Erel Uziel

Social media recommendation based on people and tags

We study personalized item recommendation within an enterprise social media application suite that includes blogs, bookmarks, communities, wikis, and shared files. Recommendations are based on two of the core elements of social media - people and tags. Relationship information among people, tags, and items, is collected and aggregated across different sources within the enterprise. Based on these aggregated relationships, the system recommends items related to people and tags that are related to the user. Each recommended item is accompanied by an explanation that includes the people and tags that led to its recommendation, as well as their relationships with the user and the item. We evaluated our recommender system through an extensive user study. Results show a significantly better interest ratio for the tag-based recommender than for the people-based recommender, and an even better performance for a combined recommender. Tags applied on the user by other people are found to be highly effective in representing that users topics of interest.

Personalized recommendation of social software items based on social relations

We study personalized recommendation of social software items, including bookmarked web-pages, blog entries, and communities. We focus on recommendations that are derived from the users social network. Social network information is collected and aggregated across different data sources within our organization. At the core of our research is a comparison between recommendations that are based on the users familiarity network and his/her similarity network. We also examine the effect of adding explanations to each recommended item that show related people and their relationship to the user and to the item. Evaluation, based on an extensive user survey with 290 participants and a field study including 90 users, indicates superiority of the familiarity network as a basis for recommendations. In addition, an important instant effect of explanations is found - interest rate in recommended items increases when explanations are provided.

Same places, same things, same people?: mining user similarity on social media

In this work we examine nine different sources for user similarity as reflected by activity in social media applications. We suggest a classification of these sources into three categories: people, things, and places. Lists of similar people returned by the nine sources are found to be highly different from each other as well as from the list of people the user is familiar with, suggesting that aggregation of sources may be valuable. Evaluation of the sources and their aggregates points at their usefulness across different scenarios, such as information discovery and expertise location, and also highlights sources and aggregates that are particularly valuable for inferring user similarity.

Social networks and discovery in the enterprise (SaND)

Traditional information discovery methods are based on content: documents, terms, and the relationships between them. In Web 2.0, people come into play as they create documents and tags in many forms. Personalized search, social graphs, content and people recommendation, are some of the tasks that can take advantage of this newly formed ecosystem. The Social Networks and Discovery (SaND) platform is an aggregation tool for information discovery and analysis over social data gathered from Web 2.0 applications in the enterprise. It leverages complex relationships between content and people as surfaced through the social applications to unleash the value of information. Its integrated index supports combining content-based analysis and people-based analysis over a rich data foundation. Enterprise social data is easily modeled and ingested into SaND, and can be further combined with data from external social applications. This demo will present three main functions provided by SaND: Social search: SaND supports search over the social data using a unified approach [1] in which all system entities (documents, people, tags) are searchable and retrievable (See Figure 1). The search UI enables the searcher to get a wider view on the query topic through results from all entity types, while uncovering the relationships between the on-screen entities. Entity recommendation: SaND can be utilized to recommend people and content for the searcher (Figure 2 shows the “Do You Know” widget for people recommendation). People are recommended according to their “social similarity” relations with the searcher, e.g. organizational and friending ties, similar tagging activity and more. Similarly, content that is related to people that are “socially related” to the searcher is recommended as valuable interesting data. Personalization: Search results are personalized by considering the relations of retrieved entities with the searcher. Entities are ranked according to their relevance to the query as well as according to their relationship strength with the searcher.

Digital Traces of Interest: Deriving Interest Relationships from Social Media Interactions

Facebook and Twitter have changed the way we consume information, allowing the people we follow to become our “social filters” and determine the content of our information stream. The capability to discover the individuals a user is most interested in following has therefore become an important aspect of the struggle against information overflow. We argue that the people users are most interested in following are not necessarily those with whom they are most familiar. We compare these two types of social relationships – interest and familiarity – inside IBM. We suggest inferring interest relationships from users’ public interactions on four enterprise social media applications. We study these interest relationships through an offline analysis as well as an extensive user study, in which we combine people-based and content-based evaluations. The paper reports a rich set of results, comparing various sources for implicit interest indications; distinguishing between content-related activities and status or network updates, showing that the former are of more interest; and highlighting that the interest relationships include very interesting individuals that are not among the most familiar ones, and can therefore play an important role in social stream filtering, especially for content-related activities.

Visual social network analytics for relationship discovery in the enterprise

As people continue to author and share increasing amounts of information in social media, the opportunity to leverage such information for relationship discovery tasks increases. In this paper, we describe a set of systems that mine, aggregate, and infer a social graph from social media inside an enterprise, resulting in over 73 million relationships between 450,000 people. We then describe SaNDVis, a novel visual analytics tool that supports people-centric tasks like expertise location, team building, and team coordination in the enterprise. We also provide details of a 12-month-long, large-scale deployment to almost 1,800 users from which we extract dominant use cases from log and interview data. By integrating social position, evidence, and facets into SaNDVis, we demonstrate how users can use a visual analytics tool to reflect on existing relationships as well as build new relationships in an enterprise setting.

Folksonomy-Based Term Extraction for Word Cloud Generation

In this work we study the task of term extraction for word cloud generation in sparsely tagged domains, in which manual tags are scarce. We present a folksonomy-based term extraction method, called tag-boost, which boosts terms that are frequently used by the public to tag content. Our experiments with tag-boost based term extraction over different domains demonstrate tremendous improvement in word cloud quality, as reflected by the agreement between manual tags of the testing items and the cloud’s terms extracted from the items’ content. Moreover, our results demonstrate the high robustness of this approach, as compared to alternative cloud generation methods that exhibit a high sensitivity to data sparseness. Additionally, we show that tag-boost can be effectively applied even in nontagged domains, by using an external rich folksonomy borrowed from a well-tagged domain.

Folksonomy-based term extraction for word cloud generation

In this work we study the task of term extraction for word cloud generation. We present a folksonomy-based term extraction method, called tag-boost, which boosts terms that are frequently used by the public to tag content. Our experiments with tag-boost-based term extraction over different domains demonstrate tremendous improvement in word cloud quality, as reflected by the agreement between extracted terms and manually assigned tags of the testing items. Additionally, we show that tag-boost can be effectively applied even in non-tagged domains, by using an external rich folksonomy borrowed from a well-tagged domain.

The Longitudinal Use of SaNDVis: Visual Social Network Analytics in the Enterprise

As people continue to author and share increasing amounts of information in social media, the opportunity to leverage such information for relationship discovery tasks increases. In this paper, we describe a set of systems that mine, aggregate, and infer a social graph from social media inside an enterprise, resulting in over 73 million relationships between 450,000 people. We then describe SaNDVis, a novel visual analytics tool that supports people-centric tasks like expertise location, team building, and team coordination in the enterprise. We provide details of a 22-month-long, large-scale deployment to over 2,300 users from which we analyze longitudinal usage patterns, classify types of visual analytics queries and users, and extract dominant use cases from log and interview data. By integrating social position, evidence, and facets into SaNDVis, we demonstrate how users can use a visual analytics tool to reflect on existing relationships as well as build new relationships in an enterprise setting.