• Corpus ID: 227177260

Towards Predicting the Subscription Status of Twitch.tv Users - ECML-PKDD ChAT Discovery Challenge 2020

@inproceedings{Kobs2020TowardsPT,
  title={Towards Predicting the Subscription Status of Twitch.tv Users - ECML-PKDD ChAT Discovery Challenge 2020},
  author={Konstantin Kobs and Martin Potthast and Matti Wiegmann and Albin Zehe and Benno Stein and Andreas Hotho},
  booktitle={ChAT@PKDD/ECML},
  year={2020}
}
We investigate whether the subscription status of active users of Twitch can be inferred from their activity patterns in the chats of streamers. To enable a diversity of solutions to this problem, this task was advertised as an ECML-PKDD discovery challenge 2020, called Chat Analytics for Twitch (ChAT). Four participants submitted their working prediction models, which were evaluated at our site. The winning approach achieved an F1 score of 0.343, outperforming the baseline by a significant… 

Figures and Tables from this paper

Team voyTECH: User Activity Modeling with Boosting Trees

The connection between target-encodings and boosting trees in the context of high cardinality categoricals is introduced and it is found that modeling user activity is more powerful then direct modeling of content when encoded properly and combined with a suitable optimization approach.

Detecting Potential Subscribers on Twitch: A Text Mining Approach with XGBoost - Discovery Challenge ChAT: CoolStoryBob

The motivation of this research is to detect potential subscribers by predicting a user’s subscription status using a trained ML model, which can then be targeted with marketing campaigns.

StinkyCheese: Chat-Based Model for Subscription Classification

The process of building a binary classifier for the Chat Analytics for Twitch challenge, which is to predict the subscription status of a user in a channel based on their comments, is reported.

Data Analysis and Friendship Prediction for Twitch Streamers

A new recommendation framework is designed based on the specific characteristics of a Twitch network gamers dataset and consists of three different layers: data, interest, and recommendation layer, each considering specific tasks.

References

SHOWING 1-10 OF 24 REFERENCES

Team voyTECH: User Activity Modeling with Boosting Trees

The connection between target-encodings and boosting trees in the context of high cardinality categoricals is introduced and it is found that modeling user activity is more powerful then direct modeling of content when encoded properly and combined with a suitable optimization approach.

Detecting Potential Subscribers on Twitch: A Text Mining Approach with XGBoost - Discovery Challenge ChAT: CoolStoryBob

The motivation of this research is to detect potential subscribers by predicting a user’s subscription status using a trained ML model, which can then be targeted with marketing campaigns.

StinkyCheese: Chat-Based Model for Subscription Classification

The process of building a binary classifier for the Chat Analytics for Twitch challenge, which is to predict the subscription status of a user in a channel based on their comments, is reported.

Towards the Understanding of Gaming Audiences by Modeling Twitch Emotes

This paper proposes two important tasks specific to the Twitch platform, namely Emote prediction; and Trolling detection, and evaluates three models: a BOW baseline, a logistic supervised classifiers based on word embeddings, and a bidirectional long short-term memory recurrent neural network (LSTM).

Time-Aware Subscription Prediction Model for User Acquisition in Digital News Media

This work designs a model that not only predicts the potential subscribers but also answers queries about the subscription occurrence time and can be used to predict the subscription time and recommend accurately the “potential users” to the current marketing campaign.

WTF: the who to follow service at Twitter

An architectural overview of the architecture of WTF is provided and a few graph recommendation algorithms implemented in Cassovary are described and evaluated, including a novel approach based on a combination of random walks and SALSA.

On crowdsourced interactive live streaming: a Twitch.tv-based measurement study

An initial investigation on the modern crowdsourced live streaming systems is presented, showing that the current delay strategy on the viewer's side substantially impacts the viewers' interactive experience, and there is significant disparity between the long broadcast latency and the short live messaging latency.

“And Today’s Top Donator is”: How Live Streamers on Twitch.tv Monetize and Gamify Their Broadcasts

This article examines cultural and economic behavior on live streaming platform Twitch.tv, and the monetization of live streamers’ content production. Twitch is approximately the thirtieth

The impacts of live streaming and Twitch.tv on the video game industry

It is concluded that live streaming is a major new force in the games industry, creating new links between developers and influencers and shifting the authors' expectations of game play and game design, and is consequently a platform whose major structural effects are only now beginning to be understood.

Collaborative filtering recommender systems

This study presents an overview of the field of recommender systems with current generation of recommendation methods and examines comprehensively CF systems with its algorithms.