The fast spreading of fake news stories on social media can cause inestimable social harm. Developing effective methods to detect them early is of paramount importance. A major challenge of fake news early detection is fully utilizing the limited data observed at the early stage of news propagation and then learning useful patterns from it for identifying fake news. In this article, we propose a novel deep neural network to detect fake news early. It has three novel components: (1) a status-sensitive crowd response feature extractor that extracts both text features and user features from combinations of users’ text response and their corresponding user profiles, (2) a position-aware attention mechanism that highlights important user responses at specific ranking positions, and (3) a multi-region mean-pooling mechanism to perform feature aggregation based on multiple window sizes. Experimental results on two real-world datasets demonstrate that our proposed model can detect fake news with greater than 90% accuracy within 5 minutes after it starts to spread and before it is retweeted 50 times, which is significantly faster than state-of-the-art baselines. Most importantly, our approach requires only 10% labeled fake news samples to achieve this effectiveness under PU-Learning settings.

Formative feedback has long been recognized as a crucial scaffold for student learning. Due to the job demand of instructors, it is impossible for them to provide individual students with on-demand formative feedback based on individual students' performance. There is a growing interest in developing better approaches to provide students with automated formative feedback to assist their learning. In this research, we design and develop an automated formative feedback system to support student learning of conceptual knowledge in the course of writing assignments. In the proposed system, formative feedback can be generated automatically with the help of concept maps constructed from instructors' lecture slides and students' writing assignments. In this paper, we present the automatic approach to generate formative feedback, discuss the system architecture, and illustrate a prototype of the proposed system.

In the midst of today's pervasive influence of social media, automatically detecting fake news is drawing significant attention from both the academic communities and the general public. Existing detection approaches rely on machine learning algorithms with a variety of news characteristics to detect fake news. However, such approaches have a major limitation on detecting fake news early, i.e., the information required for detecting fake news is often unavailable or inadequate at the early stage of news propagation. As a result, the accuracy of early detection of fake news is low. To address this limitation, in this paper, we propose a novel model for early detection of fake news on social media through classifying news propagation paths. We first model the propagation path of each news story as a multivariate time series in which each tuple is a numerical vector representing characteristics of a user who engaged in spreading the news. Then, we build a time series classifier that incorporates both recurrent and convolutional networks which capture the global and local variations of user characteristics along the propagation path respectively, to detect fake news. Experimental results on three real-world datasets demonstrate that our proposed model can detect fake news with accuracy 85% and 92% on Twitter and Sina Weibo respectively in 5 minutes after it starts to spread, which is significantly faster than state-of-the-art baselines.

A primary goal of higher education is to promote meaningful learning: the delivery of core academic content to students in innovative ways that allow them to learn and then apply what they have learned. As a pedagogical strategy, Writing-to-Learn (WTL) intends to use writing to improve students' understanding of course content and concepts. To improve students' meaningful learning of conceptual knowledge in WTL activities, the project proposes to develop the Write-and-Learn system to generate automated formative feedback by taking advantage of the concept maps constructed from instructors' lecture notes and individual students' writing assignments. The proposed research aims to provide insights into how to apply concept maps into WTL activities to generate effective formative feedback on the acquisition and development of conceptual knowledge, and explore how and to what extent concept map-based formative feedback can be utilized to scaffold and promote meaningful learning in WTL activities.

Search By Multiple Examples (SBME) is a new search paradigm that allows users to specify their information needs as a set of relevant documents rather than as a set of keywords. In this study, we propose a Transductive Positive Unlabeled learning (TPU learning) based framework for SBME. The framework consists of two steps: 1) identifying potential relevant documents for searching space reduction, and 2) adopting TPU learning methods to re-rank the documents in the new searching space. Using MAP and p@k, we evaluate two state-of-the-art PU learning algorithms and the Rocchio classifier (Rc) for document ranking in the proposed framework. We then adopt the idea of ensemble learning to combine Rc with the two state-of-the-art PU learning algorithms respectively. Experiments conducted on a benchmark dataset show that the ensemble learning based methods lead to a significant improvement in effectiveness.

