Geo-tagged Tweets Research Articles

Tourist Attraction Popularity Mapping based on Geotagged Tweets

The development of tourist attractions is now highly influenced by social media. The speed at which information can be disseminated via the Internet has become an essential factor in enabling distinct tourist attractions to potentially gain high popularity in a relatively short time. This condition was not as prevalent several years ago when tourism promotion remained limited to a certain kind of media. As a consequence, rapid change in the relative popularity of tourist attractions is inevitable. Against this, knowledge of tourist attraction hotspots is essential in tourism management. This means there is a need to study how to both quickly determine the popularity level of tourist attractions and encompass a relatively large area. This article utilised tweet data from microblogging website Twitter as the basis from which to determine the popularity level of a tourist attraction. Data mining was conducted using Python and the Tweepy module. The tweet data were collected at the end of April and early May 2017, at times when there are several long holiday weekends. A Tweet Proximity Index (TPI) was used to calculate both the density and frequency of tweets based on a defined search radius. A Density Index (DI) was also used as a technique for determining the popularity. The results from both approaches were then compared to a random survey about people’s perceptions of tourist attractions in the study area. The result shows that geotagged tweet data can be used to determine the popularity of a tourist attraction, although it still only achieved a medium level of accuracy. The TPI approach used in this study produced an accuracy of 76.47%, while the DI achieved only 58.82%. This medium accuracy does indicate that the two approaches are not yet strong enough to be used for decision-making but should be more than adequate as an initial description. Further, it is necessary to improve the method of indexing and the exploration of other aspects of Twitter data.

The Social Integration of American Cities: Network Measures of Connectedness Based on Everyday Mobility Across Neighborhoods

The social integration of a city depends on the extent to which people from different neighborhoods have the opportunity to interact with one another, but most prior work has not developed formal ways of conceptualizing and measuring this kind of connectedness. In this article, we develop original, network-based measures of what we call “structural connectedness” based on the everyday travel of people across neighborhoods. Our principal index captures the extent to which residents in each neighborhood of a city travel to all other neighborhoods in equal proportion. Our secondary index captures the extent to which travels within a city are concentrated in a handful of receiving neighborhoods. We illustrate the value of our indices for the 50 largest American cities based on hundreds of millions of geotagged tweets over 18 months. We uncover important features of major American cities, including the extent to which their connectedness depends on a few neighborhood hubs, and the fact that in several cities, contact between some neighborhoods is all but nonexistent. We also show that cities with greater population densities, more cosmopolitanism, and less racial segregation have higher levels of structural connectedness. Our indices can be applied to data at any spatial scale, and our measures pave the way for more powerful and precise analyses of structural connectedness and its effects across a broad array of social phenomena.

