toon De Pessemier +1 more
This work taps into the vast availability of social media and construct a new movie rating dataset ‘MovieTweetings’ based on public and well-structured tweets and believes this dataset can show to be very useful as an always up-to-date and natural rating dataset for movie recommenders.
A general graph-based model for recommendation in event-based social networks
120 Citations2015Tuan-Anh Nguyen Pham, Xutao Li +2 more
This paper proposes a general graph-based model, called HeteRS, to solve the three recommendation problems on EBSNs in one framework, and proposes a learning scheme to set the influence weights between different types of entities.
Weakly Supervised User Profile Extraction from Twitter
118 Citations2014Jiwei Li, Alan Ritter +1 more
This paper presents a weaklysupervised approach to user profile extraction from Twitter that takes into account network information, a unique opportunity offered by social media.
Proceedings of the International AAAI Conference on Web and Social MediaHappy, Nervous or Surprised? Classification of Human Affective States in Social Media
69 Citations2021Munmun De Choudhury, Michael Gamon +1 more
The findings indicate that different types of affect involve different emotional content and usage styles; hence the performance of the classifier on various affects can differ considerably.
Semantics-aware Graph-based Recommender Systems Exploiting Linked Open Data
49 Citations2016Cataldo Musto, Pasquale Lops +3 more
A graph-based algorithm leveraging LOD-based features is able to overcome several state of the art baselines, such as collaborative filtering and matrix factorization, thus confirming the effectiveness of the proposed approach.
Addressing the New User Problem with a Personality Based User Similarity Measure
44 Citations2011Marko Tkalčič
A personality based USM was used that yielded significantly better recommender performance in the period where the new user problem occurs and a method for establishing the boundary of the cold start period was presented.
Cross social networks interests predictions based ongraph features
30 Citations2013Amit Tiroshi, Shlomo Berkovsky +3 more
An exploratory study of the process of making recommendations using a unique multi-network dataset containing user interests across multiple domains, e.g., music, books, and movies, and assess the contribution of various network- and domain-related features to the accuracy of the recommendations.
Journal of the Association for Information Science and TechnologyGraph‐based recommendation integrating rating history and domain knowledge: Application to on‐site guidance of museum visitors
25 Citations2017Einat Minkov, Keren Kahanov +1 more
A personalized graph‐based recommender framework is proposed, representing rating history and background multi‐facet information jointly as a relational graph, and a random walk measure is applied to rank available complementary multimedia presentations by their relevancy to a visitor's profile, integrating the various dimensions.
ACM Transactions on Information SystemsImproving graph-walk-based similarity with reranking
23 Citations2010Einat Minkov, William W. Cohen
High performance, together with practical runtimes, suggest that the described framework is a useful search system in the PIM domain, as well as in other semistructured domains.
Improving business rating predictions using graph based features
19 Citations2014Amit Tiroshi, Shlomo Berkovsky +4 more
This work proposes to leverage graph based representation of the data in order to generate and automatically populate features that augment the original user-item ratings data, and evaluates the proposed methodology using the recently released Yelp business ratings dataset.
Lecture notes in computer scienceGraph-Based Recommendations: Make the Most Out of Social Data
11 Citations2014Amit Tiroshi, Shlomo Berkovsky +3 more
The evaluation shows that the social auxiliary data improves the accuracy of the recommendations, and that the greatest improvement is achieved when graph features mirroring the nature of the auxiliary data are extracted by the recommender.
Graph-based Recommendation in the Museum.
3 Citations2014Tsvi Kuflik, Einat Minkov +1 more
This work presents how graph based techniques may enable combining visitors preferences and contextual aspects for recommendations in mobile Cultural Heritage.