It is often difficult for users to adopt keywords to express their information needs. Search-By-Multiple-Examples (SBME), a promising method for overcoming this problem, allows users to specify their information needs as a set of relevant documents rather than as a set of keywords. Most of the studies on SBME adopt the Positive Unlabeled learning (PU learning) techniques by treating the users' provided examples (denote as query examples) as positive set and the entire data collection as unlabeled set. However, it is inefficient to treat the entire data collection as unlabeled set, as its size can be huge. In addition, the query examples are treated as being relevant to a single topic, but it is often the case that they can be relevant to multiple topics. As the query examples are much fewer than the unlabeled data, the system performance may downgrade dramatically because of the class imbalance problem. What's more, the experiments conducted in these studies have not taken into account the settings in online search, which are very different from the controlled experiments scenario. This proposed research seeks to explore how to improve SBME by exploring: (1) how to predict user' information needs by modeling the content of the documents using probabilistic topic models; (2) how to deal with the class imbalance problem by reducing the size of the unlabeled data and adopting machine learning techniques. We will also conduct extensive experiments to better evaluate SBME using different sizes of query examples to simulate users' information needs.

Finding the most interesting correlations among items is essential for problems in many commercial, medical, and scientific domains. Although there are numerous measures available for evaluating correlations, different correlation measures provide drastically different results. Piatetsky-Shapiro provided three mandatory properties for any reasonable correlation measure, and Tan et al. proposed several properties to categorize correlation measures; however, it is still hard for users to choose the desirable correlation measures according to their needs. In order to solve this problem, we explore the effectiveness problem in three ways. First, we propose two desirable properties and two optional properties for correlation measure selection and study the property satisfaction for different correlation measures. Second, we study different techniques to adjust correlation measures and propose two new correlation measures: the Simplified χ2with Continuity Correction and the Simplified χ2with Support. Third, we analyze the upper and lower bounds of different measures and categorize them by the bound differences. Combining these three directions, we provide guidelines for users to choose the proper measure according to their needs.

Recommender systems aim to substantially reduce information overload by suggesting lists of similar items that users may find interesting.Caching has been a useful technique for reducing stress on limited resources and improving response time. In this paper, we propose an 'active caching' technique for recommender systems based on a partial order approach that not only benefits from popularity and temporal locality, but also exploits spatial locality. This approach allows the processing of answers to neighboring non-cached queries in addition to the reporting of cached query results. Test results for several data sets and recommendation techniques show substantial improvement in the cache hit ratio and computational costs, while achieving reasonable recall rates.

Abstract On the Web, there are limited ways of finding people sharing similar interests with a given person. The current methods are either ineffective or time consuming. In this paper, we present a new approach for searching people sharing similar interests from the Web. Given a person, to find similar people from the Web, there are two major research issues: person representation and matching persons. In this study, we propose a person representation method which uses a person's website to represent this person. Our design of matching process takes person representation into consideration to allow the same representation to be used when composing the query. Under this person representation method, the proposed algorithm integrates textual content and hyperlink information of all the pages belonging to a personal website to represent a person and match persons. Other algorithms are also explored and compared to the proposed algorithm. Experimental results are presented.

This paper presents a methodology for personalized knowledge discovery from text. It derives a user's background knowledge from his/her background documents, and exploits such knowledge to evaluate the novelty of discovered knowledge in the form of association rules by measuring the semantic distance between the antecedent and the consequent of a rule in the background knowledge. The experiment results show that the proposed user-oriented novelty measure is highly correlated with the human subjective rule novelty and usefulness ratings. It outperforms seven major objective interestingness measures and the WordNet novelty measure for identifying novel and useful rules.

Abstract Document keyphrases provide a concise summary of a document's content, offering semantic metadata summarizing a document. They can be used in many applications related to knowledge management and text mining, such as automatic text summarization, development of search engines, document clustering, document classification, thesaurus construction, and browsing interfaces. Because only a small portion of documents have keyphrases assigned by authors, and it is time‐consuming and costly to manually assign keyphrases to documents, it is necessary to develop an algorithm to automatically generate keyphrases for documents. This paper describes a Keyphrase Identification Program (KIP), which extracts document keyphrases by using prior positive samples of human identified phrases to assign weights to the candidate keyphrases. The logic of our algorithm is: The more keywords a candidate keyphrase contains and the more significant these keywords are, the more likely this candidate phrase is a keyphrase. KIP's learning function can enrich the glossary database by automatically adding new identified keyphrases to the database. KIP's personalization feature will let the user build a glossary database specifically suitable for the area of his/her interest. The evaluation results show that KIP's performance is better than the systems we compared to and that the learning function is effective.

