Information Retreival

Although sketch-to-photo retrieval has a wide range of applications, it is costly to obtain paired and rich-labeled ground truth. Differently, photo retrieval data is easier to acquire. Therefore, previous works pre-train their models on rich-labeled photo retrieval data (i.e., source domain) and then fine-tune them on the limited-labeled sketch-to-photo retrieval data (i.e., target domain). However, without co-training source and target data, source domain knowledge might be forgotten during the fine-tuning process, while simply co-training them may cause negative transfer due to domain gaps. Moreover, identity label spaces of source data and target data are generally disjoint and therefore conventional category-level Domain Adaptation (DA) is not directly applicable. To address these issues, we propose an Instance-level Heterogeneous Domain Adaptation (IHDA) framework. We apply the fine-tuning strategy for identity label learning, aiming to transfer the instance-level knowledge in an inductive transfer manner. Meanwhile, labeled attributes from the source data are selected to form a shared label space for source and target domains. Guided by shared attributes, DA is utilized to bridge cross-dataset domain gaps and heterogeneous domain gaps, which transfers instance-level knowledge in a transductive transfer manner. Experiments show that our method has set a new state of the art on three sketch-to-photo image retrieval benchmarks without extra annotations, which opens the door to train more effective models on limited-labeled heterogeneous image retrieval tasks. Related codes are available at https://github.com/fandulu/IHDA.

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Many real-world applications increasingly involve both structured data and text. Hence, managing both in an efficient and integrated manner has received much attention from both the IR and database communities. To date, however, little research has been devoted to semantic issues in the integration of text and data. In this paper we introduced a problem in this realm: entity retrieval. Given data fragments that describe various aspects of a real-world entity, find all other data fragments as well as text documents that describe ...

Web yang tidak terstruktur saat ini berisi berjuta-juta dokumen yang sulit di query dan struktur yang bercampur. Data terstruktur menjadi suatu hal yang diperlukan dalam suatu sistem informasi yang saling terhubung. Dengan Semantic Web, semua informasi yang tersedia pada suatu web akan menjadi lebih terstruktur dan berkualitas. Metode yang dapat digunakan untuk dapat mengekstrak data dari halaman web dan mengubahnya menjadi data yang lebih terstruktur adalah dengan menggunakan suatu modul yang disebut wrapper. Pada awalnya proses wrapping menggunakan metode manual. Yaitu dengan beberapa tahapan, yang pertama menerima halaman web, kemudian yang kedua melakukan proses parsing dari halaman web tersebut (format HTML). Proses parsing ini tidak akan mengeluarkan semua tag HTML, tapi memilih hanya pada struktur dokumen, format logical, dan format fisik. Dan yang ketiga adalah menempatkan informasi yang telah diperoleh ke dalam bentuk yang lebih terstruktur, misalnya dalam format XML. Permasalahan yang timbul adalah terkadang menyulitkan untuk memahami isi dari masing-masing halaman web tersebut serta sangat membutuhkan tenaga kerja yang banyak. Untuk mengatasi kesulitan tersebut penulis mengembangkan suatu aplikasi wrapper yang digunakan untuk mengekstrak data yang semi-terstruktur menjadi data yang terstruktur. Yang akan memudahkan dalam menentukan class dan properties dari suatu data pada halaman web, sehingga nantinya akan memudahkan dalam memahami isi dari suatu halaman web. Aplikasi ini dibuat dengan menggunakan bahasa pemrograman Python 2.5.

Text mining is a practice which is regarded as the supporting pillars of Information Retreival. This paper is in simple terms dedicated to text mining and bear a prime focus on mining academic papers. An architecture is proposed by the authors is presented in the paper, which they have named HTPI. This framework is built upon Java eclipse using Apache Hadoop. The problem under consideration for the paper is the reference metamorphosis of the references mentioned in the references section of any scientific paper based upon the similarity score(between the referenced paper and the paper whose reference list is being reordered) retrieved. Various notions have been used in the paper like stemming, skipping and similarity calculation using Jaccard Coefficient.

Autism Spectrum Disorder (ASD) is a developmental disorder that often impairs a child's normal development of the brain. According to CDC, it is estimated that 1 in 6 children in the US suffer from development disorders, and 1 in 68 children in the US suffer from ASD. This condition has a negative impact on a person's ability to hear, socialize and communicate. Overall, ASD has a broad range of symptoms and severity; hence the term spectrum is used. One of the main contributors to ASD is known to be genetics. Up to date, no suitable cure for ASD has been found. Early diagnosis is crucial for the long-term treatment of ASD, but this is challenging due to the lack of a proper objective measures. Subjective measures often take more time, resources, and have false positives or false negatives. There is a need for efficient objective measures that can help in diagnosing this disease early as possible with less effort. EEG measures the electric signals of the brain via electrodes placed on various places on the scalp. These signals can be used to study complex neuropsychiatric issues. Studies have shown that EEG has the potential to be used as a biomarker for various neurological conditions including ASD. This chapter will outline the usage of EEG measurement for the classification of ASD using machine learning algorithms.

