• Journal of Intelligence and Information Systems
• /
• v.27 no.4
• /
• pp.1-22
• /
• 2021

Aspect Based Sentiment Analysis (ABSA), which analyzes sentiment based on aspects that appear in the text, is drawing attention because it can be used in various business industries. ABSA is a study that analyzes sentiment by aspects for multiple aspects that a text has. It is being studied in various forms depending on the purpose, such as analyzing all targets or just aspects and sentiments. Here, the aspect refers to the property of a target, and the target refers to the text that causes the sentiment. For example, for restaurant reviews, you could set the aspect into food taste, food price, quality of service, mood of the restaurant, etc. Also, if there is a review that says, "The pasta was delicious, but the salad was not," the words "steak" and "salad," which are directly mentioned in the sentence, become the "target." So far, in ABSA, most studies have analyzed sentiment only based on aspects or targets. However, even with the same aspects or targets, sentiment analysis may be inaccurate. Instances would be when aspects or sentiment are divided or when sentiment exists without a target. For example, sentences like, "Pizza and the salad were good, but the steak was disappointing." Although the aspect of this sentence is limited to "food," conflicting sentiments coexist. In addition, in the case of sentences such as "Shrimp was delicious, but the price was extravagant," although the target here is "shrimp," there are opposite sentiments coexisting that are dependent on the aspect. Finally, in sentences like "The food arrived too late and is cold now." there is no target (NULL), but it transmits a negative sentiment toward the aspect "service." Like this, failure to consider both aspects and targets - when sentiment or aspect is divided or when sentiment exists without a target - creates a dual dependency problem. To address this problem, this research analyzes sentiment by considering both aspects and targets (Target-Aspect-Sentiment Detection, hereby TASD). This study detected the limitations of existing research in the field of TASD: local contexts are not fully captured, and the number of epochs and batch size dramatically lowers the F1-score. The current model excels in spotting overall context and relations between each word. However, it struggles with phrases in the local context and is relatively slow when learning. Therefore, this study tries to improve the model's performance. To achieve the objective of this research, we additionally used auxiliary loss in aspect-sentiment classification by constructing CNN(Convolutional Neural Network) layers parallel to existing models. If existing models have analyzed aspect-sentiment through BERT encoding, Pooler, and Linear layers, this research added CNN layer-adaptive average pooling to existing models, and learning was progressed by adding additional loss values for aspect-sentiment to existing loss. In other words, when learning, the auxiliary loss, computed through CNN layers, allowed the local context to be captured more fitted. After learning, the model is designed to do aspect-sentiment analysis through the existing method. To evaluate the performance of this model, two datasets, SemEval-2015 task 12 and SemEval-2016 task 5, were used and the f1-score increased compared to the existing models. When the batch was 8 and epoch was 5, the difference was largest between the F1-score of existing models and this study with 29 and 45, respectively. Even when batch and epoch were adjusted, the F1-scores were higher than the existing models. It can be said that even when the batch and epoch numbers were small, they can be learned effectively compared to the existing models. Therefore, it can be useful in situations where resources are limited. Through this study, aspect-based sentiments can be more accurately analyzed. Through various uses in business, such as development or establishing marketing strategies, both consumers and sellers will be able to make efficient decisions. In addition, it is believed that the model can be fully learned and utilized by small businesses, those that do not have much data, given that they use a pre-training model and recorded a relatively high F1-score even with limited resources.

• The Korean Journal of Air & Space Law and Policy
• /
• v.24 no.1
• /
• pp.59-89
• /
• 2009

