Log on:
Powered by Elgg

There are no events for this day. Would you like to add one?

You must be logged in to add events in this calendar.


Sam Ojo :: Activity :: Just Me

People: Everyone | Friends & Community | Inbox | Just Me
Display: Full-text | Summary
Include: Blog Posts | Blog Comments | Files | Wiki Page | Wiki Comments

<< Older

Page 1 of 2

samueloo | page | Apr 30, 2009 - 12:02am

Chapter 3 of Always On

Baron’s chapter on controlling the volume – everyone a language czar addresses issues surrounding consumer information inflow. Issues like information overloads and complexity management. Our age is witnessing an unprecedented flow of information. It is like a fire hole turned on and directed on us. Fortunately, language technologies can help in controlling the volume of information inflow and multitasking.

Language technologies like IM, email, mobile phone, (and Chat) can help in controlling information overload in many ways. On mobile phones for example, a list of who to allow or disallow can be created, thereby helping to filter unwanted conversations. The affordance of computer mediated communication incentivizes our innate ability to manage conversation and interpersonal relationship. Using language tech like IMs, email, and chat, we are able to control our online presence and information flow.

Language technologies also enhance our ability to manage complexity. We are able to multitask conversation by and large enhancing our ability to orchestrate interpersonal communication.

Chapter 7 of The Cult of the Amateur

Keen’s argument in 1984 (version 2.0) chapter centers on issues of ownership, sharing, and retention of consumer information outflow. These outflows are captured through Web 2.0 technologies. Ownership issues bother on the ownership of personal data, queries and cookies – no clear responsibility, jurisdiction, and accountability.

Personal information objects are collected, stored, shared without the consent of the individual who originated the data. The absence of statue, policy, or program that could guide the collection, storage and sharing of these personal data has led to abuses. Examples are loss of privacy suffered by AOL user #711391, release of search data of 658,000, and more.

The loss of privacy and more importantly the loss of individuality in the world of Web 2.0 is a ticking privacy time bomb without the necessary civil discourse, debate, and policies that should guide companies, companies like Google and other search company in the management and commercialization of personal data, queries, and cookies. It is high time to begin to engage in such civil discourse and policy formations.

 

Experiment

Objective

This study explains the relationship between types of resumes and job offer by examining the impact of writing style on how two sets of resumes are perceived so that applicants can increase their chance of being offered a job. Studies have suggested that a resume is a job marketing tool that reveals a bird view prospect of an applicant. However little is known about the impact of writing style and clear communication on employers' perceived impression. Hence this study explore whether differences in writing style affects initial resume selection made by employers.

Design

This study hypothesize that resume that Resumes that have no syntax or semantic errors raise positive perception of and acceptance by employers. An experiment was designed to test the null hypothesis of no difference in the perception of employer when resume with or without syntactic and semantic errors are processed.

A matched-participant experimental design was adopted to take advantage of both between-participant and within-participant experimental designs. Two conditions are defined: R1, an experimental group of resume that describes bad resume with syntax and semantic errors, R2, a control group of resume that describes good resume without syntax and semantic errors. With this design we able to a validity that eliminate carry over effect, test effect, and demand characteristics.

Participants were chosen by random sampling and assigned to the two groups randomly. Because of this random sampling we ensure that there are no major difference between the two groups. Furthermore, we looked match education degree of each member in each group such that they all have first college degree to control educational competence confounding that might influence judgment. Next 16 questionnaires were administered while 15 participants responded, a response rate of 92%.

Measurement

Three consequent variables were devices to measure the overall impression of a resume. These include experience, interpersonal relationship, and likelihood of resume acceptance. Likert scale was used to capture the key the construct measurement answers. Thus the data level of answer data is interval having equal distance. Having performed a preliminary exploration of the data, the distribution is approximately normal. Hence we use parametric one-tailed paired T test.

Results

The data distribution shows that the sample of 16 represent a web technology job applicant population that writing style has statistical significance at the .001 level on the employers perspective and the likelihood of an applicant resume being selected.

The result shows a T(obt) = 4.94 for a one-tailed paired test with 18 degree of freedom. The critical value of t is 1.864, meaning the T(obt) falls beyond the critical value. Thus the null hypothesis is rejected. Thus, the proposition that resumes that have no syntax or semantic errors raise positive perception of and acceptance by employers is supported.

Conclusion

The results of this study support that resumes that have no syntax or semantic errors raise positive perception of and acceptance by employers. Hence, applicants in this recession season that adopt better writing style in composing their resumes will likely raise their chance of being positively perceived, thereby improving their chance of getting job offers.


