TY - GEN
T1 - ELECTRA-KG
T2 - 19th International Conference on Services Computing, SCC 2022, held as part of the Services Conference Federation, SCF 2022
AU - Kwapong, Benjamin
AU - Sen, Amartya
AU - Fletcher, Kenneth K.
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Knowledge graphs (KGs) are becoming popular in recommender systems in recent times because of the wealth of side information they provide. Many researchers rely on KGs to help resolve the issues of cold start, diversity, and explainability in recommendations. However, the existing approaches usually ignore entity descriptions, which are essential in providing content information for entities in KGs. In this work, we propose a contextual language model for KG completion known as ELECTRA-KG (Efficiently Learning an Encoder that Classifies Token Replacements Accurately). We formulate the recommendation task as a KG link prediction task where we have an incomplete knowledge graph and we use state-of-the-art approaches to complete it. We do this by identifying missing facts among entities from our test data. To evaluate and validate our method, we perform a couple of experiments. First, we run experiments to demonstrate how well our model compares to state-of-the-art KG embedding models. Second, we run further experiments with our model on the tag recommendation task and compare our results to existing baselines. Our results show that our model outperforms the existing baselines on the tag recommendation task.
AB - Knowledge graphs (KGs) are becoming popular in recommender systems in recent times because of the wealth of side information they provide. Many researchers rely on KGs to help resolve the issues of cold start, diversity, and explainability in recommendations. However, the existing approaches usually ignore entity descriptions, which are essential in providing content information for entities in KGs. In this work, we propose a contextual language model for KG completion known as ELECTRA-KG (Efficiently Learning an Encoder that Classifies Token Replacements Accurately). We formulate the recommendation task as a KG link prediction task where we have an incomplete knowledge graph and we use state-of-the-art approaches to complete it. We do this by identifying missing facts among entities from our test data. To evaluate and validate our method, we perform a couple of experiments. First, we run experiments to demonstrate how well our model compares to state-of-the-art KG embedding models. Second, we run further experiments with our model on the tag recommendation task and compare our results to existing baselines. Our results show that our model outperforms the existing baselines on the tag recommendation task.
KW - Deep learning
KW - Language modeling
KW - Recommender systems
KW - Reinforcement learning
KW - word2vec
UR - https://www.scopus.com/pages/publications/85146665127
UR - https://www.scopus.com/pages/publications/85146665127#tab=citedBy
U2 - 10.1007/978-3-031-23515-3_5
DO - 10.1007/978-3-031-23515-3_5
M3 - Conference contribution
AN - SCOPUS:85146665127
SN - 9783031235146
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 56
EP - 70
BT - Services Computing – SCC 2022 - 19th International Conference, Held as Part of the Services Conference Federation, SCF 2022, Proceedings
A2 - Qingyang, Wang
A2 - Zhang, Liang-Jie
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 10 December 2022 through 14 December 2022
ER -