Algoritma adalah urutan langkah logis yang digunakan untuk menyelesaikan suatu masalah. Singkatnya, sebuah masalah harus diselesaikan dengan beberapa Langkah logis. Dalam kehidupan sehari-hari, contoh algoritma bias digambarkan dengan berbagai macam hal/aktivitas. Sementara itu dalam ilmu computer matematika, algoritm di definisikan sebagai prosedur dari berbagai macam langkah perhitungan, penalaran otomatis, sampai pemrosesan data. Algoritma juga diartikan sebagai metofe yang diekspresikan lewat rangkaian dan instruksi yang di jabarkan untuk menghitung. Saecara garis besar, algoritma dimulai dengan input dan kondisi awal, deskripsi, eksekusi dan menghasilkan output.String matching adalah suatu algoritma yang digunakan untuk memecahkan masalah pencocokan suatu teks terhadap suatu teks lain. Disini penulis akan mencoba mengimplementasikan string matching untuk mencari suatu teks dalam beberapa dokumen agar memudahkan proses pencarian serta waktu jadi lebih efisien.Algoritma pencarian (searching algorithm) adalah algoritma yang menerima sebuah argument kunci dan dengan langkah-langkah tertentu akan mencari rekaman dengan kunci tersebut. Setelah proses pencarian dilaksanakan, akan diperoleh salah satu dari dua kemungkinan, yaitu data yang dicari ditemukan (succesfull) atau tidak ditemukan (unsuccesfull). Ada dua macam teknik pencarian yaitu pencarian skuensial (sequential search) dan pencarian biner (binary search). Perbedaan dari teknik ini terletak pada keadaan data. Pencarian skuensial digunakan apabila data dalam keadaan acak atau tidak terurut. Sebaliknya pencarian biner digunakan pada data yang sudah dalam keadaan urut. String matching merupakan salah satu algoritma yang bersifat fleksibel dan banyak diaplikasikan dalam beberapa program komputer karena dapat disesuaikan dengan kebutuhan spesifik pengguna. Metode string matching juga merupakan salah satu metode yang efektif untuk menemukan kemunculan kata atau pola pada teks. String matching juga dapat diimplementasikan dengan mudah meskipun kemampuannya yang relatif sederhana. Pencarian teks dalam koleksi dokumen merupakan proses menemukan dan mengindentifikasi dokumen yang mengandung informasi atau kata kunci tertentu. Dalam pencarian tersebut penulis menggunakan metode string matching. Namun keberhasilan pencarian juga bergantung pada keakuratan metode yang digunakan serta kemampuannya untuk menangani kompleksitas dokumen dan variasi kueri pengguna. Selain itu, pencarian teks dalam koleksi dokumen juga telah banyak dikembangkan oleh beberapa perusahaan teknologi ternama seperti WPS Office dan Ms Word.

Data miningmerupakan proses penggalian informasi tersembunyi dari suatu data yang berukuran besar. Polapola yang ditemukan dalam data mining dapat menjadi suatu informasi yang bermanfaat untuk digunakan selanjutnya seperti pada sistem pendukung keputusan. Kasus data mining masih menjadi kasus yang menarik untuk diteliti oleh para peneliti. Banyaknya software yang tersedia baik yang bersifat open source maupun berbayar yang digunakan dalam membantu dalam pengolahan data mining membuat peneliti dapat dengan mudah memilih software yang sesuai dengan kebutuhan kasus yang sedang diteliti dan kemampuan pengeluaran dari peneliti itu sendiri. Pada penulisan paper ini penulis hendak membahas perbandingan antara software berbayar dan software open source dimana penulis mengambil satu buah sample berbayar yaitu IBM SPSS Modeler dan satu buah sample yang bersifat open source yaitu WEKA. Kedua software tersebut banyak digunakan dalam membantu menyelesaikan proses data mining.

Search engines are tools used to find information on the Internet. Since the web has a plethora of websites, the engine queries the majority of active sites and builds a database organized according to keywords utilized in the search. Because of this, when a user types a few descriptive words on the home page of the search engine, the search function lists websites corresponding to these keywords. However, there are some problems with this search approach. For instance, if a user wants information about the word Jaguar, most search results are animals and cars. This is a polysemic problem that forces search engines to always provide the most popular but not the most relevant results. This article presents a study of using sentiment technology to help news classification and categorization and improve the classification accuracy. We have introduced a smart search function embedded into a search engine to tackle polysemic issues and record relevant results to determine their sentiment...

