Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning
Users often sacrifice personal data for more relevant search results, presenting a problem to communities that desire both search anonymity and relevant results. To balance these priorities, this research examines the impact of using Siamese networks to extend word embeddings into document embedding...
I tiakina i:
| Kaituhi matua: | |
|---|---|
| Hōputu: | text |
| I whakaputaina: |
AFIT Scholar
2022
|
| Ngā marau: | |
| Urunga tuihono: | https://scholar.afit.edu/etd/5327 https://scholar.afit.edu/context/etd/article/6329/viewcontent/AFIT_ENG_MS_22_M_055_Petrocelli_N.pdf |
| Ngā Tūtohu: |
Tāpirihia he Tūtohu
Kāore He Tūtohu, Me noho koe te mea tuatahi ki te tūtohu i tēnei pūkete!
|
| _version_ | 1870452256298500096 |
|---|---|
| author | Petrocelli, Niko A. |
| author_facet | Petrocelli, Niko A. |
| author_sort | Petrocelli, Niko A. |
| building | US Air Force Institute of Technology (AFIT) |
| collection | AFIT Scholar |
| description | Users often sacrifice personal data for more relevant search results, presenting a problem to communities that desire both search anonymity and relevant results. To balance these priorities, this research examines the impact of using Siamese networks to extend word embeddings into document embeddings and detect similarities between documents. The predicted similarity can locally re-rank search results provided from various sources. This technique is leveraged to limit the amount of information collected from a user by a search engine. A prototype is produced by applying the methodology in a real-world search environment. The prototype yielded an additional function of finding new documents related to a provided sample document. The prototype is evaluated using real-world search examples. Results indicate that the Siamese network can produce document embeddings superior to current encoders like the Universal Sentence Encoder. Results also show the promising performance of the prototype in improving search relevancy while limiting user data transmission. |
| format | text |
| id | afit-etd-6329 |
| institution | US Air Force Institute of Technology |
| publishDate | 2022 |
| publisher | AFIT Scholar |
| record_format | dspace |
| spelling | afit-etd-6329 Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning Petrocelli, Niko A. Users often sacrifice personal data for more relevant search results, presenting a problem to communities that desire both search anonymity and relevant results. To balance these priorities, this research examines the impact of using Siamese networks to extend word embeddings into document embeddings and detect similarities between documents. The predicted similarity can locally re-rank search results provided from various sources. This technique is leveraged to limit the amount of information collected from a user by a search engine. A prototype is produced by applying the methodology in a real-world search environment. The prototype yielded an additional function of finding new documents related to a provided sample document. The prototype is evaluated using real-world search examples. Results indicate that the Siamese network can produce document embeddings superior to current encoders like the Universal Sentence Encoder. Results also show the promising performance of the prototype in improving search relevancy while limiting user data transmission. 2022-03-01T08:00:00Z text application/pdf https://scholar.afit.edu/etd/5327 https://scholar.afit.edu/context/etd/article/6329/viewcontent/AFIT_ENG_MS_22_M_055_Petrocelli_N.pdf Theses and Dissertations AFIT Scholar Natural language processing Search relevance Machine learning Siamese networks Computer Sciences |
| spellingShingle | Natural language processing Search relevance Machine learning Siamese networks Computer Sciences Petrocelli, Niko A. Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title_full | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title_fullStr | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title_full_unstemmed | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title_short | Improving Anonymized Search Relevance with Natural Language Processing and Machine Learning |
| title_sort | improving anonymized search relevance with natural language processing and machine learning |
| topic | Natural language processing Search relevance Machine learning Siamese networks Computer Sciences |
| url | https://scholar.afit.edu/etd/5327 https://scholar.afit.edu/context/etd/article/6329/viewcontent/AFIT_ENG_MS_22_M_055_Petrocelli_N.pdf |
| work_keys_str_mv | AT petrocellinikoa improvinganonymizedsearchrelevancewithnaturallanguageprocessingandmachinelearning |