Search results for: cosine similarity

Authors: Ning Kang, Marius Doornenbal

Abstract:

Funding is a very important – if not crucial – resource for research projects. Usually, funding organizations will publish a description of the funded research to describe the scope of the funding award. Logically, we would expect research outcomes to align with this funding award. For that reason, we might be able to predict future research topics based on present funding award data. That said, it remains to be shown if and how future research topics can be predicted by using the funding information. In this paper, we extract funding project information and their generated paper abstracts from the Gateway to Research database as a group, and use the papers from the same domains and publication years in the Scopus database as a baseline comparison group. We annotate both the project awards and the papers resulting from the funded projects with linguistic features (noun phrases), and then calculate tf-idf and cosine similarity between these two set of features. We show that the cosine similarity between the project-generated papers group is bigger than the project-baseline group, and also that these two groups of similarities are significantly different. Based on this result, we conclude that the funding information actually correlates with the content of future research output for the funded project on the topical level. How funding really changes the course of science or of scientific careers remains an elusive question.

Keywords: natural language processing, noun phrase, tf-idf, cosine similarity

Procedia PDF Downloads 245

Authors: Debarag Banerjee, Pooja Singh, Arjun Avadhanam, Saksham Srivastava

Abstract:

Autonomous conversational agents, i.e., chatbots, are becoming an increasingly common mechanism for enterprises to provide support to customers and partners. In order to rate chatbots, especially ones powered by Generative AI tools like Large Language Models (LLMs), we need to be able to accurately assess their performance. This is where chatbot benchmarking becomes important. In this paper, authors propose the use of a benchmark that they call the E2E (End to End) benchmark and show how the E2E benchmark can be used to evaluate the accuracy and usefulness of the answers provided by chatbots, especially ones powered by LLMs. The authors evaluate an example chatbot at different levels of sophistication based on both our E2E benchmark as well as other available metrics commonly used in the state of the art and observe that the proposed benchmark shows better results compared to others. In addition, while some metrics proved to be unpredictable, the metric associated with the E2E benchmark, which uses cosine similarity, performed well in evaluating chatbots. The performance of our best models shows that there are several benefits of using the cosine similarity score as a metric in the E2E benchmark.

Keywords: chatbot benchmarking, end-to-end (E2E) benchmarking, large language model, user centric evaluation.

Procedia PDF Downloads 66

Authors: Amal Alshahrani, Allan Ramsay

Abstract:

Paraphrasing is one of the important tasks in natural language processing; i.e. alternative ways to express the same concept by using different words or phrases. Paraphrases can be used in many natural language applications, such as Information Retrieval, Machine Translation, Question Answering, Text Summarization, or Information Extraction. To obtain pairs of sentences that are paraphrases we create a system that automatically extracts paraphrases from a corpus, which is built from different sources of news article since these are likely to contain paraphrases when they report the same event on the same day. There are existing simple standard approaches (e.g. TF-IDF vector space, cosine similarity) and alignment technique (e.g. Dynamic Time Warping (DTW)) for extracting paraphrase which have been applied to the English. However, the performance of these approaches could be affected when they are applied to another language, for instance Arabic language, due to the presence of phenomena which are not present in English, such as Free Word Order, Zero copula, and Pro-dropping. These phenomena will affect the performance of these algorithms. Thus, if we can analysis how the existing algorithms for English fail for Arabic then we can find a solution for Arabic. The results are promising.

Keywords: natural language processing, TF-IDF, cosine similarity, dynamic time warping (DTW)

Procedia PDF Downloads 386

Authors: Sandeep Kaur Chouhan, Jatinderdeep Kaur, S. S. Bhatia

Abstract:

The existence of sine and cosine series as a Fourier series, their L1-convergence seems to be one of the difficult question in theory of convergence of trigonometric series in L1-metric norm. In the literature so far available, various authors have studied the L1-convergence of cosine and sine trigonometric series with special coefficients. In this paper, we present a modified cosine and sine sums and criterion for L1-convergence of these modified sums is obtained. Also, a necessary and sufficient condition for the L1-convergence of the cosine and sine series is deduced as corollaries.

