Achieving the Sustainable Development Goals (SDGs) requires collaboration among various stakeholders, particularly governments and non-state actors (NSAs). This collaboration results in but is also based on a continually growing volume of documents that needs to be analyzed and processed in a systematic way by government officials. Artificial Intelligence and Natural Language Processing (NLP) could, thus, offer valuable support for progressing towards SDG targets, including automating the government budget tagging and classifying NSA requests and initiatives, as well as helping uncover the possibilities for matching these two categories of activities. Many non-English speaking countries, including Indonesia, however, face limited NLP resources, such as, for instance, domain-specific pre-trained language models (PTLMs). This circumstance makes it difficult to automate document processing and improve the efficacy of SDG-related government efforts. The presented study introduces IndoGovBERT, a Bidirectional Encoder Representations from Transformers (BERT)-based PTLM built with domain-specific corpora, leveraging the Indonesian government’s public and internal documents. The model is intended to automate various laborious tasks of SDG document processing by the Indonesian government. Different approaches to PTLM development known from the literature are examined in the context of typical government settings. The most effective, in terms of the resultant model performance, but also most efficient, in terms of the computational resources required, methodology is determined and deployed for the development of the IndoGovBERT model. The developed model is then scrutinized in several text classification and similarity assessment experiments, where it is compared with four Indonesian general-purpose language models, a non-transformer approach of the Multilabel Topic Model (MLTM), as well as with a Multilingual BERT model. Results obtained in all experiments highlight the superior capability of the IndoGovBERT model for Indonesian government SDG document processing. The latter suggests that the proposed PTLM development methodology could be adopted to build high-performance specialized PTLMs for governments around the globe which face SDG document processing and other NLP challenges similar to the ones dealt with in the presented study.

Online petitions are an important means for citizens to express their concerns and to interact with government entities. Due to the increase in the number of petitions, manually attributing them to the competent unit in public administration creates a bottleneck, leading to response delays, and potentially even errors. To address this problem, a multi-label classifier using fine-tuned BERT model is suggested. The proposed model, trained on a dataset from the Taiwanese Join Platform, performs reasonably well in predicting the governmental departments in charge of petitions, even when trained on an imbalanced and rather small dataset. The obtained model manages to effectively process petitions and predicts responsible departments, achieving F1 score of 0.61 averaged over 12 categories. The proposed approach would potentially improve government responsiveness, optimize resource allocation, and facilitate online petition processing. Future work would focus on improving the model’s generalization capabilities.

Online petition systems as a crowd-sourcing and feedback tool offered by public authorities are becoming more and more common these days. The volume of incoming petitions from digital platforms is increasing. Governments typically review petitions in a screening process before they can be posted on an online platform. In order to reduce the burden on public authorities and to provide fast feedback to a petitioner on whether a petition is admissible or not, an automated system is proposed to assistance to the authorities and petitioners alike. The proposed system attempts to classify whether a petition is admissible or not. By applying a BERT-based model to a dataset of petitions submitted to Taiwan’s JOIN platform 2015-2023, the classification model achieve a classification accuracy of 67%. While the model would be improved in future work, the experiments conducted the overall feasibility of the proposed system to alleviate the overcrowding of civic tech tools.

Recently, real-time affect-awareness has been applied in several commercial systems, such as dialogue systems and computer games. Real-time recognition of affective states, however, requires the application of costly feature extraction methods and/or labor-intensive annotation of large datasets, especially in the case of Asian languages where large annotated datasets are seldom available. To improve recognition accuracy, we propose the use of cognitive context in the form of “emotion-sensitive” intentions. Intentions are often represented through dialogue acts and, as an emotion-sensitive model of dialogue acts, a tagset of interpersonal-relations-directing interpersonal acts (the IA model) is proposed. The model's adequacy is assessed using a sentiment classification task in comparison with two well-known dialogue act models, the SWBD-DAMSL and the DIT++. For the assessment, five Japanese in-game dialogues were annotated with labels of sentiments and the tags of all three dialogue act models which were used to enhance a baseline sentiment classifier system. The adequacy of the IA tagset is demonstrated by a 9% improvement to the baseline sentiment classifier's recognition accuracy, outperforming the other two models by more than 5%.