Tuesday, November 30, 2010

eHarmony Brazil 3 months

eHarmony Brazil, a 3 months old site, can not take off, slowly dying. The summer will begin soon here in South America (also in Australia).

Wednesday, November 24, 2010

summary of recommender systems

Recommender systems (a.k.a recommendation engines) can be based on:

- past actions (as the formely Beacon at Facebook)

- a pattern of personal preferences (by collaborative filtering, as the actual one at Facebook) The main disadvantage with recommendation engines based on collaborative filtering is when users instead of providing their personal preference try to guess the global preference and they introduce bias in the recommendation algorithm.

- personality traits of users.

Personality Based Recommender Systems are the next generation of recommender systems because they perform FAR better than Behavioural ones (past actions and pattern of personal preferences)
That is the only way to improve recommender systems, to include the personality traits of their users.
Have you seen they need to calculate personality similarity between users?
Have you seen there are different formulas to calculate similarity?
In case you did not notice, recommender systems are morphing to .......... compatibility matching engines!!!
They mostly use the Big5 to assess personality and the Pearson correlation coefficient to calculate similarity.
Researchers in the Personality Based Recommender Systems arena are also testing different / novel formulas to calculate similarity, useless at all because they use the Big5 to assess personality of users.

That is nothing new, nothing innovative. Online Dating Sites like eHarmony, Parship, Be2, MeeticAffinity and others had been calculating personality similarity between prospective users since several years ago with low successful rates, with a low effectiveness/efficiency level of their matching algorithms (less than 10%) because they use the normative Big5 or ipsative proprietary models instead -like Chemistry or PerfectMatch- to measure personality traits.

No one is using the 16PF5 to assess personality of members.
No one calculates similarity with a quantized pattern comparison method.
No one can show Compatibility Distribution Curves to each and every of its members.

Psychological-based Recommenders Systems


More about Personality-based Recommender Systems

The NEW era of Personality Based Recommender Systems

Personality Based Recommender Systems

The PLAGUE of recommender systems

Recommender Systems and the Social Web


A Novel K-Means Based Clustering Algorithm for High Dimensional Data Sets

Recommender System for Online Dating Service

Tuesday, November 23, 2010

Predicting relationship and life satisfaction from personality

"Predicting relationship and life satisfaction from personality in nationally representative samples from three countries: the relative importance of actor, partner, and similarity effects."


Abstract
Three very large, nationally representative samples of married couples were used to examine the relative importance of 3 types of personality effects on relationship and life satisfaction: actor effects, partner effects, and similarity effects. Using data sets from Australia (N = 5,278), the United Kingdom (N = 6,554), and Germany (N = 11,418) provided an opportunity to test whether effects replicated across samples. Actor effects accounted for approximately 6% of the variance in relationship satisfaction and between 10% and 15% of the variance in life satisfaction. Partner effects (which were largest for Agreeableness, Conscientiousness, and Emotional Stability) accounted for between 1% and 3% of the variance in relationship satisfaction and between 1% and 2% of the variance in life satisfaction. Couple similarity consistently explained less than .5% of the variance in life and relationship satisfaction after controlling for actor and partner effects.



That research has 2 weak points:

1) The Big 5 test to assess personality of spouses. The Big 5 test is an oversimplification. One extravert (a bold, fearless, high-energy type) may differ considerably from another (a sweet, warm, sensitive type), depending on the extraversion-related primary scale score patterns.
I had been suggesting to use the 16PF5 (or 15FQ+) normative test since years and discard the Big 5 test forever.

2) How they calculate different similarity indices at page 5.
"One important issue in providing a clear interpretation of the similarity literature is that there are a variety of ways to measure similarity, and the choice of method involves both theoretical concerns and statistical complexities (see Chapter 12 of Kenny, Kashy, & Cook, 2006, for a complete discussion)." at page 3.
"Perhaps the most important novel contribution of this work is the systematic examination of the ways that various similarity indices are associated with relationship and life satisfaction. In past work, researchers have debated which index should be used when examining the association between similarity and relationship outcomes." at page 11

"Methodological and data analytic advances in the study of interpersonal relationships: Introduction to the Special Issue" says ".. It is also important for investigators to challenge statisticians to create new analytic techniques when existing ones are inadequate. These tasks are left to you, the reader."