Semantic trajectory inference from geo-tagged tweets

Abstract. Individual travel trajectories denote a series of places people visit along the time. These places (e.g., home, workspace, and park) reflect people’s corresponding activities (e.g., dwelling, work, and entertainment), which are discussed as semantic knowledge and could be implicit under raw data (Yan et al. 2013, Cai et al. 2016). Traditional survey data directly describe people’ activities at certain places, while costing tremendous labors and resources (Huang and Wong 2016). GPS data such as taxi logs record exact origin-destination pairs as well as people’s stay time along the way, from which semantics can be easily inferred combining with geographical context data (Yan et al. 2013). Research has been done to understand the activity sequences indicated by either individual or collective spatiotemporal (ST) travel trajectories using those dense data. Different models are proposed for trajectory mining and activity inference, including location categorization, frequent region detection, and so on (Njoo et al. 2015). A typical method for matching a location or region with a known activity type is to detect stay points and stay intervals of trajectories and to find geographical context of these stay occurrences (Furtado et al. 2013, Njoo et al. 2015, Beber et al. 2016, Beber et al. 2017).However, limited progress has been made to mine semantics of trajectory data collected from social media platforms. Specifically, detection of stay points and their intervals could be inaccurate using online trajectories because of data sparsity. Huang et al. (2014) define the notion of activity zone to detect activity types from digital footprints. In this method, individual travel trajectories first are aggregated using spatial clustering method such as density-based spatial clustering of applications with noise (DBSCAN). Then produced clusters are classified based on a regional land use map and Google Places application programming interface (API). Such land use data are only published at specific places, such as the state cartography office’s website at University of Wisconsin-Madison. Researchers need to search for those data based on their study area. Moreover, while major land use maps can be searched for large areas such as the whole United States, detailed land use data for statewide or citywide areas are made in diverse standards, which adds extra work to classify activity zones consistently. Besides, Google Places API is a service that Google opened for developers and will return information about a place, given the place location (e.g., address or GPS coordinates), in the search request. However, API keys need to be generated before we can use these interfaces and each user can only make a limited number of free-charged requests every day (i.e., 1,000 requests per 24 hours period). In sum, previous methods to detect activity zone types using social media data are not sufficient and can hardly achieve effective data fusion. Comparing to the high cost of using officially published dataset, emerging Volunteered Geographic Information (VGI) data offer an alternative to infer the types of an individual’s activities performed in each zone (i.e., cluster).Using geo-tagged tweets as an example, this research proposes a framework for mining social media data, detecting individual semantic travel trajectories, and individual representative daily travel trajectory paths by fusing with VGI data, specifically OpenStreetMap (OSM) datasets. First, inactive users and abnormal users (e.g., users representing a company with account being shared by many employees) are removed through data pre-processing (Step 1 in Figure 1). Next, a multi-scale spatial clustering method is developed to aggregate online trajectories captured through geo-tagged tweets of a group of users into collective spatial hot-spots (i.e., activity zones; Step 2). By integrating multiple OSM datasets the activity type (e.g., dwelling, service, transportation and work) of each collective zone then can be identified (Step 3). Each geo-tagged tweet of an individual, represented as a ST point, is then attached with a collective activity zone that either includes or overlaps a buffer zone of the ST point. Herein, the buffer zone is generated by using the point as the centroid and a predefined threshold as the radius. Given an individual’s ST points with semantics (i.e., activity type information) derived from the attached collective activity zone, a semantic activity clustering method is then developed to detect daily representative activity clusters of the individual (Step 4). Finally, individual representative daily semantic travel trajectory paths (i.e., semantic travel trajectory, defined as chronological travel activity sequences) are constructed between every two subsequent activity clusters (Step 5). Experiments with the historic geo-tagged tweets collected within Madison, Wisconsin reveal that: 1) The proposed method can detect most significant activity zones with accurate zone types identified (Figure 2); and 2) The semantic activity clustering method based on the derived activity zones can aggregate individual travel trajectories into activity clusters more efficiently comparing to DBSCAN and varying DBSCAN (VDBSCAN).

Searching Activity Trajectories with Semantics

With the widespread use of smart phones and mobile Internet, social network users have generated massive geo-tagged tweets, photos and videos to form lots of informative trajectories which reveal not only their spatio-temporal dynamics, but also their activities in the physical world. Existing spatial trajectory query studies mainly focus on analyzing the spatio-temporal properties of the users’ trajectories, while leaving the understanding of their activities largely untouched. In this paper, we incorporate the semantics of the activity information embedded in trajectories into query modelling and processing, with the aim of providing end users more informative and meaningful results. To this end, we propose a novel trajectory query that not only considers the spatio-temporal closeness but also, more importantly, leverages a proven technique in text mining field, probabilistic topic modelling, to capture the semantic relatedness of the activities between the data and query. To support efficient query processing, we design a hierarchical grid-based index by integrating the probabilistic topic distribution on the substructures of trajectories and their spatio-temporal extent at the corresponding level of the index hierarchy. This specialized structure enables a top-down search algorithm to traverse the index while pruning unqualified trajectories in spatial and topical dimensions simultaneously. The experimental results on real-world datasets demonstrate the good efficiency and scalability performance of the proposed indices and trajectory search methods.