Keywords: conjugate Dirichlet kernel, Dirichlet kernel, L1-convergence, modified sums

Procedia PDF Downloads 354

Authors: Gaohuizi Guo, Ning Zhang

Abstract:

Firefly algorithm (FA) and Sine Cosine algorithm (SCA) are two very popular and advanced metaheuristic algorithms. However, these algorithms applied to multi-objective optimization problems have some shortcomings, respectively, such as premature convergence and limited exploration capability. Combining the privileges of FA and SCA while avoiding their deficiencies may improve the accuracy and efficiency of the algorithm. This paper proposes a hybridization of FA and SCA algorithms, named multi-objective firefly-sine cosine algorithm (MFA-SCA), to develop a more efficient meta-heuristic algorithm than FA and SCA.

Keywords: firefly algorithm, hybrid algorithm, multi-objective optimization, sine cosine algorithm

Procedia PDF Downloads 168

Authors: Jatinderdeep Kaur

Abstract:

In this paper, the results of Kano from one-dimensional cosine and sine series are extended to two-dimensional cosine and sine series. To extend these results, some classes of coefficient sequences such as the class of semi convexity and class R are extended from one dimension to two dimensions. Under these extended classes, I have checked the function f(x,y) is two dimensional Fourier Cosine and Sine series or equivalently it represents an integrable function. Further, some results are obtained which are the generalization of Moricz's results.

Keywords: conjugate dirichlet kernel, conjugate fejer kernel, fourier series, semi-convexity

Procedia PDF Downloads 439

Authors: Shashwat Mookherjee, Shruti Dutta

Abstract:

This study explores linguistic distinctions among American, Indian, and Irish English dialects and assesses various Language Models (LLMs) in their ability to generate British English translations from these dialects. Using cosine similarity analysis, the study measures the linguistic proximity between original British English translations and those produced by LLMs for each dialect. The findings reveal that Indian and Irish English translations maintain notably high similarity scores, suggesting strong linguistic alignment with British English. In contrast, American English exhibits slightly lower similarity, reflecting its distinct linguistic traits. Additionally, the choice of LLM significantly impacts translation quality, with Llama-2-70b consistently demonstrating superior performance. The study underscores the importance of selecting the right model for dialect translation, emphasizing the role of linguistic expertise and contextual understanding in achieving accurate translations.

Keywords: cross-dialect translation, language models, linguistic similarity, multilingual NLP

Procedia PDF Downloads 75

Authors: Ali Cheraghian, Farshid Hajati, Soheila Gheisari, Yongsheng Gao

Abstract:

In this paper, we present a novel 2.5D face recognition method based on Gabor Discrete Cosine Transform (GDCT). In the proposed method, the Gabor filter is applied to extract feature vectors from the texture and the depth information. Then, Discrete Cosine Transform (DCT) is used for dimensionality and redundancy reduction to improve computational efficiency. The system is combined texture and depth information in the decision level, which presents higher performance compared to methods, which use texture and depth information, separately. The proposed algorithm is examined on publically available Bosphorus database including models with pose variation. The experimental results show that the proposed method has a higher performance compared to the benchmark.

Keywords: Gabor filter, discrete cosine transform, 2.5d face recognition, pose

Procedia PDF Downloads 328

Authors: H. G. C. P. Dinesh, G. Tharshini, M. P. B. Ekanayake, G. M. R. I. Godaliyadda

Abstract:

This paper presents a new spatial feature extraction method based on principle component analysis (PCA) and Fisher Discernment Analysis (FDA) for facial expression recognition. It not only extracts reliable features for classification, but also reduces the feature space dimensions of pattern samples. In this method, first each gray scale image is considered in its entirety as the measurement matrix. Then, principle components (PCs) of row vectors of this matrix and variance of these row vectors along PCs are estimated. Therefore, this method would ensure the preservation of spatial information of the facial image. Afterwards, by incorporating the spectral information of the eigen-filters derived from the PCs, a feature vector was constructed, for a given image. Finally, FDA was used to define a set of basis in a reduced dimension subspace such that the optimal clustering is achieved. The method of FDA defines an inter-class scatter matrix and intra-class scatter matrix to enhance the compactness of each cluster while maximizing the distance between cluster marginal points. In order to matching the test image with the training set, a cosine similarity based Bayesian classification was used. The proposed method was tested on the Cohn-Kanade database and JAFFE database. It was observed that the proposed method which incorporates spatial information to construct an optimal feature space outperforms the standard PCA and FDA based methods.