[More]

samueloo | page | Apr 22, 2009 - 11:56pm

Language is like a child, it grows; but if a child is allowed to just grow without any guide, chances are pretty high that the child will become an asocial. English language has evolved and is still evolving but its selection mechanism needs a support system that enables both the descriptive (or informal) and prescriptive (or formal) forms to flourish, which is a major concerns expressed by Baron (2008) in chapter 8 and 9.

Chap 8: Whatever: is the internet destroying language?

Language whether in spoken or writing form exhibits the identity and social hierarchy of a people. And English for that matter depicts a ruled-governed behavior of the society depending on whether the setting is formal or informal. Formal or prescriptive mode deals with the proper syntax and semantics while informal or descriptive is heavily used in everyday informal setting.

Language technologies are having an unprecedented impact on both spoken and writing English in recent times. Technologies including IM, email, texting, chatting, blogging, and mobile phones have accentuated the use of English. They have strengthened the role of writing as a representation of the informal spoken language, informal tones and forms such as b/c, U, or R are now common. These technologies have however accelerated the uncertainty in determining which word(s) should be hyphenated or become one word. For example, the move from ‘news paper’ to ‘newspaper’ or ‘off-line’ to ‘offline’ has become vague.

Chap 9: Gresham’s Ghost: challenges to written Culture

The current written culture appears to be ignoring the advice given by Sir Thomas Gresham to Queen Elizabeth I in 1558 during her ascension to the throne that the good and bad coins cannot circulate together. This advice came on a heel in history when her predecessors Henry VIII and Edward VI had notoriously reduce the silver content of English coin which led to an economic disaster. A good coin has an appropriate percentage of precious metals while a bad coin exhibits worn off or debased metals. Good coins are hoarded while bad coins are easily exchange in circulation. In like manner, if efforts are not geared towards clearly showing the public what can be trusted as a gold standard for formal writing, sloppiness and the ‘whatever’ theory of language will characterize orthography.

Other challenges include reading, writing, authorship, copyright, publishing, and language standards. Linguistic efforts need to address these challenges if the ‘whatever’ attitude will be reversed.

Communication in the next 50 years

Writing Culture. The language technologies will continue to evolve with their attending affordance. We might work with less print materials as we see more advancement in the development of hardware and software to facilitate annotating online text so as to rival the affordances of paper.

Speaking Culture. This generation and the next will continue to witness the use of new works especially those involving abbreviations commonly used in IMs, texting, emailing, chatting, and mobile phone, by and large galvanizing our interpersonal ability.

Ability to handle complexity. Language technologies will continue to enhance our ability to multi-task, an affordance that will continue to play a crucial role in enhancing our ability to handle complexity.

Social lifestyle. We are likely to be involved in dualism communication, a concept that enables an individual to exhibit two characters: a physical and virtual character. Individuals will own averter and participate deeply in the second life communication.

For some, duality of discourse management will be a lifestyle. Drawing support from structuration theory (Giddens, 1984; Walsham, 2002), a duality discourse management is a situation where social structure exists in the virtual world which is alive in the mind of an individual. An individual action is rationalized based on what is the accepted mental norm vis-à-vis that of the virtual world.

In the final analysis, the ability to analyze, coordinate, and synthesize multiple streams of communications will galvanize human cognitive power to a dimension that has never been experienced before. Hopefully, these language technologies will be supported in a manner that enables a healthy impact of both the speaking and writing culture.

References

Baron, N. S. (2008). Always on: Language in an online and mobile world: Oxford University Press.Giddens, A. (1984). The constitution of society. Cambridge, UK: Polity Press.

Walsham, G. (2002). Cross-cultural software production and use: A structurational analysis. MIS Quarterly, 26(4), 359-380.


[More]

samueloo | page | Mar 19, 2009 - 2:51am

Part A. Chapter Summary: ‘Truth and lies’ in the chapter 3 of Cult of the Amateur

In chapter 3, Keen (2007) examines various threats posed to the truth of information received from the Web2.0 infrastructure. Threat to reliability, threat to accuracy, ease of repudiation, and threat to credibility: all orchestrated by anonymity, scaling, and wider access characteristics of the Web 2.0.

Threat to accuracy. The nature of information flow in Web 2.0 made it difficult to access the accuracy of information. In Web 2.0, all are authors, no gatekeepers (i.e. editors, fact checkers, regulator, administrators, liar catchers). This inability to ascertain the truth in the web 2.0 community is a threat to civic culture, a threat able to silence public discuss, advance fake culture, and distort world views.

