1,289 research outputs found
A Clustering-based Approach for Discovering Flaws in Requirements Specifications
No full textIn this paper, we present the application of a clustering algorithm to exploit lexical and syntactic relationships occurring between natural language requirements. Our experiments conducted on a real-world data set highlight a correlation between clustering outliers, i.e., requirements that are marked as "noisy" by the clustering algorithm, and requirements presenting "flaws". Those flaws may refer to an incomplete explanation of the behavioral aspects, which the requirement is supposed to provide. Furthermore, flaws may also be caused by the usage of inconsistent terminology in the requirement specification. We evaluate the ability of our proposed algorithm to effectively discover such kind of flawed requirements. Evaluation is performed by measuring the accuracy of the algorithm in detecting a set of flaws in our testing data set, which have been previously manually-identified by a human assessor
Mining Lifecycle Event Logs for Enhancing Service-based Applications
No full textService-Oriented Architectures (SOAs), and traditional enterprise systems in general, record a variety of events (e.g., messages being sent and received between service components) to proper log files, i.e., event logs. These files constitute a huge and valuable source of knowledge that may be extracted through data mining techniques. To this end, process mining is increasingly gaining interest across the SOA community. The goal of process mining is to build models without a priori knowledge, i.e., to discover structured process models derived from specific patterns that are present in actual traces of service executions recorded in event logs. However, in this work, the authors focus on detecting frequent sequential patterns, thus considering process mining as a specific instance of the more general sequential pattern mining problem. Furthermore, they apply two sequential pattern mining algorithms to a real event log provided by the Vienna Runtime Environment for Service-oriented Computing, i.e., VRESCo. The obtained results show that the authors are able to find services that are frequently invoked together within the same sequence. Such knowledge could be useful at design-time, when service-based application developers could be provided with service recommendation tools that are able to predict and thus to suggest next services that should be included in the current service composition
Detecting Task-Based Query Sessions Using Collaborative Knowledge
No full textOur research challenge is to provide a mechanism for splitting into user task-based sessions a long-term log of queries submitted to a Web Search Engine (WSE). The hypothesis is that some query sessions entail the concept of user task. We present an approach that relies on a centroid-based and a density-based clustering algorithm, which consider queries inter-arrival times and use a novel distance function that takes care of query lexical content and exploits the collaborative knowledge collected by Wiktionary and Wikipedia
Enhancing web search user experience : from document retrieval to task recommendation
No full textThe World Wide Web is the biggest and most heterogeneous database that humans have ever built, making it the place of choice where people search for any sort of information through Web search engines. Indeed, users are increasingly asking Web search engines for performing their daily tasks (e.g., "planning holidays", "obtaining a visa", "organizing a birthday party", etc.), instead of simply looking for Web pages. In this Ph.D. dissertation, we sketch and address two core research challenges that we claim next-generation Web search engines should tackle for enhancing user search experience, i.e., Web task discovery and Web task recommendation. Both these challenges rely on the actual understanding of user search behaviors, which can be achieved by mining knowledge from query logs. Search processes of many users are analyzed at a higher level of abstraction, namely from a "task-by-task" instead of a "query-by-query" perspective, thereby producing a model of user search tasks, which in turn can be used to support people during their daily "Web lives".Il World Wide Web è la più grande sorgente dati mai realizzata dall’uomo. Ciò ha fatto sì che il Web divenisse sempre più il “luogo” di riferimento per accedere a qualsiasi tipo di informazione, attraverso l’uso dei motori di ricerca. Infatti, gli utenti tendono a rivolgersi ai motori di ricerca non solo per consultare pagine Web ma per eseguire vere e proprie attività (ad es., per organizzare vacanze, ottenere un visto, organizzare una festa, etc.). In questa tesi di dottorato, si descrivono e affrontano due sfide fondamentali tese a migliorare l’esperienza di ricerca sul Web offerta dagli attuali motori di ricerca, ovvero la scoperta e la raccomandazione di cosiddetti “Web tasks”. Entrambe queste sfide si basano su una reale comprensione dei comportamenti di ricerca degli utenti, che può essere raggiunta mediante l’applicazione di tecniche di query log mining. I processi di ricerca degli utenti sono analizzati ad un più alto livello di astrazione, ovvero da una prospettiva “task-by-task” anziché “query-by-query”. In questo modo è possible realizzare un modello di attività di ricerca che fornisca adeguato supporto alla “vita sul Web” degli utenti
Twitter anticipates bursts of requests for Wikipedia articles
No full textMost of the tweets that users exchange on Twitter make implicit mentions of named-entities, which in turn can be mapped to corresponding Wikipedia articles using proper Entity Linking (EL) techniques. Some of those become trending entities on Twitter due to a long-lasting or a sudden effect on the volume of tweets where they are mentioned. We argue that the set of trending entities discovered from Twitter may help predict the volume of requests for relating Wikipedia articles. To validate this claim, we apply an EL technique to extract trending entities from a large dataset of public tweets. Then, we analyze the time series derived from the hourly trending score (i.e., an index of popularity) of each entity as measured by Twitter and Wikipedia, respectively. Our results reveals that Twitter actually leads Wikipedia by one or more hours
- …