Identifying Travel Regions Using Location-Based Social Network Check-in Data.

Travel regions are not necessarily defined by political or administrative boundaries. For example, in the Schengen region of Europe, tourists can travel freely across borders irrespective of national borders. Identifying transboundary travel regions is an interesting problem which we aim to solve using mobility analysis of Twitter users. Our proposed solution comprises collecting geotagged tweets, combining them into trajectories and, thus, mining thousands of trips undertaken by twitter users. After aggregating these trips into a mobility graph, we apply a community detection algorithm to find coherent regions throughout the world. The discovered regions provide insights into international travel and can reveal both domestic and transnational travel regions.

Topic modeling and sentiment analysis of global climate change tweets

Social media websites can be used as a data source for mining public opinion on a variety of subjects including climate change. Twitter, in particular, allows for the evaluation of public opinion across both time and space because geotagged tweets include timestamps and geographic coordinates (latitude/longitude). In this study, a large dataset of geotagged tweets containing certain keywords relating to climate change is analyzed using volume analysis and text mining techniques such as topic modeling and sentiment analysis. Latent Dirichlet allocation was applied for topic modeling to infer the different topics of discussion, and Valence Aware Dictionary and sEntiment Reasoner was applied for sentiment analysis to determine the overall feelings and attitudes found in the dataset. These techniques are used to compare and contrast the nature of climate change discussion between different countries and over time. Sentiment analysis shows that the overall discussion is negative, especially when users are reacting to political or extreme weather events. Topic modeling shows that the different topics of discussion on climate change are diverse, but some topics are more prevalent than others. In particular, the discussion of climate change in the USA is less focused on policy-related topics than other countries.

Real-time event embedding for POI recommendation

Location-based social networks (LBSNs) allow users to check-in and share daily lives with others. We have witnessed very rapid development of LBSNs in recent years. Point-of-Interest (POI) recommendation is one of the core services in LBSNs. In this study, we propose a real-time POI embedding model. Instead of capturing intrinsic information, the proposed approach is able to mine real-time information of places and learn the latent representations according to the corresponding geo-tagged posts. On one hand, we employ a Convolutional Neural Networks (CNN) to mine textual information of POIs and learn their intrinsic representation. On the other hand, a multimodal embedding model of location, time and text is applied to keep monitoring posts on POIs and extracts a set of features for representing events or burst information that may attract users. Furthermore, we combine real-time POI embedding with matrix factorization method and propose a more comprehensive POI recommendation algorithm. To verify the effectiveness of our proposed method, we conduct experiments on Twitter dataset with geo-tagged tweets in NYC. Experimental results show that POI recommendation system with taking real-time event into consideration can strongly improve the performance than the one without.

A combination of incidence data and mobility proxies from social media predicts the intra-urban spread of dengue in Yogyakarta, Indonesia.