Researchers in the Personality Based Recommender Systems arena are also testing different formulas to calculate similarity, useless at all because they use the Big5 to assess personality of users.
You can see:
"Using Personality Information in Collaborative Filtering for New Users" "Design and User Issues in Personality-based Recommender Systems" "Emotive and Personality Parameters in Multimedia Recommender Systems" "Personality based user similarity measure for a collaborative recommender system" "The LDOS-PerAff-1 Corpus of Face Video Clips with Affective and Personality Metadata" "Addressing the New User Problem with a Personality Based User Similarity Measure" "Improving the believability in the interaction of synthetic virtual agents: Towards Personality in Group Dynamics"

Book
"Recommender Systems based on Personality Traits: Could human psychological aspects influence the computer decision-making process?"

Moreover, this is the list of milestone papers not even cited in
"Predicting relationship and life satisfaction from personality in nationally representative samples from three countries: the relative importance of actor, partner, and similarity effects.":

* Charania & Ickes (2009) paper: "Personality influences on marital satisfaction: Integrating the empirical evidence using the Actor-Partner Interdependence Model (APIM) model"
"... substantial level of inter-partner personality similarity for seven of the thirteen personality traits studied, with four of the similarity correlations exceeding 0.38 ..."

* Rammstedt & Schupp (2008) paper: "Only the congruent survive - Personality similarities in couples. Personality and Individual Differences"
".... Results reveal that among the Big Five dimensions, there are strong differences in spouses' congruences. While for Extraversion and Emotional Stability, congruence is close to zero, correlations averaging at 0.30 are found for Agreeableness, Conscientiousness, and Openness."

* Barelds & Dijkstra (2008) paper: "Do People Know What They Want: A Similar or Complementary Partner?"
"In The Netherlands, where this study was conducted, almost 40% of the divorcees report mismatches in personalities as the major cause of their break-up (De Graaf, 2006; Amato and Previti, 2003). .... although several studies have revealed similarities between partners in their personalities (e.g., Buss, 1984; McCrae, Martin, Hrebícková, Urbánek, Boomsma et al., 2008) only few studies have investigated the extent to which similarity in personality leads to romantic attraction (Barelds and Dijkstra, 2007). From their finding that couples across age groups show the same partner similarities (McCrae et al. 2008) conclude that mate selection, rather than convergence over time, accounts for personality similarity among partners." "Finally, the present study explored a recent issue uncovered by Eastwick and Finkel 2008; also Kurzban and Weeden, 2007; Todd, Penke, Fasolo, and Lenton, 2007 who found that people often report partner preferences that are not compatible with their choices in real life."

* McCrae, Martin, Hrebícková, Urbánek, Boomsma et al. (2008) paper: "Personality Trait Similarity Between Spouses in Four Cultures"
"... Most assortment effects were small, but correlations exceeding 0.40 were seen for a subset of traits, chiefly from the Openness and Agreeableness domains. ... This suggested that mate selection, rather than convergence over time, accounted for similarity"

* Barelds & Dijkstra (2007) paper: "Love at first sight or friends first? Ties among partner personality trait similarity, relationship onset, relationship quality, and love"
"... partner personality trait similarity was related to relationship quality as a function of both relationship onset and specific personality traits. "

* Bekkers, van Aken & Denissen (2006) paper: "Social Structure and Personality Assortment Among Married Couples"
"... Personality characteristics like agreeableness and neuroticism are good predictors of marital conflicts and ultimately of union dissolution, even across different relationships (Robins, Caspi & Moffitt, 2002). .... In sum: spouses with higher levels of neuroticism and openness, spouses with lower levels of agreeableness, and couples with more dissimilar personalities at the time of marriage are more likely to divorce."

Although none of the above papers use the 16PF normative personality test (they mostly use different versions of the normative Big5 personality test instead) and linear or logistic multivariate regression equations to calculate similarity, they clearly show a connection between personality similarity and marital happiness / dyadic success (stability and satisfaction) for some persons.


That is why I had invented a quantitative method to solve the problem of similarity!


The 16PF5 normative personality test codifies personality with 16 variables taking integer values from 1 to 10 in stens (standardized tens).