In this survey paper, we reviewed the current state of the art in medical question answering (MedQuAn). The MedQuAn systems are concerned with providing relevant answers in response to questions proposed in natural language. This survey tried to review the MedQuAn conceptual model, in which the question answering (QA) is therefore composed of three different components, each of which has a core component beside other supplementary components and method/approach research. These three core components are question classification, document retrieval, and answer extraction. The final result of this survey is a contribution to the development of future research in the domain of MedQuAn especially for Indonesian medical question answering system.

Search engines are tools used to fnd information on the Internet. Since the web has a plethora of websites, the engine queries the majority of active sites and builds a database organized according to keywords utilized in the search. Because of this, when a user types a few descriptive words on the home page of the search engine, the search function lists websites corresponding to these keywords. However, there are some problems with this search approach. For instance, if a user wants information about the word Jaguar, most search results are animals and cars. Tis is a polysemic problem that forces search engines to always provide the most popular but not the most relevant results. Tis article presents a study of using sentiment technology to help news classifcation and categorization and improve the classifcation accuracy. We have introduced a smart search function embedded into a search engine to tackle polysemic issues and record relevant results to determine their sentimentality. Terefore, this study presents a topic that involves several aspects of natural language processing (NLP) and sentiment analysis for news categorization and classifcation. A web crawler was used to collect British Broadcasting Corporation (BBC) news across the Internet, carried out preprocessing of text by using NLP, and applied sentiment analysis methods to determine the polarity of the processed text data. Te sentimentality represents negative, positive, or neutral polarities assigned by the sentiment analysis algorithms. Te research utilized the BBC news site to collect diferent information using a web crawler and a database to explore the sentimentality of BBC news. Te natural language toolkit (NLTK) and BM25 indexed and preprocessed patterns in the database. Te experimental results depict the proposed search function surpassing normal search with an accuracy rate of 85%. Moreover, the results show a negative polarity of BBC news using the Sentistrength algorithm. Furthermore, the Valence Aware Dictionary and sEntiment Reasoner (VADER) was the best-performing sentiment analysis model for news classifcation. Tis model obtained an accuracy of 85% using data collected with the proposed smart function.

How to quickly and reliably learn the preferences of new users remains a key challenge in the design of recommender systems. In this paper we introduce a new type of online learning algorithm, cluster-based bandits, to address this challenge. This exploits the fact that users can often be grouped into clusters based on the similarity of their preferences, and this allows accelerated learning of new user preferences since the task becomes one of identifying which cluster a user belongs to and typically there are far fewer clusters than there are items to be rated. Clustering by itself is not enough however. Intra-cluster variability between users can be thought of as adding noise to user ratings. Deterministic methods such as decision-trees perform poorly in the presence of such noise. We identify so-called distinguisher items that are particularly informative for deciding which cluster a new user belongs to despite the rating noise. Using these items the cluster-based bandit algorithm is able to efficiently adapt to user responses and rapidly learn the correct cluster to assign to a new user.

This paper describes an extension of our work presented in the robust English-to-French bilingual task of the CLEF 2007 workshop, a knowledge-light approach for query translation in Cross-Language Information Retrieval systems. Our work is based on the direct translation of character n-grams, avoiding the need for word normalization during indexing or translation, and also dealing with out-of-vocabulary words. Moreover, since such a solution does not rely on language-specific processing, it can be used with languages of very different nature even when linguistic information and resources are scarce or unavailable. The results obtained have been very positive, and support the findings from our previous English-to-Spanish experiments.

DOOH(Digital Out Of Home) advertising targets for reaching consumers through outdoor digital display medias. Traditionally, scheduling of advertisement contents over DOOH medias is usually done by operator's strategy, but an efficient ad scheduling strategy is not easy to find under various advertising contexts. In this paper, we present a context-aware factorization machine-based recommendation model for the scheduling under various advertising contexts, and provide analysis for understanding of the contexts' effects on advertising based on the recommendation model. Through simulation results on the dataset adapted from a real dataset of RecSys challenge 2015, it is shown that the proposed model and analysis based on the model will be effective for better scheduling of ad contents under advertising contexts over DOOH networks.

Edit distance merupakan jumlah minimum point mutation yang diperlukan untuk merubah suatu string ke string yang lain. Point mutation tersebut adalah mengganti, menambah dan menghapus sebuah karakter. Konsep edit distance banyak digunakan dalam proses manipulasi data berbagai aplikasi komputer. Berbagai algoritma dapat digunakan dalam pencarian dan penentuan edit distance. Dalam penelitian ini, dilakukan eksperimen untuk mencari edit distance menggunakan algoritma dynamic programming. Sedangkan proses perhitungannya dibantu dengan fungsi-fungsi yang ada dalam Microsoft Excel. Hasil eksperimen menunjukkan kompleksitas waktu (time complexity) untuk algoritma dynamic programming adalah O(|s1|*|s2|). Jika s1 dan s2 mempunyai panjang yang hampir sama, katakanlah "n", maka kompleksitas waktunya adalah O(n 2). Kompleksitas ruang (space-complexity) untuk algoritma ini juga O(n 2), karena keseluruhan dari matriks digunakan untuk menemukan solusi yang optimal.