Only a few studies have investigated the potential of using geotagged social media data for predicting the patterns of spatio-temporal spread of vector-borne diseases. We herein demonstrated the role of human mobility in the intra-urban spread of dengue by weighting local incidence data with geo-tagged Twitter data as a proxy for human mobility across 45 neighborhoods in Yogyakarta city, Indonesia. To estimate the dengue virus importation pressure in each study neighborhood monthly, we developed an algorithm to estimate a dynamic mobility-weighted incidence index (MI), which quantifies the level of exposure to virus importation in any given neighborhood. Using a Bayesian spatio-temporal regression model, we estimated the coefficients and predictiveness of the MI index for lags up to 6 months. Specifically, we used a Poisson regression model with an unstructured spatial covariance matrix. We compared the predictability of the MI index to that of the dengue incidence rate over the preceding months in the same neighborhood (autocorrelation) and that of the mobility information alone. We based our estimates on a volume of 1·302·405 geotagged tweets (from 118·114 unique users) and monthly dengue incidence data for the 45 study neighborhoods in Yogyakarta city over the period from August 2016 to June 2018. The MI index, as a standalone variable, had the highest explanatory power for predicting dengue transmission risk in the study neighborhoods, with the greatest predictive ability at a 3-months lead time. The MI index was a better predictor of the dengue risk in a neighborhood than the recent transmission patterns in the same neighborhood, or just the mobility patterns between neighborhoods. Our results suggest that human mobility is an important driver of the spread of dengue within cities when combined with information on local circulation of the dengue virus. The geotagged Twitter data can provide important information on human mobility patterns to improve our understanding of the direction and the risk of spread of diseases, such as dengue. The proposed MI index together with traditional data sources can provide useful information for the development of more accurate and efficient early warning and response systems.

Improving Nighttime Light Imagery With Location-Based Social Media Data

Location-based social media have been extensively utilized in the concept of “social sensing” to exploit dynamic information about human activities, yet joint uses of social sensing and remote sensing images are underdeveloped at present. In this paper, the close relationship between the number of Twitter users and brightness of nighttime lights (NTL) over the contiguous United States is calculated and geotagged tweets are then used to upsample a stable light image for 2013. An associated outcome of the upsampling process is the solution of two major problems existing in the NTL image, pixel saturation, and blooming effects. Compared with the original stable light image, digital number (DN) values of the upsampled stable light image have larger correlation coefficients with gridded population (0.47 versus 0.09) and DN values of the new generation NTL image product (0.56 versus 0.52), i.e., the Visible Infrared Imaging Radiometer Suite day/night band image composite. In addition, total personal incomes of states are disaggregated to each pixel in proportion to the DN value of the pixel in the NTL images and then aggregate by counties. Personal incomes distributed by the upsampled NTL image are closer to the official demographic data than those distributed by the original stable light image. All of these results explore the potential of geotagged tweets to improve the quality of NTL images for more accurately estimating or mapping socioeconomic factors.

Locality-adapted kernel densities of term co-occurrences for location prediction of tweets

While geographical metadata referring to the originating locations of tweets provides valuable information to perform effective spatial analysis in social networks, scarcity of such geotagged tweets imposes limitations on their usability. In this work, we propose a content-based location prediction method for tweets by analyzing the geographical distribution of tweet texts using Kernel Density Estimation (KDE). The primary novelty of our work is to determine different settings of kernel functions for every term in tweets based on the location indicativeness of these terms. Our proposed method, which we call locality-adapted KDE, uses information-theoretic metrics and does not require any parameter tuning for these settings. As a further enhancement on the term-level distribution model, we describe an analysis of spatial point patterns in tweet texts in order to identify bigrams that exhibit significant deviation from the underlying unigram patterns. We present an expansion of feature space using the selected bigrams and show that it eventually yields further improvement in prediction accuracy of our locality-adapted KDE. We demonstrate that our expansion results in a limited increase in the size of feature space and it does not hinder online localization of tweets. The methods we propose rely purely on statistical approaches without requiring any language-specific setting. Experiments conducted on three tweet sets from different countries show that our proposed solution outperforms existing state-of-the-art techniques, yielding significantly more accurate predictions.

Social media and urban mobility: Using twitter to calculate home-work travel matrices

The proliferation of Big Data is beneficial to the study of mobility patterns in cities. This work investigates the use of social media as an efficient tool for urban mobility studies. In this case, the social network Twitter has been used, due to its wealth of spatial and temporal data and the possibility of accessing data free of charge. Using a database of geotagged tweets in the Madrid Metropolitan Area over a two-year period, this article describes the steps followed in the preparation and cleansing of the initial data and the visualisation of the results in Geographic Information Systems in the form of home-work matrices. The Origin-Destination matrices obtained were then compared with the official data provided by the Madrid Transport Consortium from the 2014 Synthetic Mobility Survey. The results of this comparison demonstrate that the level of precision offered by Twitter as a source of geographic information is adequate and efficient, thereby permitting a more in-depth analysis of flows between different zones of interest in the study area.

