Hila Becker

Beyond Trending Topics: Real-World Event Identification on Twitter

User-contributed messages on social media sites such as Twitter have emerged aspowerful, real-time means of information sharing on the Web. These short messages tend to reflect a variety of events in real time, making Twitter particularly well suited as a source of real-time event content. In this paper, we explore approaches for analyzing the stream of Twitter messages to distinguish between messages about real-world events andnon-event messages. Our approach relies on a rich family of aggregatestatistics of topically similar message clusters. Large-scale experiments over millions of Twitter messages show the effectiveness of our approach for surfacing real-world event content on Twitter.

Hip and trendy: Characterizing emerging trends on Twitter

Twitter, Facebook, and other related systems that we call social awareness streams are rapidly changing the information and communication dynamics of our society. These systems, where hundreds of millions of users share short messages in real time, expose the aggregate interests and attention of global and local communities. In particular, emerging temporal trends in these systems, especially those related to a single geographic area, are a significant and revealing source of information for, and about, a local community. This study makes two essential contributions for interpreting emerging temporal trends in these information systems. First, based on a large dataset of Twitter messages from one geographic area, we develop a taxonomy of the trends present in the data. Second, we identify important dimensions according to which trends can be categorized, as well as the key distinguishing features of trends that can be derived from their associated messages. We quantitatively examine the computed features for different categories of trends, and establish that significant differences can be detected across categories. Our study advances the understanding of trends on Twitter and other social awareness streams, which will enable powerful applications and activities, including user-driven real-time information services for local communities.

Identifying content for planned events across social media sites

User-contributed Web data contains rich and diverse information about a variety of events in the physical world, such as shows, festivals, conferences and more. This information ranges from known event features (e.g., title, time, location) posted on event aggregation platforms (e.g., Last.fm events, EventBrite, Facebook events) to discussions and reactions related to events shared on different social media sites (e.g., Twitter, YouTube, Flickr). In this paper, we focus on the challenge of automatically identifying user-contributed content for events that are planned and, therefore, known in advance, across different social media sites. We mine event aggregation platforms to extract event features, which are often noisy or missing. We use these features to develop query formulation strategies for retrieving content associated with an event on different social media sites. Further, we explore ways in which event content identified on one social media site can be used to retrieve additional relevant event content on other social media sites. We apply our strategies to a large set of user-contributed events, and analyze their effectiveness in retrieving relevant event content from Twitter, YouTube, and Flickr.

Context transfer in search advertising

We define and study the process of context transfer in search advertising, which is the transition of a user from the context of Web search to the context of the landing page that follows an ad-click. We conclude that in the vast majority of cases, the user is shown one of three types of pages, which can be accurately distinguished using automatic text classification.

Real-time ranking with concept drift using expert advice

In many practical applications, one is interested in generating a ranked list of items using information mined from continuous streams of data. For example, in the context of computer networks, one might want to generate lists of nodes ranked according to their susceptibility to attack. In addition, real-world data streams often exhibit concept drift, making the learning task even more challenging. We present an online learning approach to ranking with concept drift, using weighted majority techniques. By continuously modeling different snapshots of the data and tuning our measure of belief in these models over time, we capture changes in the underlying concept and adapt our predictions accordingly. We measure the performance of our algorithm on real electricity data as well as asynthetic data stream, and demonstrate that our approach to ranking from stream data outperforms previously known batch-learning methods and other online methods that do not account for concept drift.

Identification and characterization of events in social media

Millions of users share their experiences, thoughts, and interests online, through social media sites (e.g., Twitter, Flickr, YouTube). As a result, these sites host a substantial number of user-contributed documents (e.g., textual messages, photographs, videos) for a wide variety of events (e.g., concerts, political demonstrations, earthquakes). In this dissertation, we present techniques for leveraging the wealth of available social media documents to identify and characterize events of different types and scale. By automatically identifying and characterizing events and their associated user-contributed social media documents, we can ultimately offer substantial improvements in browsing and search quality for event content. To understand the types of events that exist in social media, we first characterize a large set of events using their associated social media documents. Specifically, we develop a taxonomy of events in social media, identify important dimensions along which they can be categorized, and determine the key distinguishing features that can be derived from their associated documents. We quantitatively examine the computed features for different categories of events, and establish that significant differences can be detected across categories. Importantly, we observe differences between events and other non-event content that exists in social media. We use these observations to inform our event identification techniques. To identify events in social media, we follow two possible scenarios. In one scenario, we do not have any information about the events that are re ected in the data. In this scenario, we use an online clustering framework to identify these unknown events and their associated social media documents. To distinguish between event and non-event content, we develop event classification techniques that rely on a rich family of aggregate cluster statistics, including temporal, social, topical, and platform-centric characteristics. In addition, to tailor the clustering framework to the social media domain, we develop similarity metric learning techniques for social media documents, exploiting the variety of document context features, both textual and non-textual. In our alternative event identification scenario, the events of interest are known, through user-contributed event aggregation platforms (e.g., Last.fm events, EventBrite, Facebook events). In this scenario, we can identify social media documents for the known events by exploiting known event features, such as the event title, venue, and time. While this event information is generally helpful and easy to collect, it is often noisy and ambiguous. To address this challenge, we develop query formulation strategies for retrieving event content on different social media sites. Specifically, we propose a two-step query formulation approach, with a first step that uses highly specific queries aimed at achieving high-precision results, and a second step that builds on these high-precision results, using term extraction and frequency analysis, with the goal of improving recall. Importantly, we demonstrate how event-related documents from one social media site can be used to enhance the identification of documents for the event on another social media site, thus contributing to the diversity of information that we identify. The number of social media documents that our techniques identify for each event is potentially large. To avoid overwhelming users with unmanageable volumes of event information, we design techniques for selecting a subset of documents from the total number of documents that we identify for each event. Specifically, we aim to select high-quality, relevant documents that re ect useful event information. For this content selection task, we experiment with several centrality-based techniques that consider the similarity of each event-related document to the central theme of its associated event and to other social media documents that correspond to the same event. We then evaluate both the relative and overall user satisfaction with the selected social media documents for each event. The existing tools to find and organize social media event content are extremely limited. This dissertation presents robust ways to organize and filter this noisy but powerful event information. With our event identification, characterization, and content selection techniques, we provide new opportunities for exploring and interacting with a diverse set of social media documents that re ect timely and revealing event content. Overall, the work presented in this dissertation provides an essential methodology for organizing social media documents that re ect event information, towards improved browsing and search for social media event data.