Keywords: facial expression recognition, principle component analysis (PCA), fisher discernment analysis (FDA), eigen-filter, cosine similarity, bayesian classifier, f-measure

Procedia PDF Downloads 425

Authors: Kriuk Boris, Kriuk Fedor

Abstract:

This paper presents a comparative study of fundamental similarity functions for Siamese networks in semantic textual similarity (STS) tasks. We evaluate various similarity functions using the STS Benchmark dataset, analyzing their performance and stability. Additionally, we introduce a multi-objective approach for optimal threshold selection. Our findings provide insights into the effectiveness of different similarity functions and offer a straightforward method for threshold selection optimization, contributing to the advancement of Siamese network architectures in STS applications.

Keywords: siamese networks, semantic textual similarity, similarity functions, STS benchmark dataset, threshold selection

Procedia PDF Downloads 37

Authors: Qiang Yang, Ping-An Du

Abstract:

Group decision making with multiple attribute has attracted intensive concern in the decision analysis area. This paper assumes that the contributions of all the decision makers (DMs) are not equal to the decision process based on different knowledge and experience in group setting. The aim of this paper is to develop a novel approach to determine weights of DMs in the group decision making problems. In this paper, the weights of DMs are determined in the group decision environment via angle cosine and projection method. First of all, the average decision of all individual decisions is defined as the ideal decision. After that, we define the weight of each decision maker (DM) by aggregating the angle cosine and projection between individual decision and ideal decision with associated direction indicator μ. By using the weights of DMs, all individual decisions are aggregated into a collective decision. Further, the preference order of alternatives is ranked in accordance with the overall row value of collective decision. Finally, an example in a chemical company is provided to illustrate the developed approach.

Keywords: angel cosine, ideal decision, projection method, weights of decision makers

Procedia PDF Downloads 377

Authors: Umair Rashid

Abstract:

Normally softmax loss is used as the supervision signal in face recognition (FR) system, and it boosts the separability of features. In the last two years, a number of techniques have been proposed by reformulating the original softmax loss to enhance the discriminating power of Deep Convolutional Neural Networks (DCNNs) for FR system. To learn angularly discriminative features Cosine-Margin based softmax has been adjusted as monotonically decreasing angular function, that is the main challenge for angular based softmax. On that issue, we propose monotonically decreasing element for Cosine-Margin based softmax and also, we discussed the effect of different monotonically decreasing parameters on angular Margin softmax for FR system. We train the model on publicly available dataset CASIA- WebFace via our proposed monotonically decreasing parameters for cosine function and the tests on YouTube Faces (YTF, Labeled Face in the Wild (LFW), VGGFace1 and VGGFace2 attain the state-of-the-art performance.

Keywords: deep convolutional neural networks, cosine margin face recognition, softmax loss, monotonically decreasing parameter

Procedia PDF Downloads 101

Authors: Doru Anastasiu Popescu, Dan Rădulescu

Abstract:

In this paper, we determine the similarity of two HTML web applications. We are going to use a genetic algorithm in order to determine the most significant web pages of each application (we are not going to use every web page of a site). Using these significant web pages, we will find the similarity value between the two applications. The algorithm is going to be efficient because we are going to use a reduced number of web pages for comparisons but it will return an approximate value of the similarity. The binary trees are used to keep the tags from the significant pages. The algorithm was implemented in Java language.

