2022-07-24: ACM Conference on Hypertext and Social Media Trip Report

By -



The ACM Conference on Hypertext and Social Media provided an opportunity for speakers and attendees to participate in an onsite-online hybrid model for the first time in history. The conference also ensured effective interaction between on-site participants and online participants. The conference was conducted for four days between June 28 – July 1, 2022. This year HT `22 was co-located with ACM WebSci 2022 and ACM UMAP 2022. I mainly attended the paper presentations (technical sessions) on Thursday and Friday but missed the blue sky ideas session. I continuously monitored tweets related to the conference using #HT2022, and #hypertext hashtags.

This year, our paper "Enabling Convenient Online Collaborative Writing for Low Vision Screen Magnifier Users" was accepted and the presentation was conducted on Friday. All the papers and videos submitted by the authors are available in @ACMDL and can be downloaded for free for a limited time.

Workshops 

1) HUMAN’22

HUMAN’22 has a strong focus on the user. Its user-centric view on hypertext not only includes user interfaces and interaction, but also discussions about hypertext application domains.
The main intent of this workshop was to explore different hypertext research from the viewpoint of human factors. The workshop also focused on combining original hypertext research ideas with recent hypertext research workflow.
 

2) Narrative and Hypertext 2022

The Narrative and Hypertext (NHT) workshops aim to provide an interdisciplinary forum to bring together individuals from the humanities and technological communities to share work and discuss state-of-the-art research on narrative from both a technical and aesthetic perspective.

Participants were asked to provide a short paper on the work that they are currently focusing on. This was a half-day virtual workshop where participants presented their papers and the workshop also had a detailed discussion and Q&A session. 

3) OASIS’22

The main purpose of conducting this workshop is to be able to identify relevant problems in centralized and decentralized Online Social Networks, as Online Social Networks have deeply changed the way humans interact with one another.

Nicolas Kourtellis gave a keynote during the workshop where he discussed work related to his studies on user activity on popular Online Social Networking platforms and the implications for the privacy of users on these platforms. He concluded by talking about all the future challenges in the space of Online Social Networks.


The main goal of the workshop is to discuss ideas and solutions concerning the analysis and the evolution of Social Media content, with a focus on how it can be viewed and interpreted as a real-time sensor (in any possible declination, negative or positive) of the surrounding society.

Social media has become one of the most important means of communication in the modern world. According to statistics in the year 2012, only 49 percent of US adults saw the news on social media, and this number moved up to 68 percent in 2018. These platforms have been enhanced in such a way that anyone can publish content and anyone who is interested in the content can easily obtain it, this has in a way revolutionized modern-day society. These aspects make social media services the most powerful sensor for any possible interpretation of modern reality. Hence, the whole intention of the workshop was to detect and understand social-media events and their consequent interpretations.

Thursday, June 30.    

Technical session: Digital humanities, culture, and society I  (Chair: Claus Atzenbeck)


1) Links Of Darkness: Hypertext And Horror 

This paper won the Douglas Engelbart Best Paper Award. In this work researchers, Mark Bernstein and Stee McMorris  "observe that the framework of Horror is congruent to the affordances of literary hypertext" wherein the framework of Horror refers to the four main stages of a horror story namely: Sighting, Thickening, Revel, and Aftermath, and this is in agreement with the wide scope of literary hypertext
.


2) Characterizing Vaccination Movements on YouTube in the United States and Brazil

Social networks played a crucial role in the context of COVID-19. Among various social networking platforms, YouTube had a large influence on the spread of information and misinformation. In this work researchers Marcelo Sartori Locatelli, Josemar Caetano, Wagner Meira Jr, and Virgilio Almeida posed an interesting question:
What are the nuances of vaccination campaigns in the two countries?

They investigated the role of YouTube in countering the spread of misinformation in the USA and Brazil as these were the two countries with the highest death tolls. A total of 8,540,234 comments were collected, of which 107,429 had a link to at least one of the following websites: YouTube, Facebook, Instagram, Twitter, Reddit, and Gab. To analyze the comments and videos the researchers evaluated (a) engagement, (b) tone of the language, (c) toxicity, and (d) lexical analysis. The researchers found that American anti-vaccine threads had a lot of toxic and negative discussions whereas Brazilian anti-vaccine threads were less engaging. Additionally, pro-vaccine threads were filled with a lot of conspiracy theories and misinformation and anti-vaccine threads were associated with protective measures. 

3) Is there an Author in this Labyrinth?

In this presentation, Dr. Sam Brooker explores the work "The Varieties of Authorial Intention" published in the year 2017 by John Farrell and effectively answers two questions: 
"How effectively do the three modes of authorial intention Farrell identifies - communicative, artistic, practical – map to hypertext fiction both past and future? Can this model – devised in the context of linear print writing – accommodate the unique form of textuality represented by hypertext, with its own affordances and opportunities to express intent?"

 

Technical session: Digital humanities, culture and society II (Chair: George Buchanan)


1) Hypertext’s meta-history: Documenting in-conference citations, authors, and keyword data, 1987-2021

Mark William Rawlence Anderson and David Millard investigated the challenges of recreating a conference's history based on available digital records and resources and explored some of the features and opportunities of that data. For their case study, they used the ACM Hypertext Conference (‘HT’).