Verifying genre-based clustering approach to content extraction

The content of a webpage is usually contained within a small body of text and images, or perhaps several articles on the same page; however, the content may be lost in the clutter, particularly hurting users browsing on small cell phone and PDA screens and visually impaired users relying on speed rendering of web pages. Using the genre of a web page, we have created a solution, Crunch that automatically identifies clutter and removes it, thus leaving a clean content-full page. In order to evaluate the improvement in the applications for this technology, we identified a number of experiments. In this paper, we have those experiments, the associated results and their evaluation.

Genre Classification of Websites Using Search Engine Snippets

Web pages often contain clutter (such as ads, unnecessary images and extraneous links) around the body of an article, which distracts a user from actual content. Automatic extraction of “useful and relevant” content from web pages has many applications, including browsing on small cell phone and PDA screens, speech rendering for the visually impaired, and reducing noise for information retrieval systems. Prior work has led to the development of Crunch, a framework which employs various heuristics in the form of filters and filter settings for content extraction. Crunch allows users to tune these settings, essentially the thresholds for applying each filter. However, in order to reduce human involvement in selecting these heuristic settings, we have extended this work to utilize a website’s classification, defined by its genre and physical layout. In particular, Crunch would then obtain the settings for a previously unknown website by automatically classifying it as sufficiently similar to a cluster of known websites with previously adjusted settings - which in practice produces better content extraction results than a single one-size-fits-all set of setting defaults. In this paper, we present our approach to clustering a large corpus of websites by their genre, utilizing the snippets generated by sending the website’s domain name to search engines as well as the website’s own text. We find that exploiting these snippets not only increased the frequency of function words that directly assist in detecting the genre of a website, but also allow for easier clustering of websites. We use existing techniques like Manhattan distance measure and Hierarchical clustering, with some modifications, to pre-classify websites into genres. Our clustering method does not require prior knowledge of the set of genres that websites fit into, but instead discovers these relationships among websites. Subsequently, we are able to classify newly encountered websites in linear-time, and then apply the corresponding filter settings, with no noticeable delay introduced for the content-extracting web proxy.

A Genre-based Clustering Approach to Content Extraction

The content of a webpage is usually contained within a small body of text and images, or perhaps several articles on the same page; however, the content may be lost in the clutter (defined as cosmetic features such as animations, menus, sidebars, obtrusive banners). Automatic content extraction has many applications, including browsing on small cell phone and PDA screens, speech rendering for the visually impaired, and reducing noise for information retrieval systems. We have developed a framework, Crunch, which employs various heuristics for content extraction in the form of filters applied to the webpage’s DOM tree; the filters aim to prune or transform the clutter, leaving only the content. Crunch allows users to tune what we call “settings”, consisting of thresholds for applying a particular filter and/or for toggling a filter on/off, because the HTML components that characterize clutter can vary significantly from website to website. However, we have found that the same settings tend to work well across different websites of the same genre, e.g., news or shopping, since the designers often employ similar page layouts. In particular, Crunch could obtain the settings for a previously unknown website by automatically classifying it as sufficiently similar to a cluster of known websites with previously adjusted settings. We present our approach to clustering a large corpus of websites into genres, using their pre-extraction textual material augmented by the snippets generated by searching for the website’s domain name in web search engines. Including these snippets increases the frequency of function words needed for clustering. We use existing Manhattan distance measure and hierarchical clustering techniques, with some modifications, to pre-classify the corpus into genres offline. Our method does not require prior knowledge of the set of genres that websites fit into, but to be useful a priori settings must be available for some member of each cluster or a nearby cluster (otherwise defaults are used). Crunch classifies newly encountered websites online in linear-time, and then applies the corresponding filter settings, with no noticeable delay added by our content-extracting web proxy.