Inter-document similarity is used for clustering, classification, and other purposes within information retrieval. In this thesis, we investigate several aspects of document similarity. In particular, we investigate the quality of several measures of inter-document similarity, providing a framework suitable for measuring and comparing the effectiveness of inter-document similarity measures. We also explore areas of research related to novelty and diversity in information retrieval. The goal of diversity and novelty is to be able to satisfy as many users as possible while simultaneously minimizing or eliminating duplicate and redundant information from search results. In order to evaluate the effectiveness of diversity-aware retrieval functions, user query logs and other information captured from user interactions with commercial search engines are mined and analyzed in order to uncover various informational aspects underlying queries, which are known as subtopics. We investigate the suitability of implicit associations between document content as an alternative to subtopic mining. We also explore subtopic mining from document anchor text and anchor links. In addition, we investigate the suitability of inter-document similarity as a measure for diversity-aware retrieval models, with the aim of using measured inter-document similarity as a replacement for diversity-aware evaluation models that rely on subtopic mining. Finally, we investigate the suitability and application of document similarity for requirements traceability. We present a fast algorithm that uncovers associations between various versions of frequently edited documents, even in the face of substantial changes.

Inter-document similarity is used for clustering, classification, and other purposes within information retrieval. In this thesis, we investigate several aspects of document similarity. In particular, we investigate the quality of several measures of inter-document similarity, providing a framework suitable for measuring and comparing the effectiveness of inter-document similarity measures. We also explore areas of research related to novelty and diversity in information retrieval. The goal of diversity and novelty is to be able to satisfy as many users as possible while simultaneously minimizing or eliminating duplicate and redundant information from search results. In order to evaluate the effectiveness of diversity-aware retrieval functions, user query logs and other information captured from user interactions with commercial search engines are mined and analyzed in order to uncover various informational aspects underlying queries, which are known as subtopics. We investigate the suitability of implicit associations between document content as an alternative to subtopic mining. We also explore subtopic mining from document anchor text and anchor links. In addition, we investigate the suitability of inter-document similarity as a measure for diversity-aware retrieval models, with the aim of using measured inter-document similarity as a replacement for diversity-aware evaluation models that rely on subtopic mining. Finally, we investigate the suitability and application of document similarity for requirements traceability. We present a fast algorithm that uncovers associations between various versions of frequently edited documents, even in the face of substantial changes.