Cet article s'intéresse à la description et l'évaluation d'une nouvelle méthode d'apprentissage non supervisé pour réunir des définitions en espagnol selon leur signification. Nous utilisons comme mesure de regroupement l'énergie textuelle et nous étudions une adaptation de la précision et le rappel afin d'évaluer notre méthode.

Sentiment Analyses are widely used approaches to understand and identify emotions, feelings, and opinion on social media platforms. Most sentiment analysis systems measure the presumed emotional polarity of texts. While this is sufficient for some applications, these approaches are very limiting when it comes to understanding how social media users actually use language resources to make sense of extreme events. In this paper, the authors apply a Sentiment Analysis based on the Appraisal System from the theory of communication called Systemic Functional Linguistics to understand the sentiment of event-driven social media communication. A prototype was developed to code and visualise geotagged Twitter data using the Appraisal System. This prototype was applied to tweets collected during and after the Sydney Siege, a hostage situation in a busy café in Sydney’s inner city at the 15th of December 2014. Because the Appraisal System is a theorised functional communication method, the res...

On-line information services have become widespread in the Web nowadays. However, Web users are non-specialized and have a great variety of interests. Thus, interfaces for Web databases must be simple and uniform. In this paper we present an approach, based on Bayesian networks, for querying Web databases using keywords only. According to this approach, the user inputs a query through a simple searchbox interface. From the input query, one or more plausible structured queries are derived and submitted to Web databases. The results are then retrieved and presented to the user as ranked answers. Our approach reduces the complexity of existing on-line interfaces and offers a solution to the problem of querying several distinct Web databases with a single interface. The applicability of the proposed approach was demonstrated by experimental results with 3 databases, obtained with a prototype search system that implements it. We have found that from 77% to 95% of the time, one of the top three resulting structured queries is the proper one. Further, when the user selects one of these three top queries for processing, the ranked answers present average precision figures from 60% to about 100%.

Social media marketing plays a vital role in promoting brand and product values to wide audiences. In order to boost their advertising revenues, global media buying platforms such as Facebook Ads constantly reduce the reach of branded organic posts, pushing brands to spend more on paid media ads. In order to run organic and paid social media marketing efficiently, it is necessary to understand the audience, tailoring the content to fit their interests and online behaviours, which is impossible to do manually at a large scale. At the same time, various personality type categorization schemes such as the Myers-Briggs Personality Type indicator make it possible to reveal the dependencies between personality traits and user content preferences on a wider scale by categorizing audience behaviours in a unified and structured manner. Still, McKinsey-style manual categorization is a very labour-intensive task that is probably impractical in a real-world scenario, so automated incorporation of audience behaviour and personality mining into industrial applications is necessary. This problem is yet to be studied in depth by the research community, while the level of impact of different personality traits on content recommendation accuracy has not been widely utilised and comprehensively evaluated so far. Even worse, there is no dataset available for the research community to serve as a benchmark and drive further research in this direction. The present study is one of the first attempts to bridge this important industrial gap, contributing not just a novel personality-driven content recommendation approach and dataset, but also facilitating a real-world ready solution which is scalable and sufficiently accurate to be applied in real-world settings. Specifically, in this work we investigate the impact of human personality traits on the content recommendation model by applying a novel personality-driven multi-view content recommender system called Personality Content Marketing Recommender Engine, or PersiC. Our experimental results and real-world case study demonstrate not just PersiC's ability to perform efficient human personality-driven multi-view content recommendation, but also allow for actionable

Text mining is a practice which is regarded as the supporting pillars of Information Retreival. This paper is in simple terms dedicated to text mining and bear a prime focus on mining academic papers. An architecture is proposed by the authors is presented in the paper, which they have named HTPI. This framework is built upon Java eclipse using Apache Hadoop. The problem under consideration for the paper is the reference metamorphosis of the references mentioned in the references section of any scientific paper based upon the similarity score(between the referenced paper and the paper whose reference list is being re-ordered) retrieved. Various notions have been used in the paper like stemming, skipping and similarity calculation using Jaccard Coefficient.

This paper presents an approach for extending the vector space model (VSM) to perform XML retrieval. The model is extended to support important aspects of XML structural and semantic information such as element nesting level, matching tag names in the query and the collection and the relation between tag names and content of an element. Potential use of the model for heterogeneous as well as for the unstructured collection is also shown. We compared our model with the standard vector space model and obtained a gain for unstructured and structured queries. For unstructured collections the vector space model effectiveness is preserved.

