1. Querying and Exploring Geospatial
Data
A survey (to be added)
1.1
Querying spatio-textual (geo-textual) data streams
Selected publications:
§
SSTD:
A Distributed System on Streaming Spatio-Textual
Data, PVLDB 2020
§
STAR:
A Distributed Stream Warehouse System for Spatial Data. SIGMOD Conference 2020:
2761-2764 (Demo)
§
Distributed Publish/Subscribe Query
Processing on the Spatio-Textual Data Stream (ICDE
17)
§
Diversity-aware top-k
publish/subscribe on text stream (SIGMOD 15)
§
Temporal spatial-keyword top-k
publish/subscribe on geo-textual data stream (ICDE15 and demo in VLDB14)
§
Boolean spatial-keyword
publish/subscribe on geo-textual data stream (SIGMOD13 and demo in VLDB14)
1.2
Data exploration for spatial data: Region search & topic exploration
Selected publications:
§
SURGE:
Continuous Detection of Bursty Regions Over a Stream
of Spatial Objects (TKDE19, ICDE18)
§
Finding attribute-aware similar
regions for data analysis (PVLDB 19)
§
Efficient
Similar Region Search with Deep Metric Learning (KDD 18)
§
Efficient Selection of Geospatial
Data on maps for Interactive Visualized Exploration (SIGMOD 18)
§
Towards Best Region Search for Data
Exploration (SIGMOD 2016)
§
Topic Exploration in Spatio-Temporal Document Collections(SIGMOD 2016, VLDBJ19)
1.2 Spatial keyword queries
§
On
Spatial Pattern Matching (ICDE’17, VLDBJ’19)
§
Answering
the m-closest keywords query (SIGMOD 15)
§
Search
regions of interest for user exploration (VLDB14)
§
Distributed
spatial keyword querying on road networks (EDBT14)
§
An
evaluation of 12 geo-spatial indexes (VLDB13).
Code available here.
§
An
overview paper on spatial-keyword querying (invited
paper in ER)
§
Route
planning: answering queries like “a most popular route such that it
passes by shopping malls, restaurant, and pub, and the travel time is
within 4 hours.” (PVLDB12)
§
Efficient
processing of several types of spatial keyword queries (VLDB09, PVLDB10, SIGMOD11a).
Code for our SIGMOD11 paper is available here.
An extension of our SIGMOD 11 paper is published in TODS
§
Efficient
algorithms and cost models for reverse spatial-keyword k-nearest neighbor search (SIGMOD11b, TODS14)
§ Efficient spatial keyword search in trajectory databases (unpublished paper)
2.1 Intelligent transportation using trajectory
data
Selected Publications:
§
Online Anomalous Trajectory Detection with Deep
Generative Sequence Modeling, ICDE2020
§
Spatial Transition Learning on Road Networks with
Deep Probabilistic Models, ICDE 2020
§
Learning Travel Time Distributions
with Deep Generative Model. (WWW 2019)
2.1 Data driven smart city applications
§ Periodic-CRN:
A Convolutional Recurrent Model for Crowd Density Prediction with Recurring Periodic
Patterns (IJCAI, 2018)
§ Efficient
Similar Region Search with Deep Metric Learning (KDD 2018)
2.3 Spatial graph mining, POI
recommendation & prediction
§ Densely
Connected User Community and Location Cluster Search in Location-Based Social
Networks, SIGMOD2020
§
Context-aware
Deep Model for Joint Mobility and Time Prediction, WSDM 2020
§
More
work on POI recommendation in Project 4
3. Machine Learning
for Data Management
3.1 Deep learning for databases
§ Cardinality
and selectivity estimation
§ Workload
generation
§ Indexing
optimization
3.2 Deep learning for trajectory
similarity and search
Selected Publications:
§
Efficient
and Effective Similar Subtrajectory Search with Deep
Reinforcement Learning, PVLDB 2020
§
Deep
Learning for Trajectory Similarity Search (ICDE18, ICDE19)
§
Effective
and Efficient Sports Play Retrieval with Deep Representation Learning (KDD19)
4. Recommendation, POI recommendation and User
Behaviour Modeling
4.1 Recommendation and group recommendation
Selected publications:
§
HyperML:
A Boosting Metric Learning Approach in Hyperbolic Space for Recommender
Systems. WSDM 2020 (Best paper award
runner-up)
§
Global
Context Enhanced Graph Nerual Networks for
Session-based Recommendation, SIGIR 2020
§
Interact
and Decide: Medley of Sub-Attention Networks for Effective Group Recommendation
(SIGIR 19)
§
Group
Recommendation based on topic models(KDD14)
§
HME:
A Hyperbolic Metric Embedding Approach for Next-POI Recommendation, SIGIR 2020
§
A
new POI recommentdation approach, which performs
better than previous approaches in experiments (SIGIR 2015)
§
SAR: A sentiment-aspect-region model
for user preference analysis and POI/user recommendation. The model provides
explanations for recommendation results. (ICDE 2015)
§
A general graph model for
recommendation in heterogeneous networks and its applications in
event-based social networks (ICDE 2015)
§
Diversity-aware POI recommendation
(AAAI 2015)
§
Time-aware POI recommendation (SIGIR13, CIKM14). Datasets
available here
§
Mining significant semantic locations
from user generated GPS data for recommendation (PVLDB10)
§ W4: Discovering spatio-temporal
topics for individual users and its various applications, e.g.,
requirement-aware POI recommendation (KDD13, TOIS15).
Datasets available here
5. Mining Reviews, Social Media, and Forums
§
We consider the impact of users' attributes, time factor, and novelty decay (Repeated exposures of
an individual to an idea may have diminishing influence on the individual) for
finding influential users.
§
We develop techniques
for review mining and sentiment
analysis.
§
We also develop techniques for mining social media, including
Micro-blogs (e.g., Twitter), and Community Based Question Answering Sites
(e.g., Yahoo! Q&A).
Publications:
§
Inf2vec:
Latent Representation Model for Social Influence Embedding (ICDE 18)
§
DynaDiffuse:
A dynamic diffusion model for continuous time constrained influence
maximization (AAAI 15)
§
Finding
influential event organizers in event based social networks (SIGMOD14)
§
Influence
maximization with novelty decay (AAAI14)
§
Time
constrained influence maximization in social
networks ( ICDM12
, TKDE . Source
code)
§
Computing
top-k influential nodes (KDD10,
AAAI
11)
§
Detecting user intents from tweets
(AAAI 15)
§
Coarse-to-fine review selection via
supervised joint aspect and sentiment model (SIGIR14)
§
One seed to find them all: Mining
opinion features via association (CIKM12)
§
Geolocation prediction for social
images by exploring user profiles (JASIST14)
§
On predicting popularity of newly emerging
hashtags in Twitter (JASIST13)
§
Short text classification ( WWW12 poster,
evaluation paper JASIST ) and hierarchy
maintenance ( SIGIR12).
Annotated dataset for our SIGIR12 paper is
available here.
§
Using categorization information to
improve question search in community based question answering services ( CIKM09, WWW2010, TOIS12). Annotated dataset is
available here
§
Extracting Question-Answer pairs
from forums to build the QA database (SIGIR08, ACL08)
§
Routing questions to expert users ( ICDE09)
Acknowledgement: Some of these
projects are supported by grants awarded by Ministry of Education, NRF, IAF,
Roll-Royce, and Microsoft.