Personalizing Hybrid-Based Dialogue Agents
In this paper, we present a continuation of our work on the personification of dialogue agents. We expand upon the previously demonstrated models—the ranking and generative models—and propose new hybrid models. Because there is no single definitive way to build a hybrid model, we explore various arc...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/24/4657 |
| _version_ | 1797456530763677696 |
|---|---|
| author | Yuri Matveev Olesia Makhnytkina Pavel Posokhov Anton Matveev Stepan Skrylnikov |
| author_facet | Yuri Matveev Olesia Makhnytkina Pavel Posokhov Anton Matveev Stepan Skrylnikov |
| author_sort | Yuri Matveev |
| collection | DOAJ |
| description | In this paper, we present a continuation of our work on the personification of dialogue agents. We expand upon the previously demonstrated models—the ranking and generative models—and propose new hybrid models. Because there is no single definitive way to build a hybrid model, we explore various architectures where the components adopt different roles, sequentially and in parallel. Applying the perplexity and BLEU performance metrics, we discover that the Retrieve and Refine and KG model—a modification of the Retrieve and Refine model where the ranking and generative components work in parallel and compete based on the proximity of the candidate found by the ranking model with a knowledge-grounded generation block—achieves the best performance, with values of 1.64 for perplexity and 0.231 for BLEU scores. |
| first_indexed | 2024-03-09T16:08:13Z |
| format | Article |
| id | doaj.art-84d574e74fbb44548b3a39daefbb32af |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-09T16:08:13Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-84d574e74fbb44548b3a39daefbb32af2023-11-24T16:27:30ZengMDPI AGMathematics2227-73902022-12-011024465710.3390/math10244657Personalizing Hybrid-Based Dialogue AgentsYuri Matveev0Olesia Makhnytkina1Pavel Posokhov2Anton Matveev3Stepan Skrylnikov4Information Technologies and Programming Faculty, ITMO University, 197101 Saint Petersburg, RussiaInformation Technologies and Programming Faculty, ITMO University, 197101 Saint Petersburg, RussiaInformation Technologies and Programming Faculty, ITMO University, 197101 Saint Petersburg, RussiaInformation Technologies and Programming Faculty, ITMO University, 197101 Saint Petersburg, RussiaInformation Technologies and Programming Faculty, ITMO University, 197101 Saint Petersburg, RussiaIn this paper, we present a continuation of our work on the personification of dialogue agents. We expand upon the previously demonstrated models—the ranking and generative models—and propose new hybrid models. Because there is no single definitive way to build a hybrid model, we explore various architectures where the components adopt different roles, sequentially and in parallel. Applying the perplexity and BLEU performance metrics, we discover that the Retrieve and Refine and KG model—a modification of the Retrieve and Refine model where the ranking and generative components work in parallel and compete based on the proximity of the candidate found by the ranking model with a knowledge-grounded generation block—achieves the best performance, with values of 1.64 for perplexity and 0.231 for BLEU scores.https://www.mdpi.com/2227-7390/10/24/4657personalized dialogue systemshybrid modelsretrieve modelsrefine models |
| spellingShingle | Yuri Matveev Olesia Makhnytkina Pavel Posokhov Anton Matveev Stepan Skrylnikov Personalizing Hybrid-Based Dialogue Agents Mathematics personalized dialogue systems hybrid models retrieve models refine models |
| title | Personalizing Hybrid-Based Dialogue Agents |
| title_full | Personalizing Hybrid-Based Dialogue Agents |
| title_fullStr | Personalizing Hybrid-Based Dialogue Agents |
| title_full_unstemmed | Personalizing Hybrid-Based Dialogue Agents |
| title_short | Personalizing Hybrid-Based Dialogue Agents |
| title_sort | personalizing hybrid based dialogue agents |
| topic | personalized dialogue systems hybrid models retrieve models refine models |
| url | https://www.mdpi.com/2227-7390/10/24/4657 |
| work_keys_str_mv | AT yurimatveev personalizinghybridbaseddialogueagents AT olesiamakhnytkina personalizinghybridbaseddialogueagents AT pavelposokhov personalizinghybridbaseddialogueagents AT antonmatveev personalizinghybridbaseddialogueagents AT stepanskrylnikov personalizinghybridbaseddialogueagents |