Page 41 - untitled
P. 41
Web, Internet, Networking, and Software (WINS)
Web, Internet , Networking, and Software (WINS)
ਂ݊Շଡ଼ᇍૢගٙഐό݈࿁fҢࡁٙ ҢࡁආБʕٙࠇྌ˸ܔྡҖʷʧࠦމ Internet Knowledge Extraction and Man- measurements in emerging and challenging network
ጳሳίӺνО˸͍ٙ˙όdԸܔϤ၇ ͦٙdԶԴ͜٫อᄣeࡌҷeʝਗXML༟ agement scenarios. We plan to develop several approaches
that can accurately, timely, and uncostly measure/
݈࿁ٙᅼۨd˸ʿνОਗ਼வԬᅼۨdໝྼྼ ࣘdҢࡁ͍ίක೯ɓ၇ʝਗʩForms-
This project focuses on Internet document monitor the network.
ਕɪf XMLd̴ঐಁXMLႧคପ͛˸ڌఊމਿᓾٙ extraction, classification and query processing. We
Ңࡁ࿁XMLʕٙ˖ۨ࿒່֛d̈ Դ͜٫ʧࠦdԨЪމXML˖ᇜ፨ኜf address the problem of extracting knowledge from In addition, using network measurement
web sites along two directions. In the first direction, results, we intend to develop QoS enhancement
ਞᅰʷʫ࢙ᅼۨٙ฿ׂdԨኽϤІਗܔ
'PSNT 9.- Ϟ˸ɨอٙतЍj we extract knowledge from web documents using an schemes for various network services/applications.
̈࿁Ꮠٙ˖ഐ᜕ᗇҏfϤධᅼۨމԴ͜
existing ontology. In the second direction, we recog- For instance, for mobile network applications, we
MLɓ၇ᆽڭۨ࿒τΌٙόႧԊdՈ৷චՌ (1) ̴˸ᒒеಁ˾ᏨݟႧجɪٙ፹ nize informative structure of a web site and hereof plan to develop an algorithm which can agilely Research Groups
ᅰၾਞᅰʷᅼଡ଼ഃ̌ঐ˸ආБబۨ࿒ٙXML Ⴌdה˸Ъ͜ʕٙ˖ᐼ݊ୌΥܸ֛ٙၤͦd to extract informative contents which are usually adapt network services, in accordance with real-
όணࠇٙਿᓾf͟ϤᅼۨdҢࡁɰኬ̈XML Դ͜٫ʔዄː༼ˀXMLႧجٙ፹Ⴌf coupled with undesired redundant information. We time network measurement results, in order to
are also interested in developing new methods, algo- optimize the user-perceived service quality for the
˖ᔷ౬ٙᅼଡ଼ʷଣሞf (2) Forms-XML ˕౪อᄣɓʩЇЪ͜ʩ
rithms, and technologies to extract knowledge from mobiles. Moreover, we plan to apply network mea-
Ϥ̮d࿁ XML ˖ٙЕݴஈଣ ٙቌڐᇍఖfνϤԴ͜٫ʔᆽ֛อᄣʩٙ web pages of a specific type, e.g., citation list or surement results to benefit emerging peer-to-peer,
(streaming processing) όdҢࡁɰഹ˓Ӻ͍ ͍ᆽЗໄf bibliography. overlay, mesh, sensor, ad hoc, and/or opportunistic
Research Groups
ʷٙ˙جd˸ᆽڭՉЭܓٙাኳ٤ගცӋe (3) Forms-XML Զ܄ႡʷʿฆછՓԴ͜ networks, so that not only can the network resource
In addition, the project also studies learning be effectively utilized, but QoS can also be greatly
˸ʿՉପ̈˖ഐ͍ٙᆽf ٫ʧࠦٙ̌ঐdܼ̍əᏐ͜όʧࠦʿ܄Ⴁʷ and management of knowledge from Internet docu- improved.
ίྼࠦ͜dҢࡁʊԴ͜ XML ˸ʿᗫ ˖(ஹЇ̮ٙCSSᅵόڌ)f ments, including the following topics: personalized
knowledge management, topic discovery, and event Formal Verifi cation
Ҧஔνַ̙࢝ΣඎྡdSVGɓࢁ வԬतЍԸdForms-XML̙މॹ˶
detection and tracking. The personalization system
˸༟ৃၣ༩މਿᓾٙɛɹήྡӻ୕fவࢁն XMLٝᗆٙԴ͜٫ପ͛Դ͜٫ʧࠦf Our research goal is to improve the quality
filters information for users based on user interests.
Τމၽᝄٟึήྡhttp: //tsm.iis.sinica. Topic discovery finds interesting topics via hyper- of computer systems. For the past decades, formal
ක׳όٙήଣ٤ග༟ৃҦஔ
edu.twٙӻ୕d̙ᜫԴ͜٫ί༟ৃၣ༩ɪd links and contents of Web documents. The goal of method has been used to ensure the correctness of
ڐϋԸήଣ༟ৃӻ୕Դ͜Όଢ༟ৃၣމ computer systems. We are developing and applying
Іٙ͟ค˸ʿൖᙂʷɛɹ౷ݟ༟ࣘණf event detection and life cycle tracking is to identify
ੂБձ೯̨̻࢝ٙઋҖ˚ू౷ʿdᗫٙ༟ࣘ specific events and distinguish the articles of the formal verification techniques to help system engi-
̤̮dεᏐ͜ჯਹʊҖϓʮึאᑌຑd
ʹ౬ᅺνήଣᅺͪႧԊGMLe̙ events from others from several document sources. neers build high-quality computer systems.