(A) Warmth; (B) Reasoning; (C) Emotional Stability; (E) Dominance, (F) Liveliness; (G) RuleConsciousness; (H) Social Boldness; (I) Sensitivity; (L) Vigilance; (M) Abstractedness; (N) Privateness (O) Apprehension; (Q1) Openness to Change; (Q2) SelfReliance; (Q3) Perfectionism; (Q4) Tension.


The ensemble (whole set of different valid possibilities) of the 16PF5 is: 10E16, big number as All World Population is nearly 6.7 * 10E9



The 16PF5 result is a quantized pattern, like
John's 16PF5 Profile A:6.B:7.C:6.E:8.F:9.G:6.H:7.I:7.L:8.M:7.N:2.O:5.Q1:8.Q2:7.Q3:3.Q4:4
Lucy's 16PF5 Profile A:5.B:7.C:4.E:8.F:7.G:4.H:5.I:6.L:4.M:6.N:8.O:9.Q1:6.Q2:8.Q3:4.Q4:4

I had invented a high precision quantitative method to compare similarity between those quantized patterns, e.g.:

John's pattern is 74.79865772% +/- 0.00000001% similar to Lucy's.


Latest Research in Theories of Romantic Relationships Development outlines: compatibility is all about a high level on personality similarity* between prospective mates for long term mating with commitment.
*Similarity is a word that has different meanings for different persons or companies, it exactly depends on how mathematically is defined.



Also several studies show that pill users make different mate choices, on average, compared to non-users.

1) "Human oestrus" Gangestad & Thornhill (2008)
"Only short-term but not long-term partner preferences tend to vary with the menstrual cycle"

2) "Does the contraceptive pill alter mate choice in humans?" Alvergne & Lummaa (2009)
".. whereas normally cycling women express a preference for MHC (Major Histocompatibility Complex) dissimilarity in mates, pill users prefer odours of MHC-SIMILAR men, indicating that pill use might eliminate adaptive preferences for genetic dissimilarity."






Please see how the1stMap distorts because it shows you Greenland as big as South America
But see the2ndMap, constructed using the Hobo Dyer Equal Area Projection, it correctly shows
South America: Area 17,840,000 km2 (6,890,000 sq mi)
and
Greenland: Area 2,166,086 km2 (13th) 836,109 sq mi


Fig. 1 extracted from Personal Relationships, 6 (1999), 519-534.
"Taxometric and related methods in relationships research"
Fig. 1 shows how a variable can be masked by other variable.


Please see "An exercise of similarity"

Sunday, November 21, 2010

South America




Some maps distorts because they show you
Greenland as big as South America

South America: Area
17,840,000 km2
(6,890,000 sq mi)

Greenland: Area
2,166,086 km2 (13th)
836,109 sq mi



South America is not the end of the world, it is the beginning of everything.

Friday, November 19, 2010

"The state of Internet in Argentina"

summary about the "The state of Internet in Argentina" from Comscore


* 13 million unique users -over 15 years old Home/Work Location- per month, great demand for news. (September 2010)
* The main activities of the Argentines on the Web is search, the use of social networking and e-mail.
* Unlike what happens in other markets, the country has high levels of demand for news and information on the Internet.
* An average Internet user in Argentina spent 25 hours online during September, consuming 1,810 pages of content and averaging 54 online visits during the month.
* Argentines spent more than 30 percent of their online time staying connected on Instant Messengers and Email sites.
* Microsoft Sites ranked as the most-visited Internet property with 12.2 million visitors, up 24 percent, followed by Google Sites (includes YouTube) with 11.6 million visitors, up 18 percent. Facebook ranked third with 10.7 million visitors, climbing 54 percent, as the social networking destination continued gain fans across the Latin America region.
* 7 of 10 Internet users in Argentina visited the News/Information category in September, with Grupo Clarín and Grupo La Nación leading as the most-visited destinations in the category.
* 25 percent of Argentines online visited a Travel site in September with local brand Despegar ranking as the top site.
* 3 out of 5 Argentines visited an online retail site in September with MercadoLibre leading the way as the top retail-oriented destination in the country.

---------------------------------------------



Grupo Clarín owns the online dating site "Te Busco"
Grupo La Nación owns the online dating site "Zonacitas"
Both are low quality free/cheap online dating sites, performing mostly as search engines.

Match/Chemistry?
Meetic/Meetic Affinity?
eHarmony?
Be2?