Keywords: Tag, HTML, web page, genetic algorithm, similarity value, binary tree

Procedia PDF Downloads 355

Authors: Madiha Khan, Sidrah Ramzan, Seemab Khan, Shahzad Hassan, Kamran Saeed

Abstract:

Measuring semantic similarity between texts is calculating semantic relatedness between texts using various techniques. Our web application (Measuring Relatedness of Concepts-MRC) allows user to input two text corpuses and get semantic similarity percentage between both using WordNet. Our application goes through five stages for the computation of semantic relatedness. Those stages are: Preprocessing (extracts keywords from content), Feature Extraction (classification of words into Parts-of-Speech), Synonyms Extraction (retrieves synonyms against each keyword), Measuring Similarity (using keywords and synonyms, similarity is measured) and Visualization (graphical representation of similarity measure). Hence the user can measure similarity on basis of features as well. The end result is a percentage score and the word(s) which form the basis of similarity between both texts with use of different tools on same platform. In future work we look forward for a Web as a live corpus application that provides a simpler and user friendly tool to compare documents and extract useful information.

Keywords: Graphviz representation, semantic relatedness, similarity measurement, WordNet similarity

Procedia PDF Downloads 237

Authors: Adnan A. Y. Mustafa

Abstract:

In this paper we present a quick technique to measure the similarity between binary images. The technique is based on a probabilistic mapping approach and is fast because only a minute percentage of the image pixels need to be compared to measure the similarity, and not the whole image. We exploit the power of the Probabilistic Matching Model for Binary Images (PMMBI) to arrive at an estimate of the similarity. We show that the estimate is a good approximation of the actual value, and the quality of the estimate can be improved further with increased image mappings. Furthermore, the technique is image size invariant; the similarity between big images can be measured as fast as that for small images. Examples of trials conducted on real images are presented.

Keywords: big images, binary images, image matching, image similarity

Procedia PDF Downloads 196

Authors: Guang-Ho Cha

Abstract:

This paper presents a context-sensitive media similarity search algorithm. One of the central problems regarding media search is the semantic gap between the low-level features computed automatically from media data and the human interpretation of them. This is because the notion of similarity is usually based on high-level abstraction but the low-level features do not sometimes reflect the human perception. Many media search algorithms have used the Minkowski metric to measure similarity between image pairs. However those functions cannot adequately capture the aspects of the characteristics of the human visual system as well as the nonlinear relationships in contextual information given by images in a collection. Our search algorithm tackles this problem by employing a similarity measure and a ranking strategy that reflect the nonlinearity of human perception and contextual information in a dataset. Similarity search in an image database based on this contextual information shows encouraging experimental results.

Keywords: context-sensitive search, image search, similarity ranking, similarity search

Procedia PDF Downloads 365

Authors: Liang-Ta Cheng, Ching-Yu Yang

Abstract:

Based on fast discrete cosine transform (FDCT), the authors present a high capacity and high perceived quality method for electrocardiogram (ECG) signal. By using a simple adjusting policy to the 1-dimentional (1-D) DCT coefficients, a large volume of secret message can be effectively embedded in an ECG host signal and be successfully extracted at the intended receiver. Simulations confirmed that the resulting perceived quality is good, while the hiding capability of the proposed method significantly outperforms that of existing techniques. In addition, our proposed method has a certain degree of robustness. Since the computational complexity is low, it is feasible for our method being employed in real-time applications.

Keywords: data hiding, ECG steganography, fast discrete cosine transform, 1-D DCT bundle, real-time applications

Procedia PDF Downloads 194

Authors: Delara Kazempour, Mashallah Abasi Dezfuli, Reza Javidan

Abstract:

Video compression used for channels with limited bandwidth and storage devices has limited storage capabilities. One of the most popular approaches in video compression is the usage of different transforms. Discrete cosine transform is one of the video compression methods that have some problems such as blocking, noising and high distortion inappropriate effect in compression ratio. wavelet transform is another approach is better than cosine transforms in balancing of compression and quality but the recognizing of curve curvature is so limit. Because of the importance of the compression and problems of the cosine and wavelet transforms, the contourlet transform is most popular in video compression. In the new proposed method, we used contourlet transform in video image compression. Contourlet transform can save details of the image better than the previous transforms because this transform is multi-scale and oriented. This transform can recognize discontinuity such as edges. In this approach we lost data less than previous approaches. Contourlet transform finds discrete space structure. This transform is useful for represented of two dimension smooth images. This transform, produces compressed images with high compression ratio along with texture and edge preservation. Finally, the results show that the majority of the images, the parameters of the mean square error and maximum signal-to-noise ratio of the new method based contourlet transform compared to wavelet transform are improved but in most of the images, the parameters of the mean square error and maximum signal-to-noise ratio in the cosine transform is better than the method based on contourlet transform.

