doi: 10.17586/2226-1494-2023-23-1-112-120


Software framework for hyperparameters optimization of models with additive regularization

M. A. Khodorchenko, N. A. Butakov, D. A. Nasonov, M. Y. Firulik


Read the full article  ';
Article in Russian

For citation:
Khodorchenko M.A., Butakov N.A., Nasonov D.A., Firulik M.Yu. Software framework for hyperparameters optimization of models with additive regularization. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2023, vol. 23, no. 1, pp. 112–120 (in Russian). doi: 10.17586/2226-1494-2023-23-1-112-120


Abstract
The processing of unstructured data, such as natural language texts, is one of the urgent tasks in the development of intelligent products. In turn, topic modeling as a method of working with unmarked and partially marked text data is a natural choice for analyzing document bodies and creating vector representations. In this regard, it is especially important to train high-quality thematic models in a short time which is possible with the help of the proposed framework. The developed framework implements an evolutionary approach to optimizing hyperparameters of models with additive regularization and high results on quality metrics (coherence, NPMI). To reduce the computational time, a mode of working with surrogate models is presented which provides acceleration of calculations up to 1.8 times without loss of quality. The effectiveness of the framework is demonstrated on three datasets with different statistical characteristics. The results obtained exceed similar solutions by an average of 20 % in coherence and 5 % in classification quality for two of the three datasets. A distributed version of the framework has been developed for conducting experimental studies of topic models. The developed framework can be used by users without special knowledge in the field of topic modeling due to the default data processing pipeline. The results of the work can be used by researchers to analyze topic models and expand functionality.

Keywords: AutoML framework, topic modeling, unstructured data, additive regularization, evolutionary approach, surrogate models

Acknowledgements. This research is financially supported by The Russian Science Foundation, Agreement No. 20-11-20270.

References
  1. Khanthaapha P., Pipanmaekaporn L., Kamonsantiroj S. Topic-based user profile model for POI recommendations. Proc. of the 2nd International Conference on Intelligent Systems, Metaheuristics Swarm Intelligence, 2018, pp. 143–147. https://doi.org/10.1145/3206185.3206203
  2. Peña F.J., O'Reilly-Morgan D., Tragos E.Z., Hurley N., Duriakova E., Smyth B., Lawlor A. Combining rating and review data by initializing latent factor models with topic models for top-n recommendation. Proc. of the 14th ACM Conference on Recommender Systems, 2020, pp. 438–443. https://doi.org/10.1145/3383313.3412207
  3. Sokhin T., Butakov N. Semi-automatic sentiment analysis based on topic modeling. Procedia Computer Science, 2018, vol. 136, pp. 284–292. https://doi.org/10.1016/j.procs.2018.08.286
  4. Nevezhin E., Butakov N., Khodorchenko M., Petrov M., Nasonov D. Topic-driven ensemble for online advertising generation. Proc. of the 28th International Conference on Computational Linguistics, 2020, pp. 2273–2283. https://doi.org/10.18653/v1/2020.coling-main.206
  5. Zamiralov A., Khodorchenko M., Nasonov D. Detection of housing and utility problems in districts through social media texts. Procedia Computer Science, 2020, vol. 178, pp. 213–223. https://doi.org/10.1016/j.procs.2020.11.023
  6. Shi T., Kang K., Choo J., Reddy C.K. Short-text topic modeling via non-negative matrix factorization enriched with local word-context correlations. Proc. of the World Wide Web Conference (WWW 2018), 2018, pp. 1105–1114. https://doi.org/10.1145/3178876.3186009
  7. Hofmann T. Probabilistic latent semantic indexing. Proc. of the 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '99), 1999, pp. 50–57. https://doi.org/10.1145/312624.312649
  8. Blei D.M., Ng A.Y., Jordan M.I. Latent Dirichlet allocation. Journal of Machine Learning Research, 2003, vol. 3, pp. 993–1022.
  9. Vorontsov K., Potapenko A., Plavin A. Additive regularization of topic models for topic selection and sparse factorization. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2015, vol. 9047, pp. 193–202. https://doi.org/10.1007/978-3-319-17091-6_14
  10. Card D., Tan C., Smith N.A. Neural models for documents with metadata. Proc. of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2018, pp. 2031–2040. https://doi.org/10.18653/v1/p18-1189
  11. Cao Z., Li S., Liu Y., Li W., Ji H. A novel neural topic model and its supervised extension. Proceedings of the AAAI Conference on Artificial Intelligence, 2015, vol. 29, no. 1, pp. 2210–2216. https://doi.org/10.1609/aaai.v29i1.9499
  12. Bianchi F., Terragni S., Hovy D. Pre-training is a hot topic: Contextualized document embeddings improve topic coherence. Proc. of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), 2021, pp. 759–766. https://doi.org/10.18653/v1/2021.acl-short.96
  13. Ye J., Jing X., Li J. Sentiment Analysis Using Modified LDA. Lecture Notes in Electrical Engineering, 2018, vol. 473, pp. 205–212. https://doi.org/10.1007/978-981-10-7521-6_25
  14. Bodrunova S., Koltsov S., Koltsova O., Nikolenko S., Shimorina A. Interval semi-supervised LDA: Classifying needles in a haystack. Lecture Notes in Computer Science, 2013, vol. 8265, pp. 265–274. https://doi.org/10.1007/978-3-642-45114-0_21
  15. Řehůřek R., Sojka P. Software framework for topic modelling with large corpora. Proc. of the LREC 2010 Workshop on New Challenges for NLP, 2010, pp. 45–50.
  16. Terragni S., Fersini E., Galuzzi B.G., Tropeano P., Candelieri A. OCTIS: Comparing and optimizing topic models is simple! Proc. of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, 2021, pp. 263–270. https://doi.org/10.18653/v1/2021.eacl-demos.31
  17. Khodorchenko M., Butakov N. Developing an approach for lifestyle identification based on explicit and implicit features from social media. Procedia Computer Science, 2018, vol. 136, pp. 236–245. https://doi.org/10.1016/j.procs.2018.08.262
  18. Vorontsov K., Frei O., Apishev M., Romov P., Dudarenko M. BigARTM: Open source library for regularized multimodal topic modeling of large collections. Communications in Computer and Information Science, 2015, vol. 542, pp. 370–381. https://doi.org/10.1007/978-3-319-26123-2_36
  19. Khodorchenko M., Teryoshkin S., Sokhin T., Butakov N. Optimization of learning strategies for ARTM-based topic models. Lecture Notes in Computer Science, 2020, vol. 12344, pp. 284–296. https://doi.org/10.1007/978-3-030-61705-9_24
  20. Khodorchenko M., Butakov N., Sokhin T., Teryoshkin S. Surrogate-based optimization of learning strategies for additively regularized topic models. Logic Journal of the IGPL, 2022. https://doi.org/10.1093/jigpal/jzac019
  21. Röder M., Both A., Hinneburg A. Exploring the space of topic coherence measures. Proc. of the Eighth ACM International Conference on Web Search and Data Mining (WSDM’15), 2015, pp. 399–408. https://doi.org/10.1145/2684822.2685324
  22. Newman D., Noh Y., Talley E., Karimi S., Baldwin T. Evaluating topic models for digital libraries. Proc. of the 10th Annual Joint Conference on Digital Libraries (JCDL’10), 2010, pp. 215–224. https://doi.org/10.1145/1816123.1816156


Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Copyright 2001-2024 ©
Scientific and Technical Journal
of Information Technologies, Mechanics and Optics.
All rights reserved.

Яндекс.Метрика