Traditional Keyword-based Approaches Research Articles

Towards Effective and Efficient Sparse Neural Information Retrieval

Sparse representation learning based on Pre-trained Language Models has seen a growing interest in Information Retrieval. Such approaches can take advantage of the proven efficiency of inverted indexes and inherit desirable IR priors such as explicit lexical matching or some degree of interpretability. In this work, we thoroughly develop the framework of sparse representation learning in IR, which unifies term weighting and expansion in a supervised setting. We then build on SPLADE—a sparse expansion-based retriever—and show to which extent it is able to benefit from the same training improvements as dense bi-encoders by studying the effect of distillation, hard negative mining, as well as the Pre-trained Language Model’s initialization on its effectiveness , leading to state-of-the-art results in both in- and out-of-domain evaluation settings (SPLADE++). We furthermore propose efficiency improvements, allowing us to reach latency requirements on par with traditional keyword-based approaches (Efficient-SPLADE).

A comprehensive approach to reviewing latent topics addressed by literature across multiple disciplines

This paper proposes an approach to capturing and reviewing scientific literature addressing latent topics across multiple scientific fields. As latent topics like moral values are affected by word polysemy and synonymy, a traditional keyword-based approach is often ineffective and therefore inappropriate. As a result, scientific literature addressing latent topics tends to be fragmented thereby constraining efforts to address similar and complementary research challenges. A novel approach to reviewing the literature by utilizing both semantic fields and probabilistic topic models has therefore been developed. We illustrate this approach by reviewing the literature addressing the value justice in the energy sector and compare this with a regular keyword-based approach. The new approach results in a more complete overview of the relevance of energy justice as compared to the traditional keyword-based approach. This novel approach can be applied to other latent topics including other values or phenomena such as societal resistance to technologies, thereby leading to an increased understanding of existing relevant literature and the identification of new areas of research.

An Ontology-based Term Weighting Technique for Web Document Categorization

Web document classification has become crucial as there has been a massive increase in the magnitude of web pages across the web. In the research community, an efficient approach to this problem is based on machine learning techniques. Ontology forms the heart of knowledge representation for any domain. This paper proposes an ontology-based term weighting technique which is novel and efficient for the classification of web pages. The proposed approach builds domain ontology and selects the features that significantly improve the prediction performance. Experiments were conducted on domain based web pages and classification performance was calculated with state of the art classification algorithms. The experimental analysis demonstrates that the proposed approach produces significantly better results compared to the traditional keyword-based approaches.

An Efficient Content based Image Retrieval: CBIR

Due to the exponential growth of image data there is a dire need for innovative tools which can easily manage, retrieve images and images from the large image database. The most common approach which is being used is Content-Based Image Retrieval (CBIR) system. CBIR is the popular image retrieval system by which reterived the targetted image can be retrieved by matching the features of the given image. The goal of this paper is to develop an image retrieval based on content properties such as shape, color, texture etc. usually encoded into feature vectors. One of the main advantages of the CBIR approach is the possibility of an automatic retrieval process instead of the traditional keyword-based approach. The CBIR technology has been used in several applications such as fingerprint identification, biodiversity information systems, digital libraries, medicine and historical research among others. This paper aims to develop a new efficient tool for CBIR based on above mention parameters using MATLAB.

Ontology-based Document Mining System for IT Support Service

Information Technology (IT) is a vital and an integral part of every organization. IT executives are constantly faced with problems that are difficult to tackle and time consuming. Experience is required to solve these problems easier and faster. We can utilize case-based reasoning (CBR), data mining and information retrieval (IR) techniques to automate IT problem solving and experience management. In this paper, we propose an IT ontology-based system for semantic retrieval that increases the efficiency and quality of IT support service. The proposed approach retrieves similar problem/solution pairs based on the concepts in the query and performs better than the traditional keyword-based approach especially in cases where the keywords of the relevant documents do not match the keywords in the query.

Web Information at Your Fingertips: Paper as an Interaction Metaphor

The prototype O system seamlessly integrates Web searching and browsing on touch-enabled tablets using two-handed gesture interaction, much like handling and marking content on a piece of paper. The system also leverages contextual information to infer user intent and significantly improve search results over traditional keyword-based approaches, enabling users to go beyond websites' "walled garden"' and more fully explore information objects.

Content Based Image Retrieval Using Spatial Features

The field of image retrieval technologies have enabled the creation of large image datasets. It is necessary to develop appropriate information systems to efficiently manage these collections. An image retrieval system is a computer system for searching, browsing and retrieving images from a large database of images.An approach known as Content-Based Image Retrieval (CBIR) systems used to retrieve images similar to a user-defined specification or pattern (image, sketch etc). Their goal is to support image retrieval based on content properties (colour, texture and shape), usually encoded into feature vectors. CBIR approach is an automatic retrieval process, instead of the traditional keyword-based approach, which usually is very time-consuming. The CBIR technology has been used in several applications such as, digital libraries, crime prevention, medicine, historical research, fingerprint identification, biodiversity information systems.

Sharing, finding and reusing end-user code for reformatting and validating data

To help users with automatically reformatting and validating spreadsheets and other datasets, prior work introduced a user-extensible data model called “topes” and a supporting visual programming language. However, no support has existed to date for users to exchange and reuse topes. This functional gap results in wasteful duplication of work as users implement topes that other people have already created. In this paper, a design for a new repository system is presented that supports sharing and finding of topes for reuse. This repository tightly integrates traditional keyword-based search with two additional search methods whose usefulness in repositories of end-user code has gone unexplored to date. The first method is “search-by-match”, where a user specifies examples of data, and the repository retrieves topes that can reformat and validate that data. The second method is collaborative filtering, which has played a vital role in repositories of non-code artifacts. The repository’s search functionality was empirically tested on a prototype repository implementation by simulating queries generated from real user spreadsheets. This experiment reveals that search-by-match and collaborative filtering greatly improve the accuracy of search over the traditional keyword-based approach, to a recall as high as 95%. These results show that search-by-match and collaborative filtering are useful approaches for helping users to publish, find, and reuse visual programs similar to topes.