Online scientific publications are becoming more and more popular. The number of publications we can access almost instantaneously is rapidly increasing. This makes it more challenging for researchers to pursue a topic, review literature, track research history or follow research trends. Using online resources such as search engines and digital libraries is helpful to find scientific publications, however most of the time the user ends up with an overwhelming amount of linear results to go through. This thesis proposes an alternative system, which takes advantage of citation/reference relations between publications. This demonstrates better insight of the hierarchy distribution of publications around a given topic. We also utilize information visualization techniques to represent the publications as a network. Our system is designed to automatically retrieve publications from Google Scholar and visualize them as a 2-dimensional graph representation using the citation relations. In this, the nodes represent the documents while the links represent the citation/reference relations between them. Our visualization system provides a better view of publications, making it easier to identify the research flow, connect publications, and assess similarities/differences between them. It is an interactive web based system, which allows the users to get more information about any selected publication and calculate a similarity score between two selected publications. Traditionally, similar documents are found using Natural Language Processing (NLP), which compares documents based on matching their contents. In the proposed method, similar documents are found using the citation/reference relations which are iii represented by the relationship that was originally inputted by the authors. We propose a new path based metric for measuring the similarity scores between any pair of publications. This is based on both the number of paths and the length of each path. More paths and shorter lengths increase the similarity score. We compare our similarity score results with another similarity score from Scurtu’s Document Similarity [1] that uses the NLP method. We then use the average of the similarity scores collected from 15 users as a ground truth to validate the efficiency of our method. The results indicate that our Citation Network approach yielded better scores than Scurtu’s approach.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: The World Wide Web brought with it an unprecedented level of information overload. Computers are very effective at processing and clustering numerical and binary data, however, the automated conceptual clustering of natural-language data is considerably harder to automate. Most past techniques rely on simple keyword-matching techniques or probabilistic methods to measure semantic relatedness. However, these approaches do not always accurately capture conceptual relatedness as measured by humans. In this thesis we propose and evaluate the use of novel Spreading Activation (SA) techniques for computing semantic relatedness, by modelling the article hyperlink structure of Wikipedia as an associative network structure for knowledge representation. The SA technique is adapted and several problems are addressed for it to function over the Wikipedia hyperlink structure. Inter-concept and inter-document similarity metrics are developed which make use of SA to compute the conceptual similarity between two concepts and between two natural-language documents. We evaluate these approaches over two document similarity datasets and achieve results which compare favourably with the state of the art. Furthermore, document preprocessing techniques are evaluated in terms of the performance gain these techniques can have on the well-known cosine document similarity metric and the Normalised Compression Distance (NCD) metric. Results indicate that a near two-fold increase in accuracy can be achieved for NCD by applying simple preprocessing techniques. Nonetheless, the cosine similarity metric still significantly outperforms NCD. Finally, we show that using our Wikipedia-based method to augment the cosine vector space model provides superior results to either in isolation. Combining the two methods leads to an increased correlation of Pearson p = 0:72 over the Lee (2005) document similarity dataset, which matches the reported result for the state-of-the-art Explicit Semantic Analysis (ESA) technique, while requiring less than 10% of the Wikipedia database as required by ESA. As a use case for document similarity techniques, a purely content-based news-article recommender system is designed and implemented for a large online media company. This system is used to gather additional human-generated relevance ratings which we use to evaluate the performance of three state-of-the-art document similarity metrics for providing content-based document recommendations. / AFRIKAANSE OPSOMMING: Die Wêreldwye-Web het ’n vlak van inligting-oorbelading tot gevolg gehad soos nog nooit tevore. Rekenaars is baie effektief met die verwerking en groepering van numeriese en binêre data, maar die konsepsuele groepering van natuurlike-taal data is aansienlik moeiliker om te outomatiseer. Tradisioneel berus sulke algoritmes op eenvoudige sleutelwoordherkenningstegnieke of waarskynlikheidsmetodes om semantiese verwantskappe te bereken, maar hierdie benaderings modelleer nie konsepsuele verwantskappe, soos gemeet deur die mens, baie akkuraat nie. In hierdie tesis stel ons die gebruik van ’n nuwe aktiverings-verspreidingstrategie (AV) voor waarmee inter-konsep verwantskappe bereken kan word, deur die artikel skakelstruktuur van Wikipedia te modelleer as ’n assosiatiewe netwerk. Die AV tegniek word aangepas om te funksioneer oor die Wikipedia skakelstruktuur, en verskeie probleme wat hiermee gepaard gaan word aangespreek. Inter-konsep en inter-dokument verwantskapsmaatstawwe word ontwikkel wat gebruik maak van AV om die konsepsuele verwantskap tussen twee konsepte en twee natuurlike-taal dokumente te bereken. Ons evalueer hierdie benadering oor twee dokument-verwantskap datastelle en die resultate vergelyk goed met die van ander toonaangewende metodes. Verder word teks-voorverwerkingstegnieke ondersoek in terme van die moontlike verbetering wat dit tot gevolg kan hê op die werksverrigting van die bekende kosinus vektorruimtemaatstaf en die genormaliseerde kompressie-afstandmaatstaf (GKA). Resultate dui daarop dat GKA se akkuraatheid byna verdubbel kan word deur gebruik te maak van eenvoudige voorverwerkingstegnieke, maar dat die kosinus vektorruimtemaatstaf steeds aansienlike beter resultate lewer. Laastens wys ons dat die Wikipedia-gebasseerde metode gebruik kan word om die vektorruimtemaatstaf aan te vul tot ’n gekombineerde maatstaf wat beter resultate lewer as enige van die twee metodes afsonderlik. Deur die twee metodes te kombineer lei tot ’n verhoogde korrelasie van Pearson p = 0:72 oor die Lee dokument-verwantskap datastel. Dit is gelyk aan die gerapporteerde resultaat vir Explicit Semantic Analysis (ESA), die huidige beste Wikipedia-gebasseerde tegniek. Ons benadering benodig egter minder as 10% van die Wikipedia databasis wat benodig word vir ESA. As ’n toetstoepassing vir dokument-verwantskaptegnieke ontwerp en implementeer ons ’n stelsel vir ’n aanlyn media-maatskappy wat nuusartikels aanbeveel vir gebruikers, slegs op grond van die artikels se inhoud. Joernaliste wat die stelsel gebruik ken ’n punt toe aan elke aanbeveling en ons gebruik hierdie data om die akkuraatheid van drie toonaangewende maatstawwe vir dokument-verwantskap te evalueer in die konteks van inhoud-gebasseerde nuus-artikel aanbevelings.

