Short and Sparse Text Topic Modeling via Self-Aggregation
Authors: Xiaojun Quan, Chunyu Kit, Yong Ge, Sinno Jialin Pan
IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on real-world datasets validate the effectiveness of this new model, suggesting that it can distill more meaningful topics from short texts. |
| Researcher Affiliation | Academia | Xiaojun Quan1, Chunyu Kit2, Yong Ge3, Sinno Jialin Pan4 1Institute for Infocomm Research, A*STAR, Singapore 2City University of Hong Kong, Hong Kong SAR, China 3UNC Charlotte, NC, USA 4Nanyang Technological University, Singapore |
| Pseudocode | No | The paper describes generative processes and sampling steps in prose and mathematical equations (e.g., Section 3.1 'Generative Process', Section 3.2 'Gibbs Sampling'), but does not include any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any links to open-source code for the methodology described, nor does it state that the code is publicly available. |
| Open Datasets | Yes | NIPS. The first corpus consists of 1,740 NIPS conference papers over the period of 2000 to 2012... Yahoo! Answers. This corpus, crawled from Yahoo! Answers1, consists of 88,120 questions from 11 categories... 1https://answers.yahoo.com/ |
| Dataset Splits | Yes | We conduct a short text classification task for such a purpose on the Yahoo! Answers corpus by randomly dividing it into two equal-sized subsets for training and testing... For the implementation of SVM, the LIBSVM library [Chang and Lin, 2011] is employed, with its parameters chosen by five-fold cross-validation on the training set. |
| Hardware Specification | No | The paper does not specify any hardware details such as GPU/CPU models, memory, or specific computing environments used for running the experiments. |
| Software Dependencies | No | The paper mentions the use of 'LIBSVM library [Chang and Lin, 2011]' for SVM implementation but does not provide a specific version number for this or any other software dependency. |
| Experiment Setup | Yes | First, the number of iterations for Gibbs sampling is set to 3000... Then, following previous work [Griffiths and Steyvers, 2004; Weng et al., 2010], the parameters of α and β are set to 50/T and 0.1, respectively... The parameter, K, is fixed at 20... The number of pseudo-documents of SATM is fixed at 1000. |