Ԩމəʹ౬༟ࣘϾ່֛༈ჯਹत֛ٙXMLႧ
ַ࢝ΣඎྡSVGٙ೯࢝ɰڢ੬Ҟdவ The presentation of a complicated event for reader Currently, the following projects are focused
คfXML༟ࣘࢫ၍ଣӻ୕͍နနϓᆞdXMLႧ comprehension is also an interesting research topic.
ேпόอ٤ග༟ৃӻ୕ٙක೯f͉ࠇ in the formal verification group:
ค̙ঐᜊϓਗ਼ԸΆุ༟ৃӻ୕ٙᅺᅼۨf
ཀ̘Դ͜வԬ༟ࣘʹ౬ᅺdภৣΌଢ༟ৃ WLAN/WMAN as Access Network (1) System-on-Chip verification:
Ңࡁ˸ҦஔԶᏐ٫ٙԉЍਞၾᗫᅺଡ଼
ၣҦஔ̨̻d˸ʿක׳ࡡᇁழʈՈdʊϓ
ᔌٙݺਗdतй݊ʫٙৌਕ༟ৃᅺ-XBRL This project focuses on the enhancement of
̌ٙක೯əɓࢁழʈՈd࿁Դ̨͜ᝄତϞٙ SoC systems require the integration of both
ၾषዝࣸόᅺ-HL7/CDAfҢࡁʊၾึࠇձ IEEE 802.11and 802.16 families in order to provide hardware and software components. But it is dif-
ਿ͉ήҖྡ༟ࣘࢫᅺʹ౬ࣸόᇜᇁٙᅰ QoS guarantee and multimedia communication, and
༟ৃ߅Ҧഃᗫჯਹٙኪ٫ଡ଼ϓӺྠd˸ ficult to integrate due to different development envi-
࠽ήҖྡd̙ਗ਼Չᔷ౬މ˸ήଣᅺͪႧԊ multi-hop access, such that they can be integrated as
೯࢝ʫٙXBRL taxonomiesၾක೯XBRLৌਕ ronments of hardware and software components. In
ձַ̙࢝ΣඎྡᇜᇁٙXML˖dԴவ a full-blown access network. Resource management this project, we would like to propose an integrated
༟ৃٙᗫᔷ౬ʱؓʈՈfҢࡁɰၾ̨ᝄᔼኪ and fairness are the major issues for QoS guarantee.
Ԭ˸ᔚᅺᇜᇁٙᅰ࠽ήҖྡd̙˸ܘ˙کٙ verification platform to check the correctness of
༟ৃኪึίཥɿषዝᅺՓࠈɪၡഐΥdߧ Beside theoretic study, we implement embedded
ίΌଢ༟ৃၣɪԴ͜f͉ࠇٙӺࠠᓃਗ਼׳ SoC systems. The idea is to develop a core verifica-
ɢՓ֛˸CDAމਿٙXML schemadԨ˸Ϥ system prototypes on Linux for performance evalu- tion tool which allows users to translate hardware
ί˸ᅺႧԊމਿᓾٙอήଣ༟ৃӻ୕dܔ ation purposes. Our studies cover MAC layer, net-
schema೯࢝Web-basedٙᕑҖӻ୕d੶ሜਗ਼ෂ୕ and software specifications to the core verification
ிהცٙӻ୕ᕑҖd˸᜕ᗇอٙ฿ׂfவԬอ work layer and transport layer protocols.
ٙᔼኪ༟ࣘࢫ༟ࣘᔷ౬މCDAٙᅺࣸόʘྼ language.
ٙήଣ༟ৃӻ୕dʔස̙ΥԸІεࡈԸ๕
Ъd˸ڮආ̨ᝄᔼᐕ৫הཥɿषዝٙ೯࢝ၾᏐ Network Measurement (2) OMocha model checker:
ٙମሯ༟ࣘd˸Όଢ༟ৃၣމੂБ̨̻dԨ̙
͜પᄿf
ԱԴ͜٫אԴ͜٫໊ٟٙცࠅאԴ͜ᐑྤdਂ Network measurement is a fundamental re- OMocha is a model checker for reactive mod-
อጳٙXMLႧค˚ूᄣεdᜑ್ࠅމ search problem in the field of computer networks.
ʿࣛٙ܄ႡʷяତfவධӺࠇٙͦᅺʘɓ ules. It serves as the core verification platform of
XML ༟ࣘ˖ܔͭԴ͜٫ʧࠦfՑͦۃމ The results of network measurements are critical for
уίԴ͜ක׳ٙ༟ࣘᅺeක೯ʈՈeၾҦஔ our System-on-Chip verification project. We have
˟dXML༟ࣘٙԴ͜٫ʧࠦே݊ఱत֛ٙXML efficient network design, management, and usage. also developed model checking algorithms and
̨̻dܔி̙࢝ͪϤ၇อ̌ঐٙήଣ༟ৃӻ With the growing popularity of emerging network
ႧคʿᏐࠠ͜อႡЪdϽᅇՑΆุॴXMLႧค implemented them in OMocha. We use CUDD and
୕f technologies, such as WiFi, Bluetooth, ZigBee, Wi-
ܼ̍ٙʩʿ᙮ਗႿɷϵdԴ͜٫ʧࠦ̀ chaff as the underlying BDD and SAT packages. To
Media, WiMAX, and GPRS/3G, it is becoming in- improve the quality of the model checker, we use
ءจਗ࿒XML༟ࣘഐeႧجᇍʿ̮ᝈٙя creasingly desirable to have a simple, accurate, end-
Objective Caml to implement high-level algorithms
ତഃdவࡈ˙όਗ਼ึڢ੬൬ࣛ൬ɢf to-end, and less intrusive approach to measure the in OMocha. Currently, OMocha has both BDD and
network. In this project, we intend to study network
SAT-based model checking algorithms.
30 31