Saved in:
Bibliographic Details
Main Authors: Senthilnathan, Annamalai, Arumae, Kristjan, Khalilia, Mohammed, Xing, Zhengzheng, Colak, Aaron R.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.12458
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915157571010560
author Senthilnathan, Annamalai
Arumae, Kristjan
Khalilia, Mohammed
Xing, Zhengzheng
Colak, Aaron R.
author_facet Senthilnathan, Annamalai
Arumae, Kristjan
Khalilia, Mohammed
Xing, Zhengzheng
Colak, Aaron R.
contents Analyzing long text data such as customer call transcripts is a cost-intensive and tedious task. Machine learning methods, namely Transformers, are leveraged to model agent-customer interactions. Unfortunately, Transformers adhere to fixed-length architectures and their self-attention mechanism scales quadratically with input length. Such limitations make it challenging to leverage traditional Transformers for long sequence tasks, such as conversational understanding, especially in real-time use cases. In this paper we explore and evaluate recently proposed efficient Transformer variants (e.g. Performer, Reformer) and a CNN-based architecture for real-time and near real-time long conversational understanding tasks. We show that CNN-based models are dynamic, ~2.6x faster to train, ~80% faster inference and ~72% more memory efficient compared to Transformers on average. Additionally, we evaluate the CNN model using the Long Range Arena benchmark to demonstrate competitiveness in general long document analysis.
format Preprint
id arxiv_https___arxiv_org_abs_2502_12458
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle An Empirical Evaluation of Encoder Architectures for Fast Real-Time Long Conversational Understanding
Senthilnathan, Annamalai
Arumae, Kristjan
Khalilia, Mohammed
Xing, Zhengzheng
Colak, Aaron R.
Computation and Language
Analyzing long text data such as customer call transcripts is a cost-intensive and tedious task. Machine learning methods, namely Transformers, are leveraged to model agent-customer interactions. Unfortunately, Transformers adhere to fixed-length architectures and their self-attention mechanism scales quadratically with input length. Such limitations make it challenging to leverage traditional Transformers for long sequence tasks, such as conversational understanding, especially in real-time use cases. In this paper we explore and evaluate recently proposed efficient Transformer variants (e.g. Performer, Reformer) and a CNN-based architecture for real-time and near real-time long conversational understanding tasks. We show that CNN-based models are dynamic, ~2.6x faster to train, ~80% faster inference and ~72% more memory efficient compared to Transformers on average. Additionally, we evaluate the CNN model using the Long Range Arena benchmark to demonstrate competitiveness in general long document analysis.
title An Empirical Evaluation of Encoder Architectures for Fast Real-Time Long Conversational Understanding
topic Computation and Language
url https://arxiv.org/abs/2502.12458