2) The Impact of Non-Verbalization in Think-Aloud: Understanding Knowledge Gain Indicators Considering Think-Aloud Web Searches

Marcelo Tibau, Sean W. M. Siqueira and, Bernardo Pereira Nunes seek to verify the impact of verbal and non-verbal periods in Think-Aloud protocols.

3) Learning to Adapt Domain Shifts of Moral Values via InstanceWeighting

This paper published by Xiaolei Huang, Alexandra Wormley, and Adam Cohen addresses the shift in moral values and language used on social media and also proposes an adaptive framework to model the addressed shifts and a study is conducted to model COVID-19 vaccine-related behaviors from moral values.

Technical session: Social web content, language, and networks II (Chair: Daniel Roßner)


1) ADAGIO — Automated Data Augmentation of Knowledge Graphs Using Multi-expression Learning

Researchers Kevin Dreßler, Mohamed Sherif, and Axel-Cyrille Ngonga Ngomo did some interesting work wherein they leveraged multi-expression learning to create an efficient supervised algorithm based on genetic programming for learning knowledge graphs augmentation pipelines of arbitrary length.

2) Characterizing Sponsored Content in Facebook and Instagram

In this work researchers Emanuelle A. Martins, Isadora A. de Salles, Olga Goussevskaia, and, Fabrício Benevenuto collected and characterized a large-scale dataset comprising  9.5 million sponsored posts. They performed a comparative analysis of influencer marketing evolution on Facebook and Instagram for a time period of 5 years, which also included pre- and post- Covid-19 pandemic time frames. They concluded that there has been a growth of online commerce on both social networks and especially on Instagram.

3) SpaceE: Knowledge Graph Embeddingby Relational Linear Transformation in the Entity Space

The presentation focused on representation learning which means understanding representations of input data in order to generate good decisions. Researchers Jinxing Yu, Yunfeng Cai, Mingming Sun, and Ping Li studied the ability of representation learning algorithms in knowledge graph reasoning.

Friday, July 1.

Technical session: Information exploration and visualization (Chair: Ludovico Boratto)


1 )The Effects of Spatial Visualization versus Ranked Lists on Quality, Time Efficiency, and Interaction 

This paper addresses the differences between spatial layouts and ranked lists for visualizations and interactions. For this purpose, researchers Daniel Roßner, Claus Atzenbeck, and Tom Gross conducted a study with 43 participants wherein participants solved a task based on an exploratory search. 

2)  Enabling Convenient Online Collaborative Writing for Low Vision Screen Magnifier Users

In this presentation, Dr. Vikas Ashok presented our novel approach to reducing the work of low vision users using a screen magnifier on online collaborative platforms such as addressing collaborator's comments and reviewing document changes and even other basic collaborative writing activities. We also performed a user study with 15 low-vision users to assess user satisfaction and interaction experience.


 3) Exploring the Feasibility of Crowd-Powered Decomposition of Complex User Questions in Text-to-SQL Tasks

Conversion of natural language to SQL queries is quite popular and is known as Text-to-SQL models (Natural Language Interfaces to Databases (NLIDB)) however, present-day models do not tend to work really well with complex SQL queries. In this work researchers Sara Salimzadeh, Ujwal Gadiraju, Claudia Hauf, and Arie van Deursen investigated the easiness of breaking down complex user questions in a manner that each sub-question is simple enough for existing NLIDBs to generate appropriate SQL queries.

Technical session: Personalized Recommender Systems (Chair: Noemi Mauro)


1) The Effect of Recommendation Source and Justification on Professional Development Recommendations for High School Teachers

This paper presented an interesting idea to conduct a study to create a recommendation system that provides personalized professional development pathways for high school teachers. Moreover, researchers Lijie Guo, Christopher Flathmann, Reza Ghaiumy Anaraky, Nathan J. McNeese, and Bart Knijnenburg made sure to recruit participants among actual teachers. Results from this paper show that the recommender system benefits teachers most if they are told that the recommendations are from an AI algorithm and are based on their interests.

Demos


1) EarlyAd: A System for Real-Time Surveillance of Brazilian Early Electoral Ads on Twitter

Marcelo M R Araújo introduces "EarlyAd", a system that performs, in real-time, the collection, identification, and analysis of early electoral advertisements on Twitter.


2) Telegram Monitor: Monitoring Brazilian Political Groups and Channels on Telegram

Manoel M Júnior presented this work where a web-based system named “Telegram Monitor” monitors the flow of information and provides a detailed analysis of the same. This allows agencies to identify trending conspiracy theories and misinformation campaigns during elections.



Conclusion

This was the first conference I ever attended and I am really grateful for the experience this has further motivated me to focus on systematic research. We published our paper under the Technical session: Information exploration and visualization and I did get to learn a lot with respect to the current research direction in this field. It was great seeing all the wonderful presentations and demos that this conference had to offer.

Two prestigious awards were handed over and the winners were:
2) Ted Nelson Newcomer Best Paper Award: Cross-Lingual Query-Based Summarization of Crisis-Related Social Media: An Abstractive Approach Using Transformers, Fedor Vitiugin and Carlos Castillo



-YASH PRAKASH



Comments

Post a Comment