mapping of geographical entity with meeting location from text for mobile 2011. 9. 30 kyoungryol kim
TRANSCRIPT
Mapping of Geographical Entity with Meeting Location from Text for Mobile
2011. 9. 30
Kyoungryol Kim
2
Table of Contents
1. Introduction
2. Background and Related Work
3. The Proposed System
4. Experimentation
5. Conclusion
1. Introduction1) Motivation2) Problem Definition3) Contribution
4
Motivation : IE Techniques on Smartphone
AppleiPhone
GoogleAndroid
RIMBlackberry
MSWindows
Phone
Time(Text)Recognition
Phone No.Recognition
Location(Text)Recognition
Adding event by recognized time
May 21, 2011
AddressRecognition
(Captured from Apple iPhone)
People start to pay attention to ‘Location Extraction’ technique
5
Motivation : Characteristics of Mobile Device
Memory Issue Android : 16MB heap size limit for each app. iPhone : No memory limit, but totally 512MB of RAM (iPhone4)
Speed Issue People who use mobile devices usually feel uncomfortable when it delays.
IE System Usually general Information Extraction system consists of many NLP modules
which consume more than 1GB memory, at least.
Client-Server model Client and server communicating model that every processing is done in server-
side. Need internet connection (3G or Wifi). If many clients request to the server at once, there will be overloading delays or
the server dies.IE Method Specialized on Mobile Device is Needed
6
Goal of this Research
Mapping Meeting Location text to the Geographical Locationand update it to online calendar in mobile device
The team meeting for the evaluation of first half of Univcast will be held.Date : July 19 (Sat) PM 2Location : Myeong-dong Dande-lion TerritoryDirections to Dandelion TerritoryAt Myeong-dong station gate num-ber 8, take a walk following the downtown then there it is on the first floor of YMCA building.
MeetingLocation
Name Myeong-dong Dandelion Territory
Address 1-1, Myeong-dong 1-ga, Jung-gu, Seoul, Korea
Geocode (37.5647312, 126.9861426)
Meeting AnnouncementExtractMeetingLocation
UpdateCalendar
startTime 2011-07-19T14:00
ExtractTime
7
Problem Definition
1. Extract meeting location from meeting announce-ment email
2. Disambiguate the extracted meeting location
회의는 오후 5 시 학생회관 101 호에서 열립니다 .
(Meeting will be held 5 PM at Room 101, Student
Union.)
2. Background and Related Works1) Information Extraction2) Geocoding3) Linked Open Data4) Local-Grammar Graph
10
Information Extraction
Information Extraction The objective is to construct structured database from free text or semi-
structured text (J. H. Kim 2004)
Related Work CMU Seminar Announcement Corpus 485 semi-structured seminar announcements Types : stime, etime, location, speaker Focus only on 4 types of information extraction, not on Geocoding.
Examples of seminar announcement
11
Geocoding
Geocoding The process of finding associated geographic coordinates, often expressed as lati-
tude and longitude, from other geographic data such as street addresses or zip codes (Geocoding, Wikipedia)
Related Work Geocode from the address
(Manov 2003; Jones 2003; Peng 2006; Pouliquen 2006; Volz 2007; Overell 2007; Goldberg 2007; Kauppinen 2008)
The big issue of the research is disambiguation of address (Pouliquen et al. 2006)
1. Multi-referent ambiguity two different geographic locations share the same name, e.g. "Cambridge" is it Cambridge, UK or Cambridge, Massachusetts?
2. Name variant ambiguity the same location has different names,
3. Geoname-Non Geoname ambiguity where a location name could also stand for some other word such as a person name or nouns, e.g. Metro as the city in Indonesia vs. Metro as the subway system
Focus only on Geocoding address, not all location entity e.g. "Room 101, Student Union, Hanyang University"
12
Linked Open Data
Linked Open Data URL : http://linkeddata.org The project aims to identify data sets that are available under open licenses, re-pub-
lish these in RDF on the Web and interlink them with each other Geographic Datasets are growing rapidly For only few Korean Geographical data included in LOD, we regard set of open ge-
ographical data as Linked Data, in this research.
March 2009 September 2010 September 2011
13
Local-Grammar Graph
Local-Grammar Graph The language description model which is to perform automatic analysis
and generation of natural language text, information extraction, using local language information in the form of Finite-State Automata. (J. Nam 2006)
Help to increase efficiency and accuracy by lexicalizing the knowledge forming grammar readability by consisting grammar as Directed Acyclic Graphs.
Various omission and permutation can be described which cannot be done by rules or specific features.
Example of LGG for 176 kinds of French wine
un vin rouge de Bordeauxun vin de Bordeaux rougeun rouge de Bordeauxun Bordeaux rougeun Bordeauxun rouge....du vin d'Alsace blancdu vin blanc d'Alsacedu blanc d'Alsacede l'Alsacede l'Alsace blancdu blanc
3. The Proposed System1) Preliminaries2) Overall Architecture3) Extraction Module4) Disambiguation Module
17
Overall Architecture
ExtractionModule
DisambiguationModule
Query DisambiguatedResult
MobileDevice
Server
Linked Data
Finite-StateTransducers
INPUT
OUTPUT
제목 : 팀장회의 공지 2008 년도의 마지막 팀장회의가 11 월 22 일 토요일 오후 2 시 종로 토즈에서 열립니다 . 재계약 그리고 명함 배부가 이뤄질 예정이니 팀장님 , 그리고 차기팀장님들 모두 와주시기 바랍니다 . 오시는 길 : 종로 종각역 4 번 출구에서 내려서 100m 정도 걸어오시면 오른쪽에 있습니다 .