Census Tract Food Tweets and Chronic Disease Outcomes in the U.S., 2015⁻2018.

There is a growing recognition of social media data as being useful for understanding local area patterns. In this study, we sought to utilize geotagged tweets—specifically, the frequency and type of food mentions—to understand the neighborhood food environment and the social modeling of food behavior. Additionally, we examined associations between aggregated food-related tweet characteristics and prevalent chronic health outcomes at the census tract level. We used a Twitter streaming application programming interface (API) to continuously collect ~1% random sample of public tweets in the United States. A total of 4,785,104 geotagged food tweets from 71,844 census tracts were collected from April 2015 to May 2018. We obtained census tract chronic disease outcomes from the CDC 500 Cities Project. We investigated associations between Twitter-derived food variables and chronic outcomes (obesity, diabetes and high blood pressure) using the median regression. Census tracts with higher average calories per tweet, less frequent healthy food mentions, and a higher percentage of food tweets about fast food had higher obesity and hypertension prevalence. Twitter-derived food variables were not predictive of diabetes prevalence. Food-related tweets can be leveraged to help characterize the neighborhood social and food environment, which in turn are linked with community levels of obesity and hypertension.

Twitter-based traffic delay detection based on topic propagation analysis using railway network topology

Twitter has become one of the most popular social media platforms, evidently stirred by a very popular trend of event detection with many applications, including delay detection and traffic congestion on the public transport network. In this paper, we propose a Twitter-based railway delay detection method based on topic propagation analysis of geo-tagged tweets between railway stations. In particular, we aim to discover delay events and to predict train delays due to traffic accidents by analyzing topic propagation using railway network topology of real space. To realize this, first, we construct the topology of the railway network (the physical space) as a graph in which nodes are railway stations and edges are represented as routes between them. Then, we extract the topology of the social network that is mapped on the railway network, based on topic propagation analysis of accident delays between stations and by analyzing geo-tagged tweets of each station with a neural network. This allows us to observe the influence of delays on railway stations even if there are a few tweets on them and to predict stations affected by delays with the tweets which contain indirect topics about delays such as “crowded!” and “raining!”. Overall, this paper proposes the method which enables us to analyze the topic propagation of geo-tagged tweets in order to predict accident delays by considering the railway topology of real space. In addition, we also evaluate the performance of the proposed method on datasets derived from Twitter with the actual delay information from 488 stations of 62 routes in Tokyo area in Japan.

Air pollution lowers Chinese urbanites' expressed happiness on social media.

High levels of air pollution in China may contribute to the urban population's reported low level of happiness1-3. To test this claim, we have constructed a daily city-level expressed happiness metric based on the sentiment in the contents of 210 million geotagged tweets on the Chinese largest microblog platform Sina Weibo4-6, and studied its dynamics relative to daily local air quality index and PM2.5 concentrations (fine particulate matter with diameters equal or smaller than 2.5 μm, the most prominent air pollutant in Chinese cities). Using daily data for 144 Chinese cities in 2014, we document that, on average, a one standard deviation increase in the PM2.5 concentration (or Air Quality Index) is associated with a 0.043 (or 0.046) standard deviation decrease in the happiness index. People suffer more on weekends, holidays and days with extreme weather conditions. The expressed happiness of women and the residents of both the cleanest and dirtiest cities are more sensitive to air pollution. Social media data provides real-time feedback for China's government about rising quality of life concerns.

