| 000 | 00000nam c2200205 c 4500 | |
| 001 | 000045978724 | |
| 005 | 20190416162523 | |
| 007 | ta | |
| 008 | 181226s2019 ulkd bmAC 000c eng | |
| 040 | ▼a 211009 ▼c 211009 ▼d 211009 | |
| 085 | 0 | ▼a 0510 ▼2 KDCP |
| 090 | ▼a 0510 ▼b 6D36 ▼c 1093 | |
| 100 | 1 | ▼a 이지민 ▼g 李知玟 |
| 245 | 1 0 | ▼a Incorporating neural models into open directory project based large-scale text classification / ▼d Ji-min Lee |
| 260 | ▼a Seoul : ▼b Graduate School, Korea Unversity, ▼c 2019 | |
| 300 | ▼a iv, 39장 : ▼b 도표 ; ▼c 26 cm | |
| 500 | ▼a 지도교수: 이상근 | |
| 502 | 0 | ▼a 학위논문(석사)-- ▼b 고려대학교 대학원: ▼c 컴퓨터·전파통신공학과, ▼d 2019. 2 |
| 504 | ▼a 참고문헌: 장 36-39 | |
| 530 | ▼a PDF 파일로도 이용가능; ▼c Requires PDF file reader(application/pdf) | |
| 653 | ▼a Text classification ▼a Neural network ▼a Word embeddings | |
| 776 | 0 | ▼t Incorporating Neural Models into Open Directory Project based Large-scale Text Classification ▼w (DCOLL211009)000000083457 |
| 900 | 1 0 | ▼a Lee, Ji-min, ▼e 저 |
| 900 | 1 0 | ▼a 이상근 ▼g 李尙根, ▼e 지도교수 |
| 945 | ▼a KLPA |
Electronic Information
| No. | Title | Service |
|---|---|---|
| 1 | Incorporating neural models into open directory project based large-scale text classification (28회 열람) |
View PDF Abstract Table of Contents |
Holdings Information
| No. | Location | Call Number | Accession No. | Availability | Due Date | Make a Reservation | Service |
|---|---|---|---|---|---|---|---|
| No. 1 | Location Science & Engineering Library/Stacks(Thesis)/ | Call Number 0510 6D36 1093 | Accession No. 123060849 | Availability Available | Due Date | Make a Reservation | Service |
| No. 2 | Location Science & Engineering Library/Stacks(Thesis)/ | Call Number 0510 6D36 1093 | Accession No. 123060850 | Availability Available | Due Date | Make a Reservation | Service |
Contents information
Abstract
In natural language processing, large-scale text classification has been utilized to capture the various topics in arbitrary texts. For large-scale text classification, many approaches have used the explicit representation model based on knowledge bases. Although such approaches exhibit promising results, their performance is limited to the associated knowledge base. In this thesis, we incorporate neural models into the large-scale text classification. To this end, we propose an attentive joint model and adaptive joint models that represent documents and categories using word embeddings adapted to the knowledge base. To demonstrate the efficacy of our strategy, we apply the proposed methodologies to the Open Directory Project (ODP)-based text classification task. The proposed methods outperform the recent state-of-the-art method in terms of micro-averaging F1-score, macro-averaging F1-score, and precision at k.
Table of Contents
Abstract Contents i List of Figures iii List of Tables iv 1 Introduction 1 2 Preliminary 4 2.1 ODP-based Text Classification 4 2.2 Word Embeddings 5 2.3 ODP-based Text Classification with Word Embeddings 6 3 Attentive Joint Model for ODP-based Text Classification 8 3.1 Attentive Joint Model 8 3.2 Semantic Similarity Method 12 4 Adaptive Joint Models for ODP-based Text Classification 14 4.1 Adaptive Joint Model Trained by ODP Documents 14 4.2 Adaptive Joint Model Trained by ODP Categories 16 4.3 Classification in Adaptive Joint Models 17 5 Performance Evaluation 18 5.1 Datasets 18 5.1.1 ODP Dataset 18 5.1.2 NYT Dataset 19 5.2 Evaluation Metrics 19 5.2.1 ODP Dataset 19 5.2.2 NYT Dataset 19 5.3 Experimental Setup 20 5.4 Experimental Results 22 5.4.1 Results of the ODP Dataset 22 5.4.2 Results of the NYT Dataset 24 5.5 Analysis of NYT Classification Results 26 6 Related Work 32 6.1 Large-scale Text Classification 32 6.2 ODP-based Text Classification 33 7 Conclusion 35 Bibliography 36 Acknowledgement 40