Threat to reliability. Since information contributors are anonymous, it is difficult to ascertain the integrity of the information presented. This threat is illustrated by Washington Post’s summary of the farce scripted lonelygirl15 clip which was covertly scripted as an advert in YouTube but later identified as a sham. Other examples of fraudulent acts include ‘splogs’ (i.e. a combination of spam and blogs aimed at clogging up the blogpsphere with 900,000 post per day), ‘paid to read’ rings (who manually click on links which advertisers can bill), click-fraud scheme, flogs, botnets, and clickbot.

Ease of Repudiating. Under anonymous, no one can be held accountable for Web2.0 contents. This is a threat to credibility. Vernon Robinson’s distasteful attack on Brad Miller – a distortion of the truth blamed on somebody is stack example.

Threat of corrupted search engine. Search engine especially Google search can be manipulated or corrupted to achieve selfish or malicious goals. Popular manipulations include ‘Google bombing’ – hyperlinked and cross-linking pages aimed at circumventing search algorithms to list biased pages on top of search lists.

User generated corruption. Google bombing might be use to sway popular opinion. This threat is illustrated in the blog of senatorial candidate Jon Kyle.

Limitation of search algorithm based on collective behavior. Sites that track reading behaviors of a community and then make recommendation based on aggregated preference may limit access to fair and balanced information which might lead to distorted world view.

Threat of masquerading and impersonating collective wisdom. Wall street journal research reveals 30 users at Digg who were responsible for 25,000 recommendation rankings in a community of 900,000. And a user in netscape.com was found responsible for 13% of all the stories on the site for fourteen days. At any rate, collective wisdom pulled from the crowd is not often very wise. Thus anonymous influencers at Digg or Reddit, anonymous editors at Wikipedia, and anonymous film-maker on Youtube cannot be trusted.

In all, the author submits that the experts are the arbiter of the truth of information which in my opinion is questionable. Inasmuch as the Web 2.0 allows some benefits in terms of access to information, it will flourish. However, more studies and debates are called for to conceptualize and implement a clearing house mechanism that can establish oversight and credibility in some critical aspects of Web 2.0.

Part B. System Design to capture Liars

A Company A has been successful in marketing its pain therapy Opioid agonist drug. Opioid agonist drug works best for smoker when administered through parenteral route. But there is a lie in popular social networks including Digg and Reddit’s blogs and emails that Company A’s opioid agonist drug is a killer and that it is no good compared with that of Company Z’s.

Description of a system of components

In order to catch liars’ URLs and contents, a rule-based text analysis and tagging processes are employed as follows:

1. Blog and email texts that mentioned the company name "Company A" are obtained as the domain text from the Internet.

2. Domain texts in (1) are tagged for keywords and BIGRAMs using GATE processing resources (Tokenizer, Sentence splitter, and POS tagger).

3. Entities from tagged tokens in (2) are extracted using corpus resources.

4. Extracted entities in (3) are classified into a tabular data with probability of lie sense.

5. A graphic presentation of liars’ threat data in (4) and their trends are shown on trend charts.

Description of components

The entire system comprises of six components including input texts, GATE processing resources, language resources, entity extraction, and graphic analysis presentation. The system's architecture in the attached liar catching system file shows the workflow required among the processing and language resources.

Input texts. Texts are obtained from blogs and emails from social network sites like Digg or Reddit.

GATE processing resources. These resources include Tokenizer, Sentence splitter, Orthographic coreferences, POS tagger, and JAPE. JAPE is used to implement annotations of (1) sentences with fewer or none of first-person pronouns including: I, me, mine and (2) sentences with fewer exclusionary words including: but, nor, except, and whereas.

Language resources. These resources constitute the system’s corpus which includes Company product website pages, UMLS database, MyDoctor website pages, Insurance claim website pages, and LIAR_TEXT gazetteer. Using list of NLS’ BIGRAM concept, a LIAR_TEXT gazetteer is compiled using the word list form psychology professor James Pennebaker (Mackey, 2009) liar watch word list.

Entity extraction. This component extracts lie keywords and bigrams and attaches them with probability weights.

Graphic analysis presentation.  This presentation shows liars threat and their trends for decision making.

An Example

"Company A’s Opioid agonist drug is no good for smoker. It is a killer drug. It has killed many people. Use company Z pain medicine instead."