La supervision d'evenements a travers les reseaux sociaux ont connu un engouement certain ces dernieres annees. Cependant, le nombre considerable de messages publies rend difficile, voire impossible, pour une personne de suivre "ce qui se passe autour de l'evenement". Le filtrage et la generation en temps reel d'une synthese de messages importants portant sur l'evenement permet de repondre a ce besoin. La generation de cette synthese pose plusieurs problemes qui rend cette tâche difficile. En effet, la synthese doit etre concise, non redondante et couvrant le maximum de sous evenements, et ce, en sus de la pertinence des informations qu'elle contient. Dans cet article, nous proposons une nouvelle approche de selection des messages pour la generation, en temps reel, de resumes de flux de messages courts. La decision de selectionner un message est prise instantanement, un nouveau message est ajoute au resume si ses scores afferents a l'informativite e...

Data on the web page can be available in various formats, such as table. With the growing of web pages, the need to extract data from tables is increasing. Results of the extraction can be used for integration with other web tables or stored in a database. This study discusses the extraction of data from a table on a web page using a Document Object Model (DOM) tree. The initial step of this extraction process is to transform the HTML document into a DOM tree. Then, by applying search methods Depth First Search (DFS), part of the data in the table is extracted and stored in a CSV file. An engine has been developed using Visual Basic. The results show that the engine can automatically extract data from the table that has the following characteristics: the number of rows and columns are not limited, able to handle all of the table orientation layout, and able to handle tables that are merged cells.

In recent years, user's trust has gained attention in recommender systems. Trust plays a vital role in the recommendation of online products. Trust is a dynamic feature which evolves with passage of time and varies from person to person. Trust-based cross domain recommender systems suggest items to the users usually by ratings, provided by similar users, often not available in the same domain. However, due to the sparse rating scores, recommender systems cannot generate up-to-the-mark recommendations. In this research, we solved a user cold start problem, mainly by modeling preference drift on a temporal basis. We tried to solve this problem by adopting one of the scenarios of cross domain of 'No Overlap' using cross domain information. In this work, we proposed a model called Trust Aware Cross Domain Temporal Recommendations (TrustCTR) that predict the rating of an item about an active user from the most recent time. We generated user features and item features by using latent factor model and trained the proposed model. We also introduced the concept of trust relevancy that shows the degree of trust, computed the trusted neighbors in target domain for an active user belonging to a source domain, and predicted the ratings of items for cold start users. We performed experiments on public datasets Ciao and Epinions and used these datasets in cross domain form such as the categories of Ciao as source domain and Epinions as the target domain. We selected five different domains, having a higher proportion of rating sparsity, for observing the performance of our approach using MAE, RMSE, and F-measure. Our approach is a viable solution of cold start problem and offers effective recommendations We also compared the model with state-of-the-art methods; the model generates satisfactory results.

Feature Maximization is a feature selection method that deals efficiently with textual data: to design systems that are altogether language-agnostic, parameter-free and do not require additional corpora to function. We propose to evaluate its use in text summarization, in particular in cases where documents are structured. We first experiment this approach in a single-document summarization context. We evaluate it on the DUC AQUAINT corpus and show that despite the unstructured nature of the corpus, our system is above the baseline and produces encouraging results. We also observe that the produced summaries seem robust to redundancy. Next, we evaluate our method in the more appropriate context of SciSumm challenge, which is dedicated to research publications summarization. These publications are structured in sections and our class-based approach is thus relevant. We more specifically focus on the task that aims to summarize papers using those that refer to them. We consider and evaluate several systems using our approach dealing with specific bag of words. Furthermore, in these systems, we also evaluate cosine and graph-based distance for sentence weighting and comparison. We show that our Feature Maximization based approach performs very well in the SciSumm 2016 context for the considered task, providing better results than the known results so far, and obtaining high recall. We thus B Nicolas Dugué

La F-Mesure de trait est une metrique de selection de variables statis-tique sans parametres qui a montre de bonnes performances pour la classification , l'etiquetage de clusters ou encore la mesure de qualite des clusters. Dans cet article, nous proposons d'evaluer son utilisation dans le contexte des graphes de terrain et de leur structure communautaire pour beneficier de son systeme sans parametres et de ses performances bien evaluees. Nous etudions donc sur des graphes synthetiques realistes les correlations qui existent entre la F-Mesure de trait et certaines mesures de centralite, mais surtout avec des mesures destinees a caracteriser le role communautaire des noeuds. Nous montrons ainsi que cette mesure est liee a la centralite des noeuds du reseau, et qu'elle est particulierement adaptee a la mesure de leur connectivite au regard de la structure de communautes. Nous observons par ailleurs que les mesures usuelles de detection des roles communautaires sont forteme...