They do not have traction or not have the traction they deserve here, in Argentina or other Latin American Country.
The Online Dating Market for serious daters remains enormous.

Wednesday, November 17, 2010

"Ten Questions Internet Execs Should Ask & Answer"

"Ten Questions Internet Execs Should Ask & Answer" presentation from Morgan Stanley


slide #4 Top Internet Markets – 46% of Users in 5 Countries – China, USA, Brazil, India, Russia
Brazil is a Latin American country.
Match, Meetic, Be2 and others had been spending millions of U.S. Dollars since years and until now they do not understand the Latin American culture.


slide #17 Facebook's 620MM Users + Under-Monetized Like Connections Offer Significant New Ad Opportunities

slide #19: Online Ad CPM Dislocation?
A CPM could be very low on Social Networking but I think the most interesting indicator is CTR


slide #28: Top Global 15 Publicly Traded Internet Companies by Market Value – 2010 vs. 2004
(IAC/Interactive USA sure suffered from lack of innovation)



slide #30: "Apple literally was failing when Steve went back and re-infused the innovation and risk- taking that have been phenomenal"

That shows it is not *team work*, sometimes you need a brilliant mind that behaves like a dictator, a tyrant, a despot.


slide #42: USA Federal Government – Entitlement Spending + Interest Expense are Forecast to Exceed Revenue by 2025, per Congressional Budget Office

I think a major economic crisis is going to unleash in the Northern Hemisphere much earlier.

Monday, November 15, 2010

Psychological-based Recommenders Systems

"Towards to Psychological-based Recommenders Systems: A survey on Recommender Systems"



Recommender systems (a.k.a recommendation engines) can be based on:

- past actions (as the formely Beacon at Facebook)

- a pattern of personal preferences ( by collaborative filtering, as the actual one at Facebook) The main disadvantage with recommendation engines based on collaborative filtering is when users instead of providing their personal preference try to guess the global preference and they introduce bias in the recommendation algorithm.

- personality traits of users.

Personality Based Recommender Systems are the next generation of recommender systems because they perform FAR better than Behavioural ones (past actions and pattern of personal preferences)


Have you seen Personality Based Recommender Systems need to calculate personality similarity between users?

Have you seen there are different formulas to calculate similarity?
Recommender systems are morphing to .......... compatibility matching engines!!!

That is nothing new, nothing innovative. Online Dating Sites like eHarmony, Parship, Be2, MeeticAffinity and others had been calculating personality similarity between prospective users since several years ago with low successful rates, with a low effectiveness/efficiency level of their matching algorithms (less than 10%) because they use the normative Big5 or ipsative proprietary models instead -like Chemistry or PerfectMatch- to measure personality traits.

No one is using the 16PF5 to assess personality of members.
No one calculates similarity with a quantized pattern comparison method.
No one can show Compatibility Distribution Curves to each and every of its members.


Please do not think I am rude or not polite, or do not think I am hammering the "personality similarity" concept in your head, but
Can you see where the Online Dating Industry for serious daters needs to go?


There is only one road, the road of:
the 16PF5, 15FQ+ or similar to assess personality of members
and
quantized pattern comparison method to calculate similarity.

Wednesday, November 10, 2010

Is Meetic in the same road as Be2???

Total marketing investments since the start of the year stand at 79.3 million of Euros, or 58% of 9-month revenue!!!
Net profit as % of revenue for 9 months of 2010 is 8.3% compared to
Net profit as % of revenue for 9 months of 2009 is 11.2%


Marc Simoncini, CEO of Meetic, concludes: "In 2011, Meetic ... Group intends to make the most of the boom in Smartphone services to establish itself on this new media form. ... the Group will continue to heavily invest on the Matchmaking segment ... Lastly, the Group will continue to launch new sites in order to cover segments as yet not utilised by its main brands."

ALL competitors are doing exactly the SAME. Where is the innovation?


Is Meetic in the same road as Be2???

Meetic and premium services as MeeticAffinity, MatchAffinity, DatingDirectAffinity, PartnerDE are in decadence since Q1 2009, losing traffic and customers.



Parperfeito is also in decadence.


