Ekstraksi Relasi Antar Entitas di Bahasa Indonesia Menggunakan Neural Network

Dengan perkembangan zaman yang begitu pesat, berdampak pada perkembangan data pula. Salah satu bentuk data yang paling banyak saat ini berupa data tekstual seperti artikel sederhana maupun dokumen lain yang terdapat di internet. Agar data tekstual tersebut dapat dimengerti dan dimanfaatkan dengan baik oleh manusia, maka perlu di proses dan disederhanakan agar menjadi informasi yang ringkas dan jelas. Oleh karena itu, semakin berkembang pula penelitian dalam bidang Information Extraction (IE) dan salah satu contoh penelitian di IE adalah Relation Extraction (RE). Penelitian RE sudah banyak dilakukan terutama pada Bahasa Inggris dimana resourcenya sudah termasuk banyak. Metode yang digunakan pun bermacam-macam seperti kernel, tree kernel, support vector machine, long short-term memory, convulution recurrent neural network, dan lain sebagainya. Pada penelitian kali ini adalah penelitian RE pada Bahasa Indonesia dengan menggunakan metode convulution recurrent neural network yang sudah dipergunakan untuk RE Bahasa Inggris. Dataset yang digunakan pada penelitian ini adalah dataset Bahasa Indonesia yang berasal dari file xml wikipedia. File xml wikipedia ini kemudian diproses sehingga menghasilkan dataset seperti yang digunakan pada CRNN dalam Bahasa inggris yaitu dalam format SemEval-2 Task 8. Uji coba dilakukan dengan berbagai macam perbandingan data training dan testing yaitu 80:20, 70:30, dan 60:40. Selain itu, parameter pooling untuk CRNN yang digunakan ada dua macam yaitu ‘att’ dan ‘max’. Dari uji coba yang dilakukan, hasil yang didapatkan adalah bervariasi mulai dari mendekati maupun lebih baik bila dibandingkan dengan CRNN dengan menggunakan dataset Bahasa inggris sehingga dapat disimpulkan bahwa dengan CRNN ini bisa digunakan untuk proses RE pada Bahasa Indonesia apabila dataset yang digunakan sesuai dengan penelitian sebelumnya.

Domain Adaptation via Tree Kernel Based Maximum Mean Discrepancy for User Consumption Intention Identification

Identifying user consumption intention from social media is of great interests to downstream applications. Since such task is domain-dependent, deep neural networks have been applied to learn transferable features for adapting models from a source domain to a target domain. A basic idea to solve this problem is reducing the distribution difference between the source domain and the target domain such that the transfer error can be bounded. However, the feature transferability drops dramatically in higher layers of deep neural networks with increasing domain discrepancy. Hence, previous work has to use a few target domain annotated data to train domain-specific layers. In this paper, we propose a deep transfer learning framework for consumption intention identification, to reduce the data bias and enhance the transferability in domain-specific layers. In our framework, the representation of the domain-specific layer is mapped to a reproducing kernel Hilbert space, where the mean embeddings of different domain distributions can be explicitly matched. By using an optimal tree kernel method for measuring the mean embedding matching, the domain discrepancy can be effectively reduced. The framework can learn transferable features in a completely unsupervised manner with statistical guarantees. Experimental results on five different domain datasets show that our approach dramatically outperforms state-of-the-art baselines, and it is general enough to be applied to more scenarios. The source code and datasets can be found at http://ir.hit.edu.cn/$\scriptsize{\sim}$xding/index\_english.htm.

ACV-tree: A New Method for Sentence Similarity Modeling

Sentence similarity modeling lies at the core of many natural language processing applications, and thus has received much attention. Owing to the success of word embeddings, recently, popular neural network methods have achieved sentence embedding, obtaining attractive performance. Nevertheless, most of them focused on learning semantic information and modeling it as a continuous vector, while the syntactic information of sentences has not been fully exploited. On the other hand, prior works have shown the benefits of structured trees that include syntactic information, while few methods in this branch utilized the advantages of word embeddings and another powerful technique ? attention weight mechanism. This paper makes the first attempt to absorb their advantages by merging these techniques in a unified structure, dubbed as ACV-tree. Meanwhile, this paper develops a new tree kernel, known as ACVT kernel, that is tailored for sentence similarity measure based on the proposed structure. The experimental results, based on 19 widely-used datasets, demonstrate that our model is effective and competitive, compared against state-of-the-art models.

Understanding visual scenes

A growing body of recent work focuses on the challenging problem of scene understanding using a variety of cross-modal methods which fuse techniques from image and text processing. In this paper, we develop representations for the semantics of scenes by explicitly encoding the objects detected in them and their spatial relations. We represent image content via two well-known types of tree representations, namely constituents and dependencies. Our representations are created deterministically, can be applied to any image dataset irrespective of the task at hand, and are amenable to standard NLP tools developed for tree-based structures. We show that we can apply syntax-based SMT and tree kernel methods in order to build models for image description generation and image-based retrieval. Experimental results on real-world images demonstrate the effectiveness of the framework.