Pemanfaatan data atau berita yang tersebar di internet untuk meningkatkan peluang keberhasilan dalam sebuah usaha melalui analisa trend pasar adalah hal yang sangat umum pada saat ini. Penjelajahan Web (Crawl) dan ekstraksi data dari web (Scraping) menjadi salah satu hal yang penting, agar tidak terjadi data yang kurang sempurna, dan data yang diterima adalah data yang paling baru. CSS Selector dan Xpath merupakan salah satu metode yang umum digunakan dalam melakukan proses crawling. Terdapat perbedaan dari jumlah data yang terambil, besar file output dan waktu pemrosesan dari kedua metode tersebut, dimana Xpath memiliki keunggulan pada jumlah data yang terambil dan waktu pemrosesnya yang berakibat pada ukuran file output yang lebih besar. Sedangkan untuk penggunaan memori pada kedua metode pada proses crawling tidak memiliki perbedaan yang signifikan.

Autism spectrum disorder (ASD) is a developmental disorder that often impairs a child's normal development of the brain. According to CDC, it is estimated that 1 in 6 children in the US suffer from development disorders, and 1 in 68 children in the US suffer from ASD. This condition has a negative impact on a person's ability to hear, socialize, and communicate. Subjective measures often take more time, resources, and have false positives or false negatives. There is a need for efficient objective measures that can help in diagnosing this disease early as possible with less effort. EEG measures the electric signals of the brain via electrodes placed on various places on the scalp. These signals can be used to study complex neuropsychiatric issues. Studies have shown that EEG has the potential to be used as a biomarker for various neurological conditions including ASD. This chapter will outline the usage of EEG measurement for the classification of ASD using machine learning ...

Dans le cadre de la veille ou de l'analyse prospective, il est très courant d'avoir recours aux méthodes de clustering pour traiter de gros volumes de données textuelles. Les algorithmes de clustering affichent généralement de bonnes performances dans le cas où les corpus à traiter sont de nature homogène. Cela vaut particulièrement pour les algorithmes de clustering neuronaux, et encore plus spécifiquement pour les récentes versions adaptatives de ces algorithmes, comme l'algorithme incrémental de gaz neuronal croissant (IGNG). Cependant, ce papier illustre clairement la chute drastique de performance de la plupart de ces algorithmes dans le cas plus réaliste où les corpus à traiter s'avèrent être de nature hétérogène, ou polythématique. Dans ce contexte, des mesures spécifiques de qualité de clustering et de nouvelles techniques d'étiquetage des clusters qui sont indépendantes de la méthode de clustering utilisée sont exploitées pour l'évaluation des performances des méthodes. Un nouvel algorithme de gaz neuronal croissant exploitant une mesure de similarité basée sur la maximisation de la qualité de l'étiquetage des clusters est ensuite présenté comme une alternative à l'algorithme IGNG original basé sur la distance euclidienne. Nous montrons que cette solution permet d'obtenir un accroissement très significatif de performance pour le clustering des données textuelles polythématiques. Celle-ci fournit également par ailleurs un véritable caractère incrémental à l'algorithme proposé.

This paper presents an information retrieval model on XML documents based on tree matching. Queries and documents are represented by extended trees. An extended tree is built starting from the original tree, with additional weighted virtual links between each node and its indirect descendants allowing to directly reach each descendant. Therefore only one level separates between each node and its indirect descendants. This allows to compare the user query and the document with flexibility and with respect to the structural constraints of the query. The content of each node is very important to decide weither a document element is relevant or not, thus the content should be taken into account in the retrieval process. We separate between the structure-based and the content-based retrieval processes. The content-based score of each node is commonly based on the well-known Tf × Idf criteria. In this paper, we compare between this criteria and another one we call Tf × Ief. The comparison...

Terdapat milyaran dokumen web di world wide web yang terus bertumbuh dalam volume, kecepatan dan kompleksitas yang besar dan secara alamiah sebagian besar kontennya tidak terstruktur. Diperlukan adanya teknik atau alat untuk mengekstraksi data teks dari sebuah halaman web yang dapat beradaptasi terhadap konten yang tidak terstruktur maupun semi terstruktur dari halaman web. Pada penelitian ini penulis mengajukan pustaka Java Jsoup untuk mengekstraksi dokumen web kemudian memvisualisasikan hasilnya dalam bentuk word cloud.

Dans cet article, nous présentons les améliorations que nous avons apportées au système ExtraNews de résumé automatique de documents multiples. Ce système se base sur l'utilisation d'un algorithme génétique qui permet de combiner les phrases des documents sources pour former les extraits, qui seront croisés et mutés pour générer de nouveaux extraits. La multiplicité des critères de sélection d'extraits nous a inspiré une première amélioration qui consiste à utiliser une technique d'optimisation multi-objectif en vue d'évaluer ces extraits. La deuxième amélioration consiste à intégrer une étape de pré-filtrage de phrases qui a pour objectif la réduction du nombre des phrases des textes sources en entrée. Une évaluation des améliorations apportées à notre système est réalisée sur les corpus de DUC'04 et DUC'07.