This paper presents a hybrid concept hierarchy development technique for web returned documents retrieved by a meta-search engine. The aim of the technique is to separate the initial retrieved documents into topical oriented categories, prior to the actual concept hierarchy generation. The topical categories correspond to different semantic aspects of the query. This is done using a 1-of-n automatic document classification, on the initial set of returned documents. Then, an individual topical concept hierarchy is automatically generated inside each of the resulted categories. Both steps are executed on the fly at retrieval time. Due to the efficiency constraints imposed by the web retrieval context, the algorithm only uses document snippets (rather than full web pages) for both document classification and concept hierarchy generation. Experimental results show that the algorithm is able to improve the quality of the concept hierarchy presented to the searcher; at the same time, the efficiency parameters are kept within reasonable intervals.

Document keyphrases provide semantic metadata characterizing documents and producing an overview of the content of a document. They can be used in many text-mining and knowledge management related applications. This paper describes a Keyphrase Identification Program (KIP), which extracts document keyphrases by using prior positive samples of human identified domain keyphrases to assign weights to the candidate keyphrases. The logic of our algorithm is: the more keywords a candidate keyphrase contains and the more significant these keywords are, the more likely this candidate phrase is a keyphrase. To obtain prior positive inputs, KIP first populates its glossary database using manually identified keyphrases and keywords. It then checks the composition of all noun phrases of a document, looks up the database and calculates scores for all these noun phrases. The ones having higher scores will be extracted as keyphrases.

Most of the elearning classes use discussion forums to support class interaction. This paper presents a novel approach that automatically assesses student class performance by analyzing knowledge density in online class messages. Class performance of students is evaluated from three perspectives: the quality of their course work, the quantity of their efforts, and the activeness of their participation; three measures - knowledge density, message length, and message count, are derived from class messages respectively. Together, they are used to calculate a performance indicator score. The experiment shows that there is a high correlation between the performance indicator scores and the actual grades assigned by instructors. The rank orders of students by the performance indicator score are highly correlated with those by the actual grades as well.

This paper presents a framework for user-oriented text mining. It is then illustrated with an example of discovering knowledge from competitors' websites. The knowledge to be discovered is in the form of association rules. A user's background knowledge is represented as a concept hierarchy developed from documents on his/her own website. The concept hierarchy captures the semantic usage of words and relationships among words in background documents. Association rules are identified among the noun phrases extracted from documents on competitors' websites. The interestingness measure, i.e. novelty, which measures the semantic distance between the antecedent and the consequent of a rule in the background knowledge, is computed from the co-occurrence frequency of words and the connection lengths among words in the concept hierarchy. A user evaluation of the novelty of discovered rules demonstrates that the correlation between the algorithm and the human judges is comparable to that between human judges.

In this paper we present a document representation improvement technique, named the Relevance Feedback Accumulation (RFA) algorithm. Using prior relevance feedback assessments and a data mining measure called "support", the algorithm's learning function gradually improves document representations, over time and across users. Results show that the modified document representations yield lower dimensionality while improving retrieval effectiveness. The algorithm is efficient and scalable, suited for retrieval systems managing large document collections.

In this paper, we propose a prototype system for automatic generation of concept hierarchies to be used as an overview of search results. The system sends a user's query to five search engines and receives a returned list of relevant web pages. The system then extracts query-oriented concept terms from snippets that come with the returned hits. Concept terms are organized into a concept hierarchy using a co-occurrence-based classification technique. Finally, concepts in returned documents are dynamically highlighted according to terms in the selected concept branch that lead to the chosen document. The user study shows that concept hierarchies do provide easy navigation and browsing of web returned documents. The results also show that users can find a document of interest no matter how low it is ranked in the retrieved list.