The purpose of this thesis is to contribute to the methodology at the intersection of relational and evaluative bibliometrics. Experimental investigations are presented that address the question of how we can most successfully produce estimates of the subject similarity between documents. The results from these investigations are then explored in the context of citation-based research evaluations in an effort to enhance existing citation normalization methods that are used to enable comparisons of subject-disparate documents with respect to their relative impact or perceived utility. This thesis also suggests and explores an approach for revealing the uncertainty and stability (or lack thereof) coupled with different kinds of citation indicators.This suggestion is motivated by the specific nature of the bibliographic data and the data collection process utilized in citation-based evaluation studies. The results of these investigations suggest that similarity-detection methods that take a global view of the problem of identifying similar documents are more successful in solving the problem than conventional methods that are more local in scope. These results are important for all applications that require subject similarity estimates between documents. Here these insights are specifically adopted in an effort to create a novel citation normalization approach that – compared to current best practice – is more in tune with the idea of controlling for subject matter when thematically different documents are assessed with respect to impact or perceived utility. The normalization approach is flexible with respect to the size of the normalization baseline and enables a fuzzy partition of the scientific literature. It is shown that this approach is more successful than currently applied normalization approaches in reducing the variability in the observed citation distribution that stems from the variability in the articles’ addressed subject matter. In addition, the suggested approach can enhance the interpretability of normalized citation counts. Finally, the proposed method for assessing the stability of citation indicators stresses that small alterations that could be artifacts from the data collection and preparation steps can have a significant influence on the picture that is painted by the citationindicator. Therefore, providing stability intervals around derived indicators prevents unfounded conclusions that otherwise could have unwanted policy implications. Together, the new normalization approach and the method for assessing the stability of citation indicators have the potential to enable fairer bibliometric evaluative exercises and more cautious interpretations of citation indicators.

Most topic models are constructed under the assumption that documents follow a multinomial distribution. The Poisson distribution is an alternative distribution to describe the probability of count data. For topic modelling, the Poisson distribution describes the number of occurrences of a word in documents of fixed length. The Poisson distribution has been successfully applied in text classification, but its application to topic modelling is not well documented, specifically in the context of a generative probabilistic model. Furthermore, the few Poisson topic models in literature are admixture models, making the assumption that a document is generated from a mixture of topics. In this study, we focus on short text. Many studies have shown that the simpler assumption of a mixture model fits short text better. With mixture models, as opposed to admixture models, the generative assumption is that a document is generated from a single topic. One topic model, which makes this one-topic-per-document assumption, is the Dirichlet-multinomial mixture model. The main contributions of this work are a new Gamma-Poisson mixture model, as well as a collapsed Gibbs sampler for the model. The benefit of the collapsed Gibbs sampler derivation is that the model is able to automatically select the number of topics contained in the corpus. The results show that the Gamma-Poisson mixture model performs better than the Dirichlet-multinomial mixture model at selecting the number of topics in labelled corpora. Furthermore, the Gamma-Poisson mixture produces better topic coherence scores than the Dirichlet-multinomial mixture model, thus making it a viable option for the challenging task of topic modelling of short text. The application of GPM was then extended to a further real-world task: that of distinguishing between semantically similar and dissimilar texts. The objective was to determine whether GPM could produce semantic representations that allow the user to determine the relevance of new, unseen documents to a corpus of interest. The challenge of addressing this problem in short text from small corpora was of key interest. Corpora of small size are not uncommon. For example, at the start of the Coronavirus pandemic limited research was available on the topic. Handling short text is not only challenging due to the sparsity of such text, but some corpora, such as chats between people, also tend to be noisy. The performance of GPM was compared to that of word2vec under these challenging conditions on labelled corpora. It was found that the GPM was able to produce better results based on accuracy, precision and recall in most cases. In addition, unlike word2vec, GPM was shown to be applicable on datasets that were unlabelled and a methodology for this was also presented. Finally, a relevance index metric was introduced. This relevance index translates the similarity distance between a corpus of interest and a test document to the probability of the test document to be semantically similar to the corpus of interest. / Thesis (PhD (Mathematical Statistics))--University of Pretoria, 2020. / Statistics / PhD (Mathematical Statistics) / Unrestricted

