Collaborative, Dynamic and Diversified User Profiling
Authors: Shangsong Liang4269-4276
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments were conducted on a Twitter dataset and we found that UPA outperforms state-of-the-art non-dynamic and dynamic user profiling algorithms. |
| Researcher Affiliation | Academia | Shangsong Liang1,2 1School of Data and Computer Science, Sun Yat-sen University, China 2Guangdong Key Laboratory of Big Data Analysis and Processing, Guangzhou 51006, China |
| Pseudocode | Yes | Algorithm 1: Inference for our CITM model at time t. |
| Open Source Code | No | The paper describes the proposed models and algorithms (UPA, CITM, SKDM) but does not contain any explicit statement about releasing the source code for these methods, nor does it provide a link to a code repository. |
| Open Datasets | No | We work with a dataset collected from Twitter.1 It contains 1,375 active randomly selected users and their tweets posted from the beginning of their registrations up to May 31, 2015. The footnote links to dev.twitter.com, which is the Twitter API documentation, not a dataset. No direct link to *their* collected dataset is provided. |
| Dataset Splits | Yes | For tuning parameters, λ, η1 and η2, we use a 70%/20%/10% split for our training, validation and test sets, respectively. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used to conduct the experiments. |
| Software Dependencies | No | The paper mentions methods like k-means and uses publicly available GloVe embeddings, but does not specify any software dependencies (e.g., programming languages or libraries) with their version numbers that would be required to reproduce the experiments. |
| Experiment Setup | Yes | We set the number of topics Z = 20 in all the topic models. For tuning parameters, λ, η1 and η2, we use a 70%/20%/10% split for our training, validation and test sets, respectively. We repeat the experiments 10 times and report the average results. Since we usually choose not too many keywords to describe a user s profile, we compute the scores at depth 10, i.e., let k = 10. |