Category Archives: News

Tutorial on Extracting, Mining and Predicting Users’ Interests from Social Networks

SIGIR’19, July 21-25, Paris, France

KDD 2019

KDD’19, August 4-8, Anchorage, Alaska, USA

INTRODUCTION

The abundance of user generated content on social networks provides the opportunity to build models that are able to accurately and effectively extract, mine and predict users’ interests with the hopes of enabling more effective user engagement, better quality delivery of appropriate services and higher user satisfaction. While traditional methods for building user profiles relied on AI-based preference elicitation techniques that could have been considered to be intrusive and undesirable by the users, more recent advances are focused on a non-intrusive yet accurate way of determining users’ interests and preferences. In this tutorial, we will cover five important aspects related to the effective mining of user interests:

  1. The information sources that are used for extracting user interests
  2. Various types of user interest profiles that have been proposed in the literature
  3. Techniques that have been adopted or proposed for mining user interests
  4. The scalability and resource requirements of the state of the art methods
  5. The evaluation methodologies that are adopted in the literature for validating the appropriateness of the mined user interest profiles. We will also introduce existing challenges, open research question and exciting opportunities for further work.

 

SCHEDULE 

Session A [30 Minutes]: Background and Introduction to Theory of User Interest Mining

Session B [120 Minutes]: Techniques and Methods in User Interest Mining from Social Networks

Session C [30 Minutes]: Evaluation Methodologies, Future Directions and Open Challenges

 

MOTIVATION AND OVERVIEW

Mining user interests from user behavioral data is critical for applications such as online advertising. Based on user interests, service providers such as advertisers, can significantly reduce service delivery costs by offering the most relevant products (e.g., ads) to their customers. The challenge of accurately and efficiently identifying user interests has been the subject of increasing attention in the
past several years. Early approaches were based on explicit input from individuals about their own interests. To avoid the extra burden of manually filling in and maintaining interest profiles, most
methods in the past two decades have focused on the development of techniques that can automatically and unobtrusively determine users’ interests based on user behavioral data from data sources such as browsing history, page visits, the links they click on, the searches they perform and the topics they interact with. With the emergence and growing popularity of social networks such as blogging systems, wikis, social bookmarking, and microblogging services, many users are extensively engaged in at least some of these applications to express their feelings and views about a wide variety of social events/topics as they happen in real time by commenting, tagging, joining, sharing, liking, and publishing posts. This has made social networks an exciting and unique source of information about users’ interests. For instance, when looking at Twitter data during the first week of March 2019, the rivalry between the two English Premier League soccer clubs, Tottenham Hotspur and Arsenal, is a topic that has attracted a lot of discussion and interest. The development of techniques that can automatically detect such topics and model users’ interests towards them from online social networks would be highly important and have the potential to improve the quality of applications that work on a user modeling basis, such as filtering twitter streams, news recommendation and retweet prediction, among others. In this tutorial, we comprehensively introduce different strategies proposed in the literature, including our own work, for mining user interests from social networks with respect to the following five perspectives:

  1. Information Sources: The type of information sources used for extracting user interests from within social networks such as textual content (comments, #tags), social network structure, and
    images. Additionally, we will review external background knowledge sources such as semantic web resources and knowledge graphs that have been incorporated by some researchers
    to enhance the accuracy of user profiles.
  2. Profile Types: Most of works in user interest mining from social networks extract users’ explicit interests that are directly observable from user content. However, given the increasingly noticeable free-rider, some other techniques focus on passive users and extract their implicit interests by considering the interaction patterns between users and topics. There is another line of work that is dedicated to predicting users’ future interests instead of modeling current or past interests of users. These works are primarily focused on predicting if and which users would be interested in future topics on social networks. The accurate identification of users’ future interests on social networks allows one to perform future planning by studying how users will react if certain topics emerge in the future.
  3. Underlying Techniques: Previous methods have employed different techniques to build user profiles including neural embeddings, collaborative filtering, topic modeling, link prediction, regression, graph-based methods and Semantic Web technologies. We will review the techniques that have been used for identifying user interests and their different architectural variations.
  4. Scalability and Resource Requirements: Scalability is fundamental to user interest mining in order to accommodate torrents of social content. To this end, we provide a comprehensive overview of the speed-accuracy (efficiency-accuracy) trade-off when building user interest profiles for existing techniques of the literature. In particular, we present a critical review of those which scale to online vs. offline for massive streaming social content.
  5. Evaluation Methodology: Intrinsic vs. extrinsic evaluations are two main evaluation techniques, which have been widely adopted in the literature. Intrinsic evaluation helps to assess the quality of the constructed user interest profiles based on user studies while extrinsic evaluations measure the quality of the user interest profiles by looking at its impact on the effectiveness of other applications such as news recommendation and retweet prediction. We will review how each of these evaluation methodologies have been used in the literature.