Throughout the history, humans continue to generate an ever-growing volume of documents about a wide range of topics. We now rely on computer programs to automatically process these vast collections of documents in various applications. Many applications require a quantitative measure of the document similarity. Traditional methods first learn a vector representation for each document using a large corpus, and then compute the distance between two document vectors as the document similarity.In contrast to this corpus-based approach, we propose a straightforward model that directly discovers the topics of a document by clustering its words, without the need of a corpus. We define a vector representation called normalized bag-of-topic-embeddings (nBTE) to encapsulate these discovered topics and compute the soft cosine similarity between two nBTE vectors as the document similarity. In addition, we propose a logistic word importance function that assigns words different importance weights based on their relative discriminating power.Our model is efficient in terms of the average time complexity. The nBTE representation is also interpretable as it allows for topic discovery of the document. On three labeled public data sets, our model achieved comparable k-nearest neighbor classification accuracy with five stateof-art baseline models. Furthermore, from these three data sets, we derived four multi-topic data sets where each label refers to a set of topics. Our model consistently outperforms the state-of-art baseline models by a large margin on these four challenging multi-topic data sets. These works together provide answers to the research question of this thesis:Can we construct an interpretable document represen-tation by clustering the words in a document, and effectively and efficiently estimate the document similarity? / Under hela historien fortsätter människor att skapa en växande mängd dokument om ett brett spektrum av publikationer. Vi förlitar oss nu på dataprogram för att automatiskt bearbeta dessa stora samlingar av dokument i olika applikationer. Många applikationer kräver en kvantitativmått av dokumentets likhet. Traditionella metoder först lära en vektorrepresentation för varje dokument med hjälp av en stor corpus och beräkna sedan avståndet mellan two document vektorer som dokumentets likhet.Till skillnad från detta corpusbaserade tillvägagångssätt, föreslår vi en rak modell som direkt upptäcker ämnena i ett dokument genom att klustra sina ord , utan behov av en corpus. Vi definierar en vektorrepresentation som kallas normalized bag-of-topic-embeddings (nBTE) för att inkapsla de upptäckta ämnena och beräkna den mjuka cosinuslikheten mellan två nBTE-vektorer som dokumentets likhet. Dessutom föreslår vi en logistisk ordbetydelsefunktion som tilldelar ord olika viktvikter baserat på relativ diskriminerande kraft.Vår modell är effektiv när det gäller den genomsnittliga tidskomplexiteten. nBTE-representationen är också tolkbar som möjliggör ämnesidentifiering av dokumentet. På tremärkta offentliga dataset uppnådde vår modell jämförbar närmaste grannklassningsnoggrannhet med fem toppmoderna modeller. Vidare härledde vi från de tre dataseten fyra multi-ämnesdatasatser där varje etikett hänvisar till en uppsättning ämnen. Vår modell överensstämmer överens med de högteknologiska baslinjemodellerna med en stor marginal av fyra utmanande multi-ämnesdatasatser. Dessa arbetsstöd ger svar på forskningsproblemet av tisthesis:Kan vi konstruera en tolkbar dokumentrepresentation genom att klustra orden i ett dokument och effektivt och effektivt uppskatta dokumentets likhet?

In this paper, we propose a novel similarity measure for document data processing and apply it to text classification and clustering. For two documents, the proposed measure takes three cases into account: (a) The feature considered appears in both documents, (b) the feature considered appears in only one document, and (c) the feature considered appears in none of the documents. For the first case, we give a lower bound and decrease the similarity according to the difference between the feature values of the two documents. For the second case, we give a fixed value disregarding the magnitude of the feature value. For the last case, we ignore its effectiveness. We apply it to the similarity based single-label classifier k-NN and multi-label classifier ML-KNN, and adopt these properties to measure the similarity between a document and a specific set for document clustering, i.e., k-means like algorithm, to compare the effectiveness with other measures. Experimental results show that our proposed method can work more effectively than others.

Help documents are important in an organization to use the technology applications licensed from a vendor. Customers and internal employees frequently use and interact with the help documents section to use the applications and know about the new features and developments in them. Help documents consist of various knowledge base materials, question and answer documents and help content. In day- to-day life, customers go through these documents to set up, install or use the product. Recommending similar documents to the customers can increase customer engagement in the product and can also help them proceed without any hurdles. The main aim of this study is to build a recommendation system by exploring different machine-learning techniques to recommend the most relevant and similar help document to the user. To achieve this, in this study a hybrid-based recommendation system for help documents is proposed where the documents are recommended based on similarity of the content using content-based filtering and similarity between the users using collaborative filtering. Finally, the recommendations from content-based filtering and collaborative filtering are combined and ranked to form a comprehensive list of recommendations. The proposed approach is evaluated by the internal employees of the company and by external users. Our experimental results demonstrate that the proposed approach is feasible and provides an effective way to recommend help documents.