Keywords: video compression, contourlet transform, discrete cosine transform, wavelet transform

Procedia PDF Downloads 443

Authors: Gi Ryung Song, Kyoung Seok Kim

Abstract:

This study reviewed the literature that focused on similarity of various characteristics such as values, personality, or demographics between employee and other elements in its organization for example employee with leader, job, and organization. We divided a body of this study into two parts and organized and demonstrated recent studies in first part. Three issues appeared in this part, which are statistical ways of measuring similarity, supervisor-subordinate similarity, and person-organization fit with person-job fit. In the latter part, based on the three issues of recent studies, we suggested three propositions about points that the recent studies missed or the studies did not orient. First proposition argued about the direction of similarity, which could also be interpreted as there is causal relation between employee and its workplace environments. Second, we suggested a consideration of eliminating common variance buried in one’s characteristics or its profiles. Third proposition was about the similarity of extra role behavior between individual and organization, and we treated this organization’s level of extra role behavior as a kind of its culture. In doing so, similarity of individual’s extra role behavior and organization’s has the meaning that individual’s congruence against their organization culture.

Keywords: similarity, person-organization fit, supervisor-subordinate similarity, literature review

Procedia PDF Downloads 283

Authors: Fatimah Zawani Abdullah, Shereena Mohd Arif, Nurul Malim

Abstract:

The study of molecular similarity search in chemical database is increasingly widespread, especially in the area of drug discovery. Similarity search is an application in the field of Chemoinformatics to measure the similarity between the molecular structure which is known as the query and the structure of chemical compounds in the database. Similarity search is also one of the approaches in virtual screening which involves computational techniques and scoring the probabilities of activity. The main objective of this work is to determine the best fingerprint when compared to the other five fingerprints selected in this study using PubChem chemical dataset. This paper will discuss the similarity searching process conducted using 6 types of descriptors, which are ECFP4, ECFC4, FCFP4, FCFC4, SRECFC4 and SRFCFC4 on 15 activity classes of PubChem dataset using Tanimoto coefficient to calculate the similarity between the query structures and each of the database structure. The results suggest that ECFP4 performs the best to be used with Tanimoto coefficient in the PubChem dataset.

Keywords: 2D fingerprints, Tanimoto, PubChem, similarity searching, chemoinformatics

Procedia PDF Downloads 293

Authors: M. Kamel El-Sayed

Abstract:

The process of determining the degree of membership for an element to an uncertain concept has been found in many ways, using equivalence and symmetry relations in information systems. In the case of similarity, these methods did not take into account the degree of symmetry between elements. In this paper, we use a new definition for finding the membership based on the degree of symmetry. We provide an example to clarify the suggested methods and compare it with previous methods. This method opens the door to more accurate decisions in information systems.

Keywords: information system, uncertain concept, membership function, similarity relation, degree of similarity

Procedia PDF Downloads 223

Authors: Salima Kouici, Abdelkader Khelladi

Abstract:

In this work, we begin with the presentation of the Tθ family of usual similarity measures concerning multidimensional binary data. Subsequently, some properties of these measures are proposed. Finally, the impact of the use of different inter-elements measures on the results of the Agglomerative Hierarchical Clustering Methods is studied.

Keywords: binary data, similarity measure, Tθ measures, agglomerative hierarchical clustering

Procedia PDF Downloads 481

Authors: Abdelkrim Alfalah, Lahcen Ouarbya, John Howroyd

Abstract:

This paper investigates and compares the performance of four existing distances and similarity measures between partitions. The partition measures considered are Rand Index (RI), Adjusted Rand Index (ARI), Variation of Information (VI), and Normalised Variation of Information (NVI). This work investigates the ability of these partition measures to capture three predefined intuitions: the variation within randomly generated partitions, the sensitivity to small perturbations, and finally the independence from the dataset scale. It has been shown that the Adjusted Rand Index performed well overall, with regards to these three intuitions.