MEET PA shares
21 OCT 2007 Euros 31.43 and dropped to
17 OCT 2008 Euros 8.99 then recovered to
09 NOV 2010 Euros 21.90


Perhaps the CEO knows Meetic can not innovate to survive and he wants to convert all his shares in cash as soon as possible; before Meetic dies.

22.805.260 shares * 27% stake * Euros 20.00 (average?) == 123.15 million of Euros
22.805.260 shares * 27% stake * Euros 8.00 (low?) == 49.26 million of Euros

Friday, November 5, 2010

Online Dating Industry summary 2010

Lack Of Innovation & Decadence can summarize the Online Dating Industry 2010


- Lack Of Innovation, the big dinosaurs
* Match/Chemistry
During the tenure of Tim Sullivan as CEO, Match entered the compatibility matching era with WeAttract's test.
During the tenure of Jim Safka, the IAC (Match) launched Chemistry (the last innovation at Match, it was during 2005).
During the tenure of Thomas Enraght-Mooney, the IAC acquired 27% stake in Meetic but lost the war in Europe.
During the tenure of Greg Blatt, the IAC bought some online dating sites like PeopleMedia's Communities, SinglesNet and YahooPersonals unit.

* eHarmony, a 10 years old obsolete site.

* Meetic (Meetic Affinity)

- tried to innovate this year
* FriendScout-24 with the BQ-Check (a hoax)
* PlentyOfFish, tried to innovate with the POF Marriage Predictor (and failed)
* Personality-based recommender systems
Have you seen the next generation of recommender systems include normative personality traits?
That is the only way to improve recommender systems, to include the personality traits of their users.
Have you seen they need to calculate personality similarity between users?
Have you seen there are different formulas to calculate similarity?
In case you did not notice, recommender systems are morphing to .......... compatibility matching engines!!!

- failed proposals:

* 12Like
* LittleHint

- free sites: are only useful to send prospective customers to paid sites and paid sites send customers to Offline Chains.
* OKCupid
* PlentyOfFish, tried to innovate with the POF Chemistry Predictor and the POF Marriage Predictor and failed
* OasisActive / AmorEnLinea, getting some traction in Latin American countries


- can not take off:
* eVow, the paid site from PlentyOfFish
* Jazzed the "free" site from eHarmony
* eHarmonyBrazil

* eCift in Turkey (a copycat of Parship)


- in decadence
* Be2 (game over)
* Parship (will be game over)
* ParPerfeito
* PerfectMatch

- White label proposals, they do not have weigh in the Online Dating Industry for serious daters.

----------------------------


WorldWide, there are over 5,000 (five thousand) online dating sites; "dating" is a broad word to define a large group of different proposals.
Online Dating Sites can be classified as:
1.0: "Browsing/Searching Options, Powerful Searching Engines"
1.5: "Unidirectional Recommendation Engines"
2.0: "Matching based on Self-Reported Data / Bidirectional Recommendation Engines"
3.0: "Compatibility Matching Algorithms"

All major actual paid Online Dating sites are old and in decadence losing traffic since January 2009. It is not due to the traction of Facebook, it is the Online Dating Fatigue Phenomenon due to Obsolescence of Dating Sites offering Compatibility Matching Methods.
Social networking/applications could merge with online dating for fun, for flirting, for entertainment purposes, for instant gratification.
Social networking and online dating for serious daters are like water and oil, they will never mix.

Owners and C-level executives of major online dating sites are more worried in attracting, converting and retaining subscribers with automatic renewal of their subscriptions and other credit-card billing trickery, than offering a good compatibility matching method.

The entire Online Dating Industry for serious daters in 1st World Countries is a HOAX, performing as a Big Online Casino, with a low effectiveness/efficiency level of their matching algorithms (less than 10%), and new proposals are not better than the old ones.

Actual online dating sites offering compatibility matching methods are only fueled by big marketing budgets and not by serious scientific evidence. No online dating site can prove its matching algorithm can match prospective partners who will have more stable and satisfying relationships than couples matched by chance, astrological destiny, personal preferences, searching on one's own, or other technique as the control group in a peer_reviewed Scientific Paper. They only use different versions of the normative Big5 or ipsative proprietary models instead (like Chemistry or PerfectMatch) to measure personality traits. They are all like placebo, because they are delivering ON AVERAGE 3 to 4 prospective mates as highly compatible persons per 1,000 persons screened, in exactly the same range as searching on one's own, with a high percentage of false positives, like gun machines shooting flowers.