Extracted keywords from the above liar concept statement are:

Company A

Opioid agonist

Drug_is

No_good

For_smoker

Killer_drug.

Killed_many

Many_people

company Z

pain

medicine_instead

These words are grouped in a lie tabular form using the company product documentation to cross-check the truth of drug administration. A presentation report in form of pie and trend charts is made.

Example of the corpus: Company A product webpage

Company A website documented authentic Opioid agonist drug’s administration which entails

1. Route administration including parenteral, transdermal, spinal and nasal methods

2. Specific characteristics including onset of action, duration of action

3. Possible uses including use in renal insufficiency and neuropathic pain

4. Side effects including constipation and dry mouth

5. Nursing role starting person who smoke or drink alcohol or both should use larger doses of opioids to obtain pain control

6. Dosage entails proving 650 mg every 4 hours to relief mild or moderate pain

// Example of a JAPE to match lie strings

// Author: Olusola Samuel-Ojo

Phase:lieCatcher

Input: Token SpaceToken

Options: control = appelt

Rule: lieCatcher

(

//match I, me, mine, but, nor, except, whereas

(

({Token.string=="I"}|{Token.string=="me"}|{Token.string=="mine"}|

{Token.string=="but"}|{Token.string=="nor"}|{Token.string=="except"}|

|{Token.string=="whereas"})

):left

-->

{

gate.AnnotationSet toRemove = (gate.AnnotationSet)bindings.get("left");

annotations.removeAll(toRemove);

//get the tokens

java.util.ArrayList tokens = new java.util.ArrayList(toRemove);

//define a comparator for annotations by start offset

Collections.sort(tokens, new gate.util.OffsetComparator());

String text = "";

Iterator tokIter = tokens.iterator();

while(tokIter.hasNext())

text += (String)((Annotation)tokIter.next()).getFeatures().get("string");

gate.FeatureMap features = Factory.newFeatureMap();

features.put("kind", "word");

features.put("string", text);

features.put("length", Integer.toString(text.length()));

features.put("orth", "apostrophe");

annotations.add(toRemove.firstNode(), toRemove.lastNode(), "Token", features);

}

References

Keen, A. (2007). The cult of the amateur: Doubleday.

Mackey, M. (2009). How to spot a liar. Retrieved Mar 18, 2009, 2009, from http://www.rd.com


[More]

samueloo | page | Mar 19, 2009 - 12:07am

Chapter Summary: ‘Truth and lies’ in the chapter 3 of Cult of the Amateur

In chapter 3, Keen (2007) examines various threats posed to the truth characteristics of information received from the Web2.0 infrastructure. These threats include threat of reliability, accuracy, repudiation, and credibility: all orchestrated by anonymity, scaling, and wider access.

Accuracy. The nature of information flow in Web 2.0 made it difficult to access the accuracy of information. In Web 2.0, all are authors, no gatekeepers (i.e. editors, fact checkers, regulator, administrators, liar catchers). This inability to ascertain the truth in the web 2.0 community is a threat to civic culture, a threat able to silencing public discuss, advancing fake culture, and distort world views.

Reliability. Since information contributors are anonymous, it is difficult to ascertain the integrity of the information presented. Washington Post’s summary of the farce scripted lonelygirl15 clip that was covertly scripted as an advert but attracted huge audience in YouTube, which was later identified as a sham. Other examples include ‘splogs’, i.e. a combination of spam and blogs aimed at clogging up the blogpsphere with 900,000 post per day, ‘paid to read’ rings who manually click on links which advertisers can bill, click-fraud scheme, flogs, botnets, and clickbot.

Ease of Repudiating. Under anonymous, no one can be held accountable for the Web2.0 contents. This is a threat to credibility. Vernon Robinson’s distasteful attack on Brad Miller – a distortion of the truth blamed on somebody is stack example.

Corrupted search engine. Search engine especially Gooogle search can be manipulated or corrupted to achieved selfish or malicious goals. ‘Google bombing’ – hyperlinked and cross-linking pages aimed at circumventing search algorithms to list biased pages on top of search lists.

User generated corruption. Strategies like google bombing might be use to sway popular opinion. Example of senatorial candidate Jon Kyle.

Limitation of search algorithm based on collective behavior. Sites that track reading behavior of a community and then make recommendation based on aggregated preference.

May limit access to fair and balanced information which might lead to distorted world view.