 

PRESENTERS

Fattane ZarrinkalamDr. Fattane Zarrinkalam is a Postdoctoral Fellow at the Laboratory of Systems, Software and Semantics (LS3) at Ryerson University, where she works on projects related to Semantic-enabled Social Network Analysis. During her PhD studies, she focused on the identification of social media users’ interests based on their individual and collective behavior on social networks especially Twitter. She has published her work in venues such as CIKM, ESWC and ECIR. In addition, she has published journal papers in premier journals including Information Retrieval and Information Processing and Management. Further, during her PhD, she was involved in two patent applications that were filed with USPTO.

hosseinHossein Fani is a PhD student at the University of New Brunswick and research assistant at the Laboratory of Systems, Software and Semantics (LS3) at Ryerson University, Canada. Hossein has worked in the broad area of Social Network Analytics with special attention to content-based and temporal user community identification. Hossein has extensively published during his PhD studies in venues such as CIKM, ECIR and WSDM. His peer-reviewed journal publications also appear in Wiley’s Computational Intelligence and Springer’s Social Network Analysis and Mining. His PhD work has resulted in a provisional patent with USPTO. He has also reviewed for conferences such as ECIR and NAACL.

Dr. Ebrahim BagheriDr. Ebrahim Bagheri is an Associate Professor and the Director for the Laboratory for Systems, Software and Semantics (LS3) at Ryerson University. He also holds a Canada Research Chair (Tier II) in Software and Semantic Computing as well as an NSERC Industrial Research Chair in Social Media Analytics. He has been PI on projects worth over $8M funded by partners such as NSERC, AIF and IBM. Most recently in 2018, he was the Program Committee co-Chair for the Canadian Conference on Artificial Intelligence and also the Industry Program Committee co-Chair at IEEE/ACM International Conference on  ASONAM and an Area Chair for NAACL-HLT 2019. He also serves on the Program Committee of venues such as RecSys and ICWSM as well guest-editor for international journals such as Information Systems and Information Processing and Management

 

Special Issue on Mining Social Influence and Actionable Insights from Social Networks

Special Issue on Mining Social Influence and Actionable Insights from Social Networks
Elsevier’s Information Processing and Management Journal

AIM AND SCOPE
In the last 10 years, the dissemination and use of social networks have grown significantly worldwide. Online social networks have billions of users and are able to record hundreds of data from each of its users. The wide adoption of social networks resulted in an ocean of data which presents an interesting opportunity for performing data mining and knowledge discovery in a real-world context. The enormity and high variance of the information that propagates through large user communities influences the public discourse in society and sets trends and agendas in topics that range from marketing, education, business and medicine to politics, technology and the entertainment industry. Mining the contents of social networks provides an opportunity to discover social structure characteristics, analyze action patterns qualitatively and quantitatively, and gives the ability to predict future events. In recent years, decision makers have become savvy about how to translate social data into actionable information in order to leverage them for a competitive edge. Moreover, social networks expose different aspects of the social behavior of its users. In this respect, many users of the social networks are known as influencers. The influencers are users that usually publish their opinions about different topics, products and services on the social networks, and then affect intentionally or unintentionally the opinions, emotions, or behaviors of other users on the social networks. Because of the high impact of influencers on the opinions and behaviors of other users, many companies and organizations are interested in discovering influencers on social networks to increase the promotion and sale of their products and services. However, the discovering of influencers on social networks is a really complex problem that requires developing models, techniques and algorithms for an appropriate analysis.