This report describes the joint work by Kyoto University and the University of Melbourne for the TREC Genomics Track in 2007. As with 2006, the task for this year was the retrieval of passages from a biomedical document collection. The overall framework of our system from 2006 remains unchanged and is comprised of two parts: a paragraph-level retrieval system and a passage extraction system. These two systems are based on the vector space model and a probabilistic word-based aspect model, respectively. This year, we have adopted numerous changes to our 2007 system which we believe corrected some problems.

In contrast with their monolingual counterparts, little attention has been paid to the effects that misspelled queries have on the performance of Cross-Language Information Retrieval (CLIR) systems. The present work makes a first attempt to fill this gap by extending our previous work on monolingual retrieval in order to study the impact that the progressive addition of misspellings to input queries has, this time, on the output of CLIR systems. Two approaches for dealing with this problem are analyzed in this paper. Firstly, the use of automatic spelling correction techniques for which, in turn, we consider two algorithms: the first one for the correction of isolated words and the second one for a correction based on the linguistic context of the misspelled word. The second approach to be studied is the use of character n-grams both as index terms and translation units, seeking to take advantage of their inherent robustness and language-independence. All these approaches have been tested on a from-Spanish-to-English CLIR system, that is, Spanish queries on English documents. Real, user-generated spelling errors have been used under a methodology that allows us to study the effectiveness of the different approaches to be tested and their behavior when confronted with different error rates. The results obtained show the great sensitiveness of classic word-based approaches to misspelled queries, although spelling correction techniques can mitigate such negative effects. On the other hand, the use of character n-grams provides great robustness against misspellings.

This paper describes an extension of our work presented in the robust English-to-French bilingual task of the CLEF 2007 workshop, a knowledge-light approach for query translation in Cross-Language Information Retrieval systems. Our work is based on the direct translation of character n-grams, avoiding the need for word normalization during indexing or translation, and also dealing with out-of-vocabulary words. Moreover, since such a solution does not rely on language-specific processing, it can be used with languages of very different nature even when linguistic information and resources are scarce or unavailable. The results obtained have been very positive, and support the findings from our previous English-to-Spanish experiments.

Les réseaux sociaux contiennent une quantité importante d'information qui peut être divisée en deux dimensions : d'abord une dimension structurelle composée par l'information concernant les connexions entre les acteurs du réseau, puis une dimension de composition qui décrit chacun des acteurs de façon individuelle. Pour analyser un réseau, dans la plupart de cas, seulement une des dimensions est prise en compte. Cela implique une restriction de la portée de l'information qui peut être extraite du réseau. Nous proposons donc, une approche conçue pour intégrer les deux dimensions dans un processus de détection de communautés. Ensuite, nous proposons un modèle visuel pour analyser et exploiter l'information produite par le processus de détection de communautés. ABSTRACT. Social networks contain an important amount of information that can be divided into two dimensions: first a structural dimension, containing all the information about the connections and relationship types between the actors of the network; then a compositional dimension that describes each actor of the network individually. Due to the nature of each of these dimensions most of the analyzes are made using only one of them, discarding the other. This implies a restriction on the scope of information that can be extracted from the network. In this paper we propose, first an approach for integrating both dimensions in a community detection process, and second a visual model that allows for the analysis and use of the partition produced by the community detection process.

Implemented with a fuzzy logic language by using the FLOPER tool developed in our research group, we have recently designed a fuzzy dialect of the popular XPath language for the flexible manipulation of XML documents. In this paper we focus on the ability of Fuzzy XPath for exploring derivation trees generated by FLOPER once they are exported in XML format, which somehow serves as a debugging/analizing tool for discovering the set of fuzzy computed answers for a given goal, performing depth/breadth-first traversals of its associated derivation tree, finding non fully evaluated branches, etc., thus reinforcing the bi-lateral synergies between Fuzzy XPath and FLOPER.

Due to the emerging technology era, today a number of firms share their service/product descriptions. Such a group of information in the textual form has some structured information, which is beneath the unstructured text. A new attainment which facilitates the form of a structured metadata by recognizing documents which are likely to have some type and this information is then used for both segregation and search process. The idea of this advent describes some attributes of a text that will match with the query object which acts as identifier both for segregation as well as for storage and retrieval. An adaptive technique is proposed to deal with relevant attributes to annotate a document by satisfying the users querying needs. The solution for annotation-attribute suggestion problem is not based on the probabilistic model or prediction but it is based on the basic keywords that a user can use to query a database to retrieve a document. Experiment results show that Querying value and Content Value approach is much useful in predicting a tag for a document and thus prediction is also based on Querying value and Content value which greatly improves the utility of shared data which is a drawback in the existing system. This approach is different, as we consider only the basic keywords to be matched with the content of a document. When compared with other approaches in the existing system, Clarity is a primary goal as we expect that the annotator may improve the annotations on process. The discovered tags assist on quest of retrieval as an alternative to bookmarking.