Threat of masquerading and impersonating collective wisdom. Wall street journal research reveals 30 users at Digg responsible for 25,000 recommendation rankings in a community of 900,000. And a user in netscape.com responsible for 13% of all the stories on the site for fourteen days. At any rate, collective wisdom pulled from the crowd is not often very wise. Thus, anonymous influencers at Digg or Reddit, anonymous editors at Wikipedia, anonymous film-maker on Youtube cannot be trusted. the ‘arbiters of truth should be the experts’ p.96.

In all, the author submits the experts are the arbiter of the truth which is questionable. Inasmuch as the Web 2.0 allows some benefits, it should be allowed to flourish but under a watchful eye to establish credibility in some aspects.

 

Part B. System Design to capture Liars

A Company A has been successful in marketing its pain therapy Opioid agonist drug. Opioid agonist drug works best for smoker when administered through parenteral route. But there is a rumors in popular social networks including Digg or Reddit blogs and emails that Company A’s Opioid agonist drug is a killer drug and that it is no good compared with that of Company Z.

Description of a system of components

In order to catch liars’s URLs and contents, a text analysis process is defined as follows:

1. Blog and email texts that mentioned the company name "Company A" are obtained as the domain text from the Internet.

2. Domain texts are tagged for keywords and BIGRAMs using GATE processing (Tokenizer, Sentence splitter, and POS tagger.

3. Entities are extracted using corpus resources

4. Extracted entities are classified into a tabular data with probability of lier sense.

5. A graphic presentation of liars threat and their trends are shown on a trend charts.

Description of components

 

The entire system comprises input texts, GATE processing resources, language resources, entity extraction, and graphic analysis presentation. Please see figure 1 showing system architecture by following the link below.

…………

Input texts. Texts are obtained from blogs and emails from social network sites like Digg or Reddit.

GATE processing resources. These resources include Tokenizer, Sentence splitter, orthographic coreferences, POS tagger, and JAPE. JAPE is used to implement annotations of sentences with fewer or none of first-person pronouns including: I, me, mine and sentences with fewer exclusionary words including: but, nor, except, and whereas.

Language resources include Company product website pages, UMLS database, MyDoctor website pages, Insurance claim website pages, and LIAR_TEXT gazetteer. Using list of NLS’ BIGRAM concept, a LIAR_TEXT gazetteer is compiled using the word list form psychology professor James Pennebaker (Mackey, 2009) liar watch word list.

Entity extraction entails extracting liar keywords and bigrams and attaching probability weights.

Graphic analysis presentation shows liars threat and their trends.

An Example

"Company A’s Opioid agonist drug is no good for smoker. It is a killer drug. It has killed many people. Use company Z pain medicine instead."

Extracted keywords from the above liar concept statement are:

Company A

Opioid agonist

Drug_is

No_good

For_smoker

Killer_drug.

Killed_many people.

Many_people

company Z

pain

medicine_instead.

These words are grouped in a liar tabular form using the company product documentation to cross-check the truth of drug administration. A presentation report in form of pie and trend charts are made.

Example of the corpus: Company A product webpage

Company A website documented authentic Opioid agonist drug’s administration which entails

1. Route administration including parenteral, transdermal, spinal and nasal methods

2. Specific characteristics including onset of action, duration of action

3. Possible uses including use in renal insufficiency and neuropathic pain

4. Side effects including constipation and dry mouth

5. Nursing role starting person who smoke or drink alcohol or both should use larger doses of opioids to obtain pain control

6. Dosage entails proving 650 mg every 4 hours to relief mild or moderate pain

 

References

Keen, A. (2007). The cult of the amateur: Doubleday.

Mackey, M. (2009). How to spot a liar. Retrieved Mar 18, 2009, 2009, from http://www.rd.com

 


[More]

samueloo | page | Feb 10, 2009 - 11:23pm


[More]

samueloo | page | Feb 10, 2009 - 11:21pm


[More]

samueloo | page | Feb 10, 2009 - 11:00pm


[More]

samueloo | page | Jan 7, 2009 - 4:06pm

This page holds work for the Fall 2008 IS 328 course with Dr. Gudea. You can view other people's portfolios by clicking on the IS328 tag on the bottom of this page, or by editing this page and looking in the sidebar.


[More]

samueloo | page | Jan 6, 2009 - 5:22am

This page holds work for the Fall 2008 IS 328 course with Dr. Gudea. You can view other people's portfolios by clicking on the IS328 tag on the bottom of this page, or by editing this page and looking in the sidebar.


[More]

samueloo | page | Jan 6, 2009 - 3:18am


[More]

<< Older

Page 1 of 2