Exploring the uncertainty of activity zone detection using digital footprints with multi-scaled DBSCAN

ABSTRACTThe density-based spatial clustering of applications with noise (DBSCAN) method is often used to identify individual activity clusters (i.e., zones) using digital footprints captured from social networks. However, DBSCAN is sensitive to the two parameters, eps and minpts. This paper introduces an improved density-based clustering algorithm, Multi-Scaled DBSCAN (M-DBSCAN), to mitigate the detection uncertainty of clusters produced by DBSCAN at different scales of density and cluster size. M-DBSCAN iteratively calibrates suitable local eps and minpts values instead of using one global parameter setting as DBSCAN for detecting clusters of varying densities, and proves to be effective for detecting potential activity zones. Besides, M-DBSCAN can significantly reduce the noise ratio by identifying all points capturing the activities performed in each zone. Using the historic geo-tagged tweets of users in Washington, D.C. and in Madison, Wisconsin, the results reveal that: 1) M-DBSCAN can capture dispersed clusters with low density of points, and therefore detecting more activity zones for each user; 2) A value of 40 m or higher should be used for eps to reduce the possibility of collapsing distinctive activity zones; and 3) A value between 200 and 300 m is recommended for eps while using DBSCAN for detecting activity zones.

An Improved Density-Based Approach to Spatio-Textual Clustering on Social Media

Density-based spatial clustering of applications with noise (DBSCAN) is the most commonly used density-based clustering algorithm but may not be sufficient when the input data type is heterogeneous in terms of textual description. When we aim to discover clusters of geo-tagged records relevant to a particular point of interest (POI) on social media, examining only one type of input data (e.g., the tweets relevant to a POI) may draw an incomplete picture of clusters due to noisy regions. To overcome this problem, we introduce DBSTexC, a newly defined density-based clustering algorithm using spatio-textual information on social media (e.g., Twitter). We first characterize the POI-relevant and POI-irrelevant geotagged tweets as the texts that include and do not include a POI name or its semantically coherent variations, respectively. By leveraging the proportion of the POI-relevant and POI-irrelevant tweets, the proposed algorithm demonstrates much higher clustering performance than the DBSCAN case in terms of F 1 score and its variants. While DBSTexC performs exactly as DBSCAN with the textually homogeneous inputs, it far outperforms DBSCAN with the textually heterogeneous inputs. Furthermore, to further improve the clustering quality by fully capturing the geographic distribution of geo-tagged points, we present fuzzy DBSTexC (F-DBSTexC), an extension of DBSTexC, which incorporates the notion of fuzzy clustering into the DBSTexC. We then demonstrate the consistent superiority of F-DBSTexC over the original DBSTexC via intensive experiments. The computational complexity of our algorithms is also analytically and numerically shown.

Face off: Travel Habits, Road Conditions and Traffic City Characteristics Bared Using Twitter

The adequacy of traditional transport related issues detection is often limited by physical sparse sensor coverage and reporting incident/issues to the emergency response system is labor intensive. The social media tweet text have been mined so as to identify the complaints regarding various road transportation issues of traffic, accident, and potholes. In order to identify and segregate tweets related to different issues, keyword-based approaches have been used previously, but these methods are solely dependent on seed keywords which are manually given and these set of keywords are not sufficient to cover all tweets posts. So, to overcome this issue, a novel approach has been proposed that captures the semantic context through dense word embedding by employing word2vec model. However, the process of tweet segregation on the basis of semantic similar keywords may suffer from the problem of pragmatic ambiguity. To handle this, Word2Vec model has been applied to match the semantically similar tweets with respect to each category. Furthermore, the hotspots have been identified corresponding to each category. However, due to the scarcity of geo-tagged tweets, we have proposed a hybrid method which amalgamates Named Entity Recognition (NER), Part of speech (POS), and Regular Expression (RE) to extract the location information from the tweet textual content. Due to the lack of availability of the ground truth dataset, model feasibility has been validated from the existing data records (i.e., published by government official accounts and reported on news media) and the evaluation results signify that the stated approach identifies few additional hotspots as compared to the existing reports while analyzing the tweets.