Keywords: clustering, comparing partitions, similarity measure, partition distance, partition metric, similarity between partitions, clustering comparison.

Procedia PDF Downloads 202

Authors: K. P. Sandesh, M. H. Suman

Abstract:

Text processing plays an important role in information retrieval, data-mining, and web search. Measuring the similarity between the documents is an important operation in the text processing field. In this project, a new similarity measure is proposed. To compute the similarity between two documents with respect to a feature the proposed measure takes the following three cases into account: (1) The feature appears in both documents; (2) The feature appears in only one document and; (3) The feature appears in none of the documents. The proposed measure is extended to gauge the similarity between two sets of documents. The effectiveness of our measure is evaluated on several real-world data sets for text classification and clustering problems, especially in banking and health sectors. The results show that the performance obtained by the proposed measure is better than that achieved by the other measures.

Keywords: document classification, document clustering, entropy, accuracy, classifiers, clustering algorithms

Procedia PDF Downloads 518

Authors: Doru Anastasiu Popescu, Raducanu Dragos Ionut

Abstract:

In this paper the presentation of a tool which measures the similarity between two websites is made. The websites are compound only from webpages created with HTML. The tool uses three ways of calculating the similarity between two websites based on certain results already published. The first way compares all the webpages within a website, the second way compares a webpage with all the pages within the second website and the third way compares two webpages. Java programming language and technologies such as spring, Jsoup, log4j were used for the implementation of the tool.

Keywords: Java, Jsoup, HTM, spring

Procedia PDF Downloads 385

Authors: Shakila N. Senarath, Dinesh Asanka, Janaka Wijayanayake

Abstract:

The necessity of cosmetic products is arising to fulfill consumer needs of personality appearance and hygiene. A cosmetic product consists of various chemical ingredients which may help to keep the skin healthy or may lead to damages. Every chemical ingredient in a cosmetic product does not perform on every human. The most appropriate way to select a healthy cosmetic product is to identify the texture of the body first and select the most suitable product with safe ingredients. Therefore, the selection process of cosmetic products is complicated. Consumer surveys have shown most of the time, the selection process of cosmetic products is done in an improper way by consumers. From this study, a content-based system is suggested that recommends cosmetic products for the human factors. To such an extent, the skin type, gender and price range will be considered as human factors. The proposed system will be implemented by using Machine Learning. Consumer skin type, gender and price range will be taken as inputs to the system. The skin type of consumer will be derived by using the Baumann Skin Type Questionnaire, which is a value-based approach that includes several numbers of questions to derive the user’s skin type to one of the 16 skin types according to the Bauman Skin Type indicator (BSTI). Two datasets are collected for further research proceedings. The user data set was collected using a questionnaire given to the public. Those are the user dataset and the cosmetic dataset. Product details are included in the cosmetic dataset, which belongs to 5 different kinds of product categories (Moisturizer, Cleanser, Sun protector, Face Mask, Eye Cream). An alternate approach of TF-IDF (Term Frequency – Inverse Document Frequency) is applied to vectorize cosmetic ingredients in the generic cosmetic products dataset and user-preferred dataset. Using the IF-IPF vectors, each user-preferred products dataset and generic cosmetic products dataset can be represented as sparse vectors. The similarity between each user-preferred product and generic cosmetic product will be calculated using the cosine similarity method. For the recommendation process, a similarity matrix can be used. Higher the similarity, higher the match for consumer. Sorting a user column from similarity matrix in a descending order, the recommended products can be retrieved in ascending order. Even though results return a list of similar products, and since the user information has been gathered, such as gender and the price ranges for product purchasing, further optimization can be done by considering and giving weights for those parameters once after a set of recommended products for a user has been retrieved.