Traditional research in social network mining mainly focuses on theories and methodologies for community discovery, pattern detection and evolution, behavioural analysis and anomaly (misbehaviour) detection. While interesting and definitely worthwhile, the main distinguishing focus of this joint workshop will be the use of social network data for building predictive models that can be used to uncover hidden and unexpected aspects of user-generated content in order to extract actionable insights from them and for analyzing different aspects of social influence, such as influence maximization and discovering influencers. Thus, the focus is on algorithms and methods for (social) network analysis, data mining techniques to gain actionable real-world insights, and models and approaches for understanding influence dissemination and discovering influential users in social networks.

In this special issue, we solicit manuscripts from researchers and practitioners, both from academia and industry, from different disciplines such as computer science, data mining, machine learning, network science, social network analysis and other related areas to share their ideas and research achievements in order to deliver technology and solutions for mining actionable insight from social network data.

TOPICS OF INTEREST
We solicit original, unpublished and innovative research work on all aspects around, but not limited to, the following themes:

  • Predictive modeling based on social networks such as

o Box office prediction
o Election prediction
o Flu prediction

  • Product adaptation models with social networks such as

o Sale price prediction
o New product popularity prediction
o Brand popularity
o Business downfall prediction

  • User modeling and social networks including

o Predict users daily activities including recurring events
o User churn prediction
o Determining user similarities, trustworthiness and reliability

  • Social networks and information/knowledge dissemination

o Topic and trend prediction
o Prediction of information diffusion patterns
o Identification of causality and correlation between event/topics/communities

  • Social network analysis and measures

o Network topology
o Centrality measures
o Community detection
o Dynamic network models
o Diffusion models

  • Information diffusion modeling with social networks

o Information propagation and assimilation in social networks
o Sentiment diffusion in social networks
o Competitive intelligence from social networks

  • Social influence analysis on online social networks

o Systems and algorithms for discovering influential users
o Recommending influential users in online social networks
o Social influence maximization
o Modeling social networks and behavior for discovering influential users
o Discovering influencers for advertising and viral marketing in social networks
o Decision support systems and influencer discovering

  • Trust and reputation in social networks
  • Merging internal (proprietary) data with social data
  • Feature Engineering from Social Networks
  • Datasets and Evaluation methodologies for predictive modeling in social networks

IMPORTANT DATES
* Submission deadline: October 15, 2018 Nov 15, 2018
* First Notification: Jan 1, 2019 Feb 1, 2019
* Revisions Due: Feb 1, 2019 March 1, 2019
* Final Notification: April 1, 2019 May 1, 2019

GUEST EDITORS
• Marcelo G. Armentano, ISISTAN Research Institute (CONICET- UNICEN), Argentina
• Ebrahim Bagheri, Ryerson University, Canada
• Frank Takes, University of Amsterdam, The Netherlands
• Virginia D. Yannibelli, ISISTAN Research Institute (CONICET- UNICEN), Argentina

Paper Submission Details
Papers submitted to this special issue for possible publication must be original and must not be under consideration for publication in any other journal or conference. Previously published or accepted conference papers must contain at least 30% new material to be considered for the special issue.

All papers are to be submitted through the journal editorial submission system (https://www.evise.com/profile/#/IPM/login). At the beginning of the submission process in the submission system, authors need to select “Mining Social Influence and Actionable Insights from Social Networks” as the article type. All manuscripts must be prepared according to the journal publication guidelines which can also be found on the website provided above. Papers will be evaluated following the journal’s standard review process.

UMLS to DBpedia Mapping

We are pleased to announce our work on mapping between UMLS and DBpedia concepts has been published in the “Journal of the American Medical Informatics Association” (https://doi.org/10.1093/jamia/ocy021) with the full dataset, online browser, and RESTful API available at http://denote.rnet.ryerson.ca/umlsMap/. This dataset includes 221 690 “exact match” relations, 26 276 “close match”, and 6 784 322 “see also” pairings. We hope you find this useful in facilitating effective interchange and integration between general knowledge databases (DBpedia) and those from (bio)medical ontologies (UMLS).