The objectives of the 1963 Tokyo Convention cover a variety of subjects, with the intention of providing safety in aircraft, protection of life and property on board, and promoting the security of civil aviation. These objectives will be treated as follows: first, the unification of rules on jurisdiction; second, the question of filling the gap in jurisdiction; third, the scheme of maintaining law and order on board aircraft; fourth, the protection of persons acting in accordance with the Convention; fifth, the protection of the interests of disembarked persons; sixth, the question of hijacking of aircraft; and finally some general remarks on the objectives of the Convention. The Tokyo Convention mainly deals with general crimes such as murder, violence, robbery on board aircraft rather than aviation terrorism. The Article 11 of the Convention deals with hijacking in a simple way. As far as aviation terrorism is concerned 1970 Hague Convention and 1971 Montreal Convention cover the hijacking and sabotage respectively. The Problem of national jurisdiction over the offence and the offender was as tangled at the Hague and Montreal Convention, as under the Tokyo Convention. Under the Tokyo Convention the prime base of jurisdiction is the law of the flag (Article 3), but concurrent jurisdiction is also allowed on grounds of: territorial principle, active nationality and passive personality principle, security of the state, breach of flight rules, and exercise of jurisdiction necessary for the performance of obligations under multilateral agreements (Article 4). No Criminal jurisdiction exercised in accordance with national law is excluded [Article 3(2)]. However, Article 4 of the Hague Convention(hereafter Hague Article 4) and Article 5 of the Montreal Convention(hereafter Montreal Article 5), dealing with jurisdiction have moved a step further, inasmuch as the opening part of both paragraphs 1 and 2 of the Hague Article 4 and the Montreal Article 5 impose an obligation on all contracting states to take measures to establish jurisdiction over the offence (i.e., to ensure that their law is such that their courts will have jurisdiction to try offender in all the circumstances covered by Hague Article 4 and Montreal Article 5). The state of registration and the state where the aircraft lands with the hijacker still on board will have the most interest, and would be in the best position to prosecute him; the paragraphs 1(a) and (b) of the Hague Article 4 and paragraphs 1(b) and (c) of the Montreal Article 5 deal with it, respectively. However, paragraph 1(b) of the Hague Article 4 and paragraph 1(c) of the Montreal Article 5 do not specify if the aircraft is still under the control of the hijacker or if the hijacker has been overpowered by the aircraft commander, or if the offence has at all occurred in the airspace of the state of landing. The language of the paragraph would probably cover all these cases. The weaknesses of Hague Article 4 and Montreal Article 5 are however, patent. The Jurisdictions of the state of registration, the state of landing, the state of the lessee and the state where the offender is present, are concurrent. No priorities have been fixed despite a proposal to this effect in the Legal Committee and the Diplomatic Conference, and despite the fact that it was pointed out that the difficulty in accepting the Tokyo Convention has been the question of multiple jurisdiction, for the reason that it would be too difficult to determine the priorities. Disputes over the exercise of jurisdiction can be endemic, more so when Article 8(4) of the Hague Convention and the Montreal Convention give every state mentioned in Hague Article 4(1) and Montreal Article 5(1) the right to seek extradition of the offender. A solution to the problem should not have been given up only because it was difficult. Hague Article 4(3) and Montreal Article 5(3) provide that they do not exclude any criminal jurisdiction exercised in accordance with national law. Thus the provisions of the two Conventions create additional obligations on the state, and do not exclude those already existing under national laws. Although the two Conventions do not require a state to establish jurisdiction over, for example, hijacking or sabotage committed by its own nationals in a foreign aircraft anywhere in the world, they do not preclude any contracting state from doing so. However, it has be noted that any jurisdiction established merely under the national law would not make the offence an extraditable one under Article 8 of the Hague and Montreal Convention. As far as international aviation terrorism is concerned 1988 Montreal Protocol and 1991 Convention on Marking of Plastic Explosives for the Purpose of Detention are added. The former deals with airport terrorism and the latter plastic explosives. Compared to the other International Terrorism Conventions, the International Aviation Terrorism Conventions do not have clauses of the passive personality principle. If the International Aviation Terrorism Conventions need to be revised in the future, those clauses containing the passive personality principle have to be inserted for the suppression of the international aviation terrorism more effectively. Article 3 of the 1973 Convention on the Prevention and Punishment of Crimes Against Internationally Protected Persons, Including Diplomatic Agents, Article 5 of the 1979 International Convention against the Taking of Hostages and Article 6 of the 1988 Convention for the Suppression of Unlawful Acts Against the Safety of Maritime Navigation would be models that the revised International Aviation Terrorism Conventions could follow in the future.