Keywords: content-based filtering, cosmetics, machine learning, recommendation system

Procedia PDF Downloads 134

Authors: Deokho Kim, Wonwoo Lee, Jaewoong Lee, Teresa Ng, Gun-Ill Lee, Jiwon Jeong

Abstract:

This paper presents a method for improving object search accuracy using a deep learning model. A major limitation to provide accurate similarity with deep learning is the requirement of huge amount of data for training pairwise similarity scores (metrics), which is impractical to collect. Thus, similarity scores are usually trained with a relatively small dataset, which comes from a different domain, causing limited accuracy on measuring similarity. For this reason, this paper proposes a deep learning model that can be trained with a significantly small amount of data, a clustered data which of each cluster contains a set of visually similar images. In order to measure similarity distance with the proposed method, visual features of two images are extracted from intermediate layers of a convolutional neural network with various pooling methods, and the network is trained with pairwise similarity scores which is defined zero for images in identical cluster. The proposed method outperforms the state-of-the-art object similarity scoring techniques on evaluation for finding exact items. The proposed method achieves 86.5% of accuracy compared to the accuracy of the state-of-the-art technique, which is 59.9%. That is, an exact item can be found among four retrieved images with an accuracy of 86.5%, and the rest can possibly be similar products more than the accuracy. Therefore, the proposed method can greatly reduce the amount of training data with an order of magnitude as well as providing a reliable similarity metric.

Keywords: visual search, deep learning, convolutional neural network, machine learning

Procedia PDF Downloads 215

Authors: Ian A. McCulloh, Madelaine Zinser, Jesse Patsolic, Michael Ramos

Abstract:

Recommender systems are a common artificial intelligence (AI) application. For any given input, a search system will return a rank-ordered list of similar items. As users review returned items, they must decide when to halt the search and either revise search terms or conclude their requirement is novel with no similar items in the database. We present a statistically designed experiment that investigates the impact of similarity ratings on human judgement to conclude a search item is novel and halt the search. 450 participants were recruited from Amazon Mechanical Turk to render judgement across 12 decision tasks. We find the inclusion of ratings increases the human perception that items are novel. Percent similarity increases novelty discernment when compared with star-rated similarity or the absence of a rating. Ratings reduce the time to decide and improve decision confidence. This suggests the inclusion of similarity ratings can aid human decision-makers in knowledge search tasks.

Keywords: ratings, rankings, crowdsourcing, empirical studies, user studies, similarity measures, human-centered computing, novelty in information retrieval

Procedia PDF Downloads 131

Authors: Omid Shahmirzadi, Adam Lugowski, Kenneth Younge

Abstract:

Automatic measurement of semantic text similarity is an important task in natural language processing. In this paper, we evaluate the performance of different vector space models to perform this task. We address the real-world problem of modeling patent-to-patent similarity and compare TFIDF (and related extensions), topic models (e.g., latent semantic indexing), and neural models (e.g., paragraph vectors). Contrary to expectations, the added computational cost of text embedding methods is justified only when: 1) the target text is condensed; and 2) the similarity comparison is trivial. Otherwise, TFIDF performs surprisingly well in other cases: in particular for longer and more technical texts or for making finer-grained distinctions between nearest neighbors. Unexpectedly, extensions to the TFIDF method, such as adding noun phrases or calculating term weights incrementally, were not helpful in our context.

Keywords: big data, patent, text embedding, text similarity, vector space model

Procedia PDF Downloads 175

Authors: Musaab Riyadh, Norwati Mustapha, Dina Riyadh

Abstract:

Trajectory similarity can be defined as the cost of transforming one trajectory into another based on certain similarity method. It is the core of numerous mining tasks such as clustering, classification, and indexing. Various approaches have been suggested to measure similarity based on the geometric and dynamic properties of trajectory, the overlapping between trajectory segments, and the confined area between entire trajectories. In this article, an evaluation of these approaches has been done based on computational cost, usage memory, accuracy, and the amount of data which is needed in advance to determine its suitability to stream mining applications. The evaluation results show that the stream mining applications support similarity methods which have low computational cost and memory, single scan on data, and free of mathematical complexity due to the high-speed generation of data.

Keywords: global distance measure, local distance measure, semantic trajectory, spatial dimension, stream data mining

Procedia PDF Downloads 396