Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Source code of CHAMELEON - A Deep Learning Meta-Architecture for News Recommender Systems
| Date | Stars |
|---|---|
| 2026-07-24 | 278 |
| 2026-07-25 | 278 |
| 2026-07-28 | 278 |
| 2026-07-30 | 278 |
| 2026-07-31 | 278 |
| 2026-08-06 | 278 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# CHAMELEON - A Deep Learning Meta-Architecture for News Recommender Systems CHAMELEON is a Deep Learning Meta-Architecture for News Recommender Systems [1]. It has being developed as part of the Doctoral research of Gabriel de Souza Pereira Moreira, at the Brazilian Aeronautics Institute of Technology ([ITA](http://www.ita.br/)). The initial version ([v1.0](https://github.com/gabrielspmoreira/chameleon_recsys/commits/v1.0)) of CHAMELEON source code allows reproducibility of the experiments reported in a [paper](https://arxiv.org/abs/1808.00076) [2] published at the [DLRS'18](https://recsys.acm.org/recsys18/dlrs/), co-located with ACM RecSys'18. The version [v1.7.3](https://github.com/gabrielspmoreira/chameleon_recsys/commits/v1.7.3) was released for reproducibility of the experiments reported in our papers at the [INRA'19](https://arxiv.org/abs/1907.07629), co-located with [ACM RecSys'19](https://recsys.acm.org/recsys19/inra/) [3], and at the [IEEE Access journal](https://ieeexplore.ieee.org/document/8908688) [4]. The version v1.7.3 includes metrics for item coverage, novelty and diversity, and item-cold start, an optimized instantiation of CHAMELEON meta-architecture and experiments with two datasets are made available: [G1 (Globo.com)](https://www.kaggle.com/gspmoreira/news-portal-user-interactions-by-globocom) and [Adressa](http://reclab.idi.ntnu.no/dataset). Furthermore, a recent baseline using Graph Neural Networks (SR-GNN) was included and some additional instantiations of the ACR module were implemented, using GRUs for both supervised and unsupervised training (Sequence Denoising Autoencoder) of the Article Content Embeddings. ## Publications [1] Gabriel de Souza Pereira Moreira, Felipe Ferreira, and Adilson Marques da Cunha. 2018. *CHAMELEON: A Deep Learning Meta-Architecture for News Recommender Systems*. In Proceedings of Doctoral Symposium of the 12th ACM RecSys'18, October 6, 2018, Vancouver, BC, Canada. ACM, New York, NY, USA, 9 pages. https://doi.org/10.1145/3240323.3240331 [2] Gabriel de Souza Pereira Moreira, Felipe Ferreira, and Adilson Marques da Cunha. 2018. *News Session-Based Recommendations using Deep Neural Networks*. In 3rd Workshop on Deep Learning for Recommender Systems (DLRS 2018), October 6, 2018, Vancouver, BC, Canada. ACM, New York, NY, USA, 9 pages. https://doi.org/10.1145/3270323.3270328 [3] Gabriel de Souza Pereira Moreira, Dietmar Jannach, and Adilson Marques da Cunha. 2019. *On the Importance of News Content Representation in Hybrid Neural Session-based Recommender Systems*. In 7th International Workshop on News Recommendation and Analytics (INRA 2019), in conjunction with RecSys 2019, September 19, 2019, Copenhagen, Denmark. https://arxiv.org/abs/1907.07629 [4] Gabriel de Souza Pereira Moreira, Dietmar Jannach, and Adilson Marques da Cunha. 2019. *Contextual Hybrid Session-based News Recommendation with Recurrent Neural Networks*. IEEE Access, v. 7, p. 169185-169203, 2019. https://doi.org/10.1109/ACCESS.2019.2954957 [5] Gabriel de Souza Pereira Moreira, 2019. *CHAMELEON: A Deep Learning Meta-Architecture for News Recommender Systems*. Doctoral Thesis, Instituto Tecnológico de Aeronáutica (ITA), Brazil. https://arxiv.org/abs/2001.04831 ## Implementation This implementation uses **Python 3** (with Pandas, Scikit-learn and SciPy modules) and **TensorFlow 1.12**. CHAMELEON modules were implemented using TF [Estimators](https://www.tensorflow.org/guide/estimators) and [Datasets](https://www.tensorflow.org/guide/datasets). The CHAMELEON modules training and evaluation can be performed either locally (GPU highly recommended) or using [Google Cloud Platform ML Engine](https://cloud.google.com/ml-engine/) managed service. ## Dataset for reproducibility The experiments reported in the papers [3], [4] and [5] use the following datasets: * [Globo.com (G1) dataset](https://www.kaggle.com/gspmoreira/news-portal-user-interactions-by-globocom) - Globo.com is the most popular media company
Excerpt of 26,655 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:d02b4b23a161e061, topic:deep-learning, topic:tensorflow