팀장회의 공지
장소
명칭 종로 토즈
주소대한민국 서울특별시 종로구 종로 1.2.3.4
가동 84-8
GPS 좌표
(37.569914, 126.984924)
Template Generator
PersonalGeoData
18
Extraction Module (1/2)
1. Construct Local-Grammar Graph (LGG) Find local patterns around meeting location, inductively. Scope of local patterns :
Previous/Next/Current sentence including meeting location. Describe local patterns with 110 information types under 7 categories.
Location, Time, Title, Actor, Label, Connecting words, Etc. e.g. ‘ 장소 : ‘ is ‘locLbl’ information type under ‘Label’ category.
2. Convert LGG to Finite-State Transducer (FST)
3. Extract Meeting Location by FST
2. 학술대회 일정 : 2003 년 5 월 17 일 ( 토요일 ) 10:30 ~ 16:303. 학술대회 장소 : 성공회대학교 피츠버그관4. 학술대회 순서
19
Extraction Module (2/2)
Category of LGG for Meeting Location
1. 개최장소 1 개
1.1. 장소
1.1.1. 장소
1.1.2. 장소 1_1 | 장소 1_2
1.1.3. 장소 1_1 | 장소 1_2 | 장소 1_3
1.2. 장소 + 랜드마크
1.2.1. 장소 | 랜드마크
1.2.2. 장소 1_1 | 장소 1_2 | 랜드마크
1.2.3. 장소 | 랜드마크 1 | 랜드마크 2
1.3. 장소 + 주소
1.3.1. 장소 | 주소
1.3.2. 장소 1_1 | 장소 1_2 | 주소
1.3.3. 장소 1_1 | 장소 1_2 | 장소 1_3 | 주소
1.3.4. 장소 | 랜드마크 | 주소
2. 개최장소 N 개 (N>1)
2.1. 개최장소 2 개
2.2. 개최장소 3 개
2.3. 개최장소 4 개
1. 일시 및 장소 : 2010. 5. 12( 수 ) 14:00~16:00, 무역협회 중회의실
( 삼성동 트레이드 타워 51 층 )
3. 장 소 : 울산광역시 울주군 상북면 등억리 27 번지
먹고쉬었다가 (052-263-1206)
20
Disambiguation Module (1/2)
Problem Multi-reference ambiguity (Pouliquen et al. 2006)
two different geographic locations share the same name e.g. "Cambridge" is it Cambridge, UK or Cambridge, Massachusetts?
Disambiguation by Linked Data Personal Geo Data
Personalized OpenStreetMap User can map and save geographical location to the ‘meeting location’ (should be applied, consulting by Claus at Leipzig Univ.)
Open Geo Data Naver Local Search API Yahoo! POI Search API Seoul Bus-stop DB
Disambiguation by applying Ranking algorithm (idea will be borrowed from meta-search researches) disambiguate with 1st ranked geographical location
21
Disambiguation Module (2/2)
Personal Geo Data
Query : 동측식당Email : [email protected] Data
Personal Geo Data
<36.369051,127.363757>
NaverLocal API
Yahoo!POI API
SeoulBus-stop
Open Geo Data
Disambiguation
동측식당 <37.19051,123.363757>동측식당 <36.347001,127.396285>동측식당 <36.998166,126.894287>
동측식당 <37.55111,126.93219>.......
동측식당<36.369051,127.363757>
4. Experimentation1) Experiment Data2) Extraction Module3) Disambiguation Module
23
Experiment Data
Meeting announcement corpus1101 meeting announcementsCollected from the web, with keyword ‘notice’Annotation
10 types of term, 13 types of relation 3 human annotators with COAT annotation toolkit
24
Extraction Module
Exp1. Extraction speed/memory comparison Baseline system : ML based system Dataset :
already gathered corpus (training/test set)
Exp2. Extraction performance comparison Baseline system : ML based system Evaluation : Precision/Recall/F-measure Dataset :
already gathered corpus (training/test set) newly gathering corpus
(Experimentation should be followed)
25
Disambiguation Module
Exp1. Accuracy in distance 6 types of distance :
0≤x≤100m, 100m≤x<1km, 1km≤x<2km, 2km≤x<3km, 3km≤x<5km and 5km≤x
Exp2. Accuracy Improvement with Personal Geo Data Evaluation :
hard to show the performance show some scenarios how can it be applied so that it can improve accuracy.
Exp3. Performance of Ranking Algorithm comparison
Exp4. Disambiguation speed/memory comparison processing and communication speed/memory comparison on Server vs. on Mobile device
(Experimentation should be followed)