Implicit feedback recommender systems provide personalized suggestions for items that are predicted to be of interest to the user, by collecting online users’ activity and inferring from it users’ preferences. While the state-of-the-art models are built by employing observations of user actions of one single type, the usage of multiple action types enables to observe more information and to build more effective recommenders. This paper presents an ongoing research in the field of implicit feedback recommender systems employing information about multiple action types. It highlights the peculiarities of existing models and identifies the research questions that should be solved to build solutions that can exploit multiple action types.

Mes dédicaces ne sont que l'expression de mes profondes gratitudes, de mes salutations chaleureuses et de ma sincère reconnaissance à tous ceux qui comblent ma vie et y confèrent son goût et sa saveur. Je dédie cette thèse à : Mon père Ce symbole de sacrifice et de dévouement, ses aides et ses recommandations m'ont souvent incité à persévérer dans l'effort et à progresser dans ma vie universitaire et professionnelle. Ma mère Ce rayon de soleil qui ne cesse d'éclairer ma vie, cette source inépuisable d'amour, de tendresse et d'affection, grâce à ses conseils, j'ai pu frayer mon chemin et savourer le goût de la réussite et du succès.

Being lightweight and cost-effective, IR-based approaches for bug localization have shown promise in finding software bugs. However, the accuracy of these approaches heavily depends on their used bug reports. A significant number of bug reports contain only plain natural language texts. According to existing studies, IR-based approaches cannot perform well when they use these bug reports as search queries. On the other hand, there is a piece of recent evidence that suggests that even these natural language-only reports contain enough good keywords that could help localize the bugs successfully. On one hand, these findings suggest that natural language-only bug reports might be a sufficient source for good query keywords. On the other hand, they cast serious doubt on the query selection practices in the IR-based bug localization. In this article, we attempted to clear the sky on this aspect by conducting an in-depth empirical study that critically examines the state-of-the-art query selection practices in IR-based bug localization. In particular, we use a dataset of 2,320 bug reports, employ ten existing approaches from the literature, exploit the Genetic Algorithm-based approach to construct optimal, near-optimal search queries from these bug reports, and then answer three research questions. We confirmed that the state-of-the-art query construction approaches are indeed not sufficient for constructing appropriate queries (for bug localization) from certain natural language-only bug reports. However, these bug reports indeed contain high-quality search keywords in their texts even though they might not contain explicit hints for localizing bugs (e.g., stack traces). We also demonstrate that optimal queries and non-optimal queries chosen from bug report texts are significantly different in terms of several keyword characteristics (e.g., frequency, entropy, position, part of speech). Such an analysis has led us to four actionable insights on how to choose appropriate keywords from a bug report. Furthermore, we demonstrate 27%-34% improvement in the performance of non-optimal queries through the application of our actionable insights to them. Finally, we summarize our study findings with future research directions (e.g., machine intelligence in keyword selection).

Malaria, Dengue Hemorrhagic Fever, and Bird Flu are three examples of various infectious diseases that spread in Indonesia in recent years. For example, Dengue Hemorrhagic Fever (DHF) is the main problem in tropical countries that emerges at the same time with rainy season. Especially in Indonesia, DHF is reported experiencing a quick increase in the number of case and death

In the past few decades recommender system has reshaped the way of information filtering between websites and the users. It helps in identifying user interest and generates product suggestions for the active users. This paper presents an enlightening analysis of various recommender system such as content-based, collaborative-based and hybrid recommendation techniques along with few optimization models that has been applied to improvise the parameters being considered by the aforementioned techniques. We explored 125 articles published from 1992 to 2019 in order to discuss the problems associated with the existing models. Various advantages and disadvantages of each recommendation model including the input methods has been elaborated. Critical review on research problems based on the explored techniques and future directions has also been covered.

XQuery is the standard language for querying XML documents using structural and content restrictions. In this paper, we propose to extend XQuery with selection operations that allow for the selection of the interesting subset of elements from each intermediate result of a query. To make this possible, intermediate results must be available during the construction of the query. This helps the user in building a query to retrieve the desired result. XQuery is being extended with a Full-Text language that allows to perform operations on text treating it as a sequence of words, units of punctuation, and spaces. This language includes a score clause that associates relevance measures to the results. The computation of these measures can be done using the method we also describe in this paper.

In the biomedical domain, there are immense data and tremendous increase of genomics and biomedical relevant publications. The wealth of information has led to an increasing amount of interest in and need for applying information retrieval techniques to access the scientific literature in genomics and related biomedical disciplines. In many cases, the desired information of a query asked by biologists is a list of a certain type of entities covering different aspects that are related to the question, such as cells, genes, diseases, proteins, mutations, etc. Hence, it is important of a biomedical IR system to be able to provide relevant and diverse answers to fulfill biologists' information needs. However traditional IR model only concerns with the relevance between retrieved documents and user query, but does not take redundancy between retrieved documents into account. This will lead to high redundancy and low diversity in the retrieval ranked lists. In this paper, we propose a...