原文传递
Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning.
| 题名： | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning. |
| 作者： | Petrocelli, N. A. |
| 摘要： | Users often sacrifice personal data for more relevant search results, presenting aproblem to communities that desire both search anonymity and relevant results. Tobalance these priorities, this research examines the impact of using Siamese networksto extend word embeddings into document embeddings and detect similarities be-tween documents. The predicted similarity can locally re-rank search results providedfrom various sources. This technique is leveraged to limit the amount of informationcollected from a user by a search engine. A prototype is produced by applying themethodology in a real-world search environment. The prototype yielded an addi-tional function of finding new documents related to a provided sample document.The prototype is evaluated using real-world search examples. Results indicate thatthe Siamese network can produce document embeddings superior to current encoderslike the Universal Sentence Encoder. Results also show the promising performance ofthe prototype in improving search relevancy while limiting user data transmission. |
| 总页数： | 104 pages |
相关文献
- Certification for Machine Learning.
- Introduction to NASA Goddard Workshop on Artificial Intelligence.
- Estimating Classifier Accuracy Using Noisy Expert Labels.
- Addressing Challenges of Machine Translation of Inuit Languages.
- Predicting Arrival and Departure Runway Assignments with Machine Learning.
- Predicting Arrival and Departure Runway Assignments with Machine Learning.
检索历史
应用推荐