Capturing and mapping quality of life using Twitter data

There is an ongoing discussion about the applicability of social media data in scientific research. Moreover, little is known about the feasibility to use these data to capture Quality-of-Life (QoL). This study explores the use of social media in QoL research by capturing and mapping people’s perceptions about their life based on geo-located Twitter data. The methodology is based on a mixed-method approach, combining manual coding of the messages, automated classification, and spatial analysis. Bristol is used as a case study, with a dataset containing 1,374,706 geotagged Tweets. Based on the manual coding results, three QoL domains were analysed. Results show the difference between Bristol wards in number and type of QoL perceptions in every domain, spatial distribution of positive and negative perceptions, and differences between the domains. Furthermore, results from this study are compared to the official QoL survey results from Bristol, statistically and spatially. Overall, three main conclusions are underlined. First, to an extent, Twitter data can be used to evaluate QoL. Second, based on people’s perceptions, there is a difference in QoL between neighbourhoods in Bristol. And, third, Twitter messages can be used to complement QoL surveys, but not act as a proxy for traditional survey results. The main contribution of this study is in recognising the potential Twitter data have in QoL research. This potential lies in producing additional knowledge about QoL that can be placed in a planning context and effectively used to improve the decision-making process and enhance quality-of-life of residents.

Development of a national-scale real-time Twitter data mining pipeline for social geodata on the potential impacts of flooding on communities

Social media, particularly Twitter, is increasingly used to improve resilience during extreme weather events/emergency management situations, including floods: by communicating potential risks and their impacts, and informing agencies and responders. In this paper, we developed a prototype national-scale Twitter data mining pipeline for improved stakeholder situational awareness during flooding events across Great Britain, by retrieving relevant social geodata, grounded in environmental data sources (flood warnings and river levels). With potential users we identified and addressed three research questions to develop this application, whose components constitute a modular architecture for real-time dashboards. First, polling national flood warning and river level Web data sources to obtain at-risk locations. Secondly, real-time retrieval of geotagged tweets, proximate to at-risk areas. Thirdly, filtering flood-relevant tweets with natural language processing and machine learning libraries, using word embeddings of tweets. We demonstrated the national-scale social geodata pipeline using over 420,000 georeferenced tweets obtained between 20 and 29th June 2016.

Event Detection and Recommendation Using Social Media for Hyperlocal Communities

Individuals often rely on newspapers and television to gather information about the communities they live in to be aware of and to engage with their communities. Social media have become increasingly important as a new source of local information. Thus far, the literature on using social media to create desirable societal benefits, such as in civic awareness and engagement, is still in its infancy in the Information System (IS) field. One key challenge in this research stream is to timely and accurately distill local information from noisy Twitter streams to community members. In this work, we develop several machine learning techniques and a mobile application utilizing these techniques to facilitate information seeking of local communities on Twitter. Specifically, we first propose a novel local event detection algorithm which scans spatially and temporally to detect local events as unusual spikes from a large-scale Twitter feeds. Our algorithm significantly improves several competing methods with much higher precision and recall in detecting local events on a real-world dataset of millions of geotagged tweets. Next, the volume of detected local events can be enormous, especially in large and populous local areas. This motivates us to develop a personalized local event recommender system that systematically integrates several contextual cues such as topical, geographical and social proximity in a learning to rank framework. Based on a comprehensive evaluation with two large-scale datasets of local events, the results show that our approach significantly outperforms competitive methods in precision as well as its robustness. Finally, we build a mobile application to demonstrate the practical value of our event detection and recommendation techniques. This application automatically infers and recommends neighborhood-specific information based on Twitter feeds. We test its effectiveness and usefulness through a comprehensive user study. The outcomes demonstrate that most participants prefer it over Twitter as it is easier to use for exploring their neighborhoods and provides better relevant information about their neighborhoods.