The IUP Journal of Knowledge Management
Similarity Search Algorithms in Knowledge Management Systems: A Review

Article Details
Pub. Date : Apr, 2022
Product Name : The IUP Journal of Knowledge Management
Product Type : Article
Product Code : IJKM050422
Author Name : Sumiran Naman, Sekhar Vadari and P H Anantha Desik
Availability : YES
Subject/Domain : Management
Download Format : PDF Format
No. of Pages : 13



The similarity search algorithm plays a major role in text-based Knowledge Management (KM) systems for information retrieval, search, and acquisition. The KM systems in business application domains are mainly required to use the existing history of the application knowledge for information retrieval and for the search to be faster and results to be accurate. Typically, KM systems generate information from internal sources, and therefore require greater accuracy and faster response times. There are already many models available in the Natural Language Processing (NLP) and Deep Learning (DL) arena which support similarity search. The DL models, though are very accurate, require a lot of memory and a lot of data, and are yet slower compared to the NLP methods. This paper primarily discusses a few text-based search and similarity methods, proposes three methods which are different in nature but faster and accurate when applied to KM system of business applications. A comparison of these techniques in terms of accuracy and speed on given datasets with examples is presented.


In Knowledge Management (KM) systems of business applications, textual data plays an important role for historical data representation compared to image and video. In KM systems, similarity measures play an increasingly key role in text-related research and applications in tasks such as information retrieval, text classification, document clustering, topic detection, topic tracking, question generation, question answering, essay scoring, short answer scoring, error handling, in Chatbots KMs, text summarization and others. The text similarity algorithms will be part of Artificial Intelligence (AI)/Machine Learning (ML) model implementation for information retrieval, storing and process in KM systems (Vadari Sekhar et al., 2021).