Online dating sites should had killed offline dating proposals (chains, professional matchmakers, etc) since some years ago, but offline proposals are still alive and with good health.
Divorce rates are still high.

The Online Dating Industry needs innovations, but the innovations the Online Dating Industry needs will come from only one source: the latest discoveries in theories of romantic relationships development with commitment.


1) What is important in attracting people to one another may not be important in making couples happy.

2) Latest Research in Theories of Romantic Relationships Development outlines: compatibility is all about a high level on personality* similarity* between prospective mates for long term mating with commitment.
*personality measured with a normative test.
*similarity: there are different ways to calculate similarity, it depends on how mathematically is defined.


----------------------------------


WorldWide, there are 5,000 -over five thousand- online dating sites
but no one
is using the 16PF5 to assess personality of its members!
but no one
calculates similarity with a quantized pattern comparison method!
but no one
can show Compatibility Distribution Curves to each and every of its members!
but no one
is scientifically proven!

Thursday, November 4, 2010

Clicck / Holistic Approach

I had secretely reviewed Clicck some time ago:


It is based on 2 terrible mistakes.


1st mistake:
"Clicck's current questionnaire based on Big Five theory and the Neo is being reworked to generate a more in-depth review into what makes a potential mate tick, providing users with a much clearer and potentially more successful match."


based on Big Five theory ???

The ENSEMBLE (whole set of different valid possibilities) of the primary domains of the Big5 is 10E5 (quantized from 1 to 10) and 10E10 (by percentile)

The ensemble of the 16PF5 is: 10E16, big number as All World Population is nearly 6.7 * 10E9


The Big 5 traits personality model is good for orientative purposes but not good enough for predictive purposes.


"Because the Big Five groups the more specific primary-level factors, feedback organized around the five Global Factor scales is more easily understood. For detailed feedback or predictive purposes, one should assess the more specific primary factors. Research has shown that more specific factors like the primary scales of the 16PF Questionnaire predict actual behavior better than the Big 5 Global Factors. For example, one extravert (a bold, fearless, high-energy type) may differ considerably from another (a sweet, warm, sensitive type), depending on the extraversion-related primary scale score patterns, so deeper analysis is typically warranted."
Extracted from the 16PF5 Manual

The 1st version of the 16PF test was invented in 1949 and no actual Online Dating site is using it. (Now in the 5th version, the 16PF5 actualized after year 2000)
MORE than 60 years of experience with the 16PF test!!!


--------------------------------------
2nd mistake:
"Clicck matches users through a combination of its patent-pending facial recognition software, compatibility test and cutting edge biochemical attraction analysis."


If they plan to use DNA matching for heterosexual couples, they are dead, because:
Basisnote, GenePartner, ScientificMatch and other proposals offering DNA matching are only based on the T_shirt Experiment that only proved:
normally cycling women (not pregnant and not taking contraceptive pills) are (temporarily) attracted by the perspiration scent of clothes used by men with a Major Histocompatibility Complex MHC more dissimilar to theirs,
and not proved: women attracted by those men for long term mating with commitment.

2 Scientific Papers debunk their claims.
1) "Human oestrus" Gangestad & Thornhill (2008)
"Only short-term but not long-term partner preferences tend to vary with the menstrual cycle"

2) "Does the contraceptive pill alter mate choice in humans?" Alvergne & Lummaa (2009)
".. whereas normally cycling women express a preference for MHC (Major Histocompatibility Complex) dissimilarity in mates, pill users prefer odours of MHC-SIMILAR men, indicating that pill use might eliminate adaptive preferences for genetic dissimilarity."
................
"Recently, Roberts et al. attempted to eliminate these potential confounds by adopting a within-subject design in which women's mate preferences were assessed before and after they began taking the pill. Women starting the pill showed a significant preference shift towards MHC SIMILARITY compared with three months before the pill was taken, a shift that was not observed in the control group of normally cycling women."


------------------------------

What is important in attracting people to one another may not be important in making couples happy.

Latest Research in Theories of Romantic Relationships Development outlines: compatibility is all about a high level on personality* similarity* between prospective mates for long term mating with commitment.
*personality measured with a normative test.
*similarity: there are different ways to calculate similarity, it depends on how mathematically is defined.