Our team won the CIKM Cup 2017!
To stand out from the crowd, sellers employ creative, sometimes disruptive titles for their products in online stores to improve their search relevancy or attract the attention of customers. As a part of the CIKM AnalytiCup 2017, the challenge is to build a product title quality model that can automatically grade the clarity and the conciseness of a product title. Our proposed “Bagging Model for Product Title Quality with Noise” could leave others behind in performance and become the winner of the CIKM Cup 2017 competition.

Check out our approach for details:[Code][Report]

The Ambivalence of Social Health Information @TEDxRyersonU

The Ambivalence of Social Health Information

https://www.youtube.com/watch?v=kS_TjzOQtNs

A recent study by PwC Health Research Institute reported that 90% of the participating millennials in the study said that they trust health-related information on social networks. Also, forty percent of the study participants reported that the information they find on social media shapes their approach to diet and exercise, and even their choice of physician. In this talk, we will explore how much impact social media and networks have on public’s health and medical related perception and behaviour.

Ebrahim Bagheri is a Canada Research Chair in Software and Semantic Computing, an NSERC Industrial Research Chair in Social Media Analytics and an Associate Professor in Electrical and Computer Engineering at Ryerson University. His work has produced leading computer software technologies that help machines understand the meaning and make sense of user-generated content, such as social network and media information. His research explores, from a computational perspective, how users’ interests are shaped on social networks, how communities are formed and can be detected and how opinions are expressed and propagated.

Ebrahim Bagheri has been awarded the prestigious NSERC Industrial Research Chair in Social Media Analytics.

Ebrahim Bagheri has been awarded the prestigious NSERC Industrial Research Chair in Social Media Analytics.

The motivation for the establishment of this chair, valued at $1.5 million over five years, is to develop a world-class research program and increased research capacity in the area of social and semantic technologies, and particularly their application to derive actionable insight through social media analytics.

According to Natural Sciences and Engineering Research Council of Canada, Industrial Research Chairs (IRC) are intended to:

  • assist universities in building on existing strengths to achieve the critical mass required for a major research endeavour in natural sciences and engineering of interest to industry; and/or
  • assist in the development of research efforts in fields that have not yet been developed in Canadian universities but for which there is an important industrial need; and

  • provide an enhanced training environment for graduate students and, where appropriate, postdoctoral fellows by exposing them to research challenges unique to industry and the opportunity for significant ongoing interactions with the industrial partner(s).

Postdoc/Research Associate in Web Search/Social Network Analytics

Laboratory for Systems, Software and Semantics (LS3)
Ryerson University, Toronto, Canada

http://ls3.rnet.ryerson.ca/

Post-Doctoral or Research Associate Position

We are seeking a Postdoctoral Fellow or Research Associate to work on social media/network analytics and/or web search. The ideal candidate will have excellent research capabilities as well as competitive development skills.

The candidate should have expertise in data mining, machine learning, social network analytics or web search. A PhD in Computer Science or related fields is required. Expertise with NoSQL technology, Java, Javascript and frameworks for distributed processing in Java (e.g. Apache Spark, Hadoop MapReduce) will be an asset.

Salary and benefits are competitive and commensurate with expertise. The position is for one year with the possibility of extension.

Further inquiries may be directed to Dr. Ebrahim Bagheri (bagheri@ryerson.ca).
Interested candidates should submit an application letter, a résumé, three publications relevant to the post-doctoral project and contact information of three reference persons to Dr. Ebrahim Bagheri (bagheri@ryerson.ca ). Candidates will be screened until the position is filled.

The Laboratory for Systems, Software and Semantics (LS3) is a part of the Faculty of Engineering, and Architectural Sciences at Ryerson University. Ryerson University is a public research university located in downtown Toronto, Ontario, Canada. Its urban campus surrounds the Yonge-Dundas Square, located at the busiest intersection in downtown Toronto. The university is composed of 36,000+ undergraduate students, 2,000+ graduate students, and 70,000 yearly certificate and continuing education registrations. Ryerson has been one of the fastest growing research institutions in Canada.

Ryerson University values diversity and is committed to equal opportunity in employment.

International Workshop on Mining Actionable Insights from Social Networks

International Workshop on Mining Actionable Insights from Social Networks
MAISoN 2017: http://ls3.rnet.ryerson.ca/MAISoN/2017/

Co-located with Tenth ACM International Web Search and Data Mining (WSDM) Conference in Cambridge, UK February 6-10, 2017.
http://www.wsdm-conference.org/2017/

The wide adoption of social network churns out an ocean of data which presents an interesting opportunity for mining the data and discover new knowledge to predict real-world outcomes. The enormity and high variance of the information that propagates through large user communities influence the public discourse in society and sets trends and agendas in topics that range from marketing, education, business and medicine to politics, technology and the entertainment industry. Mining the attributes and contents of social network provides an opportunity to discover social structure characteristics, analyze action patterns qualitatively and quantitatively, and the ability to predict future events. In recent years, decision makers have become savvy about how to translate social data into actionable information in order to leverage them for a competitive edge. In particular, marketers aggregate the opinions of the collective population to dynamically calibrate, anticipate and offer products and services that meet perpetually shifting consumer demands in a hyper-competitive marketplace. The traditional research in social network mainly focus on theories and methodologies on community discovery, pattern detection and evolution, behavioural analysis and anomaly and misbehaviour detection. The main distinguishing focus of this workshop will be the use of social media data for building predictive models that can be used to uncover hidden and unexpected aspects of user-generated content in order to extract actionable insight from them. The objectives will be to transform the insight into effective actions which could help organizations improve and refine their strategies.

In this workshop, we invite researchers and practitioners from different disciplines such as computer science, big data mining, machine learning, social network analysis and other related areas to share their ideas and research achievements in order to deliver technology and solutions for mining actionable insight from social network data.

Topics
We solicit original, unpublished and innovative research work on all aspects around, but not limited to, the following themes:
• Applications of Social Network Analysis
• Predictive modeling based on social networks such as
◦ Box office prediction
◦ Election prediction
◦ Flu prediction
• Product adaptation models with social networks such as
◦ Sale price prediction
◦ New product popularity prediction
◦ Brand popularity
◦ Business downfall prediction
• User modeling and social networks including
◦ Predict users daily activities including recurring events
◦ User churn prediction
◦ Determining user similarities, trustworthiness and reliability
• Social networks and information/knowledge dissemination
◦ Topic and trend prediction
◦ Prediction of information diffusion patterns
◦ Identification of causality and correlation between event/topics/communities
• Information diffusion modeling with social networks
◦ Sentiment diffusion in social network
◦ Competitive intelligence from social networks
• Merging internal (proprietary) data with social data
• Search Behaviour Analytics with Social Networks
• Feature Engineering from Social Networks
• Social Networks and Recommender Systems
• Sentiment Analysis and Prediction on Social Networks
• Datasets and Evaluation methodologies for predictive modeling in social networks

Format and Submission
We invite the submission of regular research papers (6-8 pages) as well as position papers (2-4 pages). We recommend papers to be formatted according to the standard double-column ACM Guidelines. All papers will be peer reviewed by three reviewers. 
All submissions must be submitted in PDF format according to the guidelines through the Easychair installation: https://easychair.org/conferences/?conf=maison2017

Important Dates
• Submission: November 10, 2016
• Decisions: December 5, 2016
Submissions will be due Midnight AoE time.

Organizers
• Ebrahim Bagheri, Ryerson University
• Zeinab Noorian, Ryerson University
• Faezeh Ensan, Ferdowsi University of Mashhad

Ebrahim Bagheri awarded Ontario’s Engineering Medal

Dr. Ebrahim Bagheri has received Ontario Professional Engineers, Engineering Medal – Young Engineer (2016), for exceptional achievements in career, community and professional participation before the age of 35. (www.ospe.on.ca)

LS3 director awarded a Tier 2 Canada Research Chair (CRC) in Software and Semantic Computing

Ebrahim Bagheri awarded a Tier 2 Canada Research Chair (CRC) in Software and Semantic Computing.

Dr. Bagheri is developing data analytics software methodologies and platforms that will help to understand and analyze the large volume of user-generated data created in recent years. Through his research, he wants to give software engineers and data scientists the tools and techniques they need to create software applications that can handle the demands of this growing industry.
Announcement Link: http://www.ee.ryerson.